Description:BACKGROUND
Technical Field
[0001] Embodiments of this disclosure generally relate tothe field of sports safety systems and automated sporting equipment, and more particularly, tosystems and methods for detecting the presence and proper use of safety gear by a player in a cricket environment using computer vision, and for controlling an automated cricket bowling machine based on the detection results.
Description of the Related Art
[0002] Indoor cricket training facilities often employ automated bowling machines to simulate realistic playing conditions for a batsman. These machines can deliver cricket balls at high velocities with considerable accuracy, offering players an effective and consistent method for honing their batting skills. However, the use of such automated machinery inherently introduces safety risks, especially in controlled indoor environments where response times are limited and supervision may be intermittent or inconsistent.
[0003] Historically, the enforcement of player safety protocolssuch as mandatory use of helmets, leg pads, gloves, and other protective gearhas relied on manual observation by coaches or facility staff. Similarly, detection of hazards such as obstructions on the playing lane, player falls, or dangerous impacts from cricket balls has traditionally been a reactive process, often identified only after an injury or incident occurs. These manual systems are not only prone to human error but also fail to provide the immediacy and objectivity necessary to maintain consistent safety standards.
[0004] While advances in computer vision and machine learning have found some applications in sports analytics, their integration into automated safety systems, especially in the context of cricket training,remains underdeveloped. Existing systems are largely incapable of dynamically monitoring safety gear compliance in real-time, detecting player falls or concussive impacts, or taking proactive control of training equipment such as bowling machines to mitigate injury risks.
[0005] Accordingly, there exists a need for a system that can continuously and autonomously monitor indoor cricket training lanes using visual data, analyze player safety compliance and potential hazards in real-time using artificial intelligence, and respond immediately by disabling automated equipment and notifying relevant personnel to prevent injury. Such a system would ideally reduce the reliance on human supervision, enhance the safety of athletes, and provide data-driven insights to inform ongoing training and risk mitigation efforts.
SUMMARY
[0006] In view of foregoing, an embodiment herein provides a system for automatically determining whether a batsman is wearing at least one required safety gear and controlling anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gear. The system comprises an edge device operatively connected to one or more cameras and the automated cricketbowling machine. The edge device comprises a processor anda non-transitory memory storing instructions that, when executed by the processor, cause the edge device toreceive one or more real-time images from theone or morecameras. The one or more camerasis configured to capture the one or morereal-time images of a cricket lane at one or more angles. The processor processes the one or more real-time images to extract depth information for accurate object localization. The processorapplies a computer vision model to the one or more real-time images that are processed to detect the presence and positioning ofone or more safety gears on the batsman. The computer vision model is trained to (i)identify objects corresponding to the one or moresafety gears comprising a helmet, hand gloves, shoes, and leg pads, and (ii) determine a spatial relationship between the one or more safety gears that are detected and the batsman by(a) detecting key points on the batsman's body, (b) associating the one or more safety gears that are detected with one or morespecific body regions comprising head, hands and legs of the batsman based on predefined positional criteria, and(c) verifying that each safety gear is properly worn by confirming that the helmet is on the head, thehand gloves are on the hands, and the leg pads are on the legs of the batsman. The processor automatically determines, based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman, whether the batsman is wearing all required safety gears, and automatically controls the operation of the automated cricket bowling machine to prevent the delivery of a ball if the batsman is not wearing at least one required safety gear.
[0007] The system proactively enforces safety compliance by verifying whether a batsman is wearing essential protective gear such as the helmet, thehand gloves, or the leg pads before allowing ball delivery. This automated decision-making by the system eliminates human intervention and ensures that player protection is prerequisite to commencing play. By integrating image capture with depth estimation and object localization through the computer vision model, the system provides high accuracy in detecting safety gear presence and placement in real-time, thus minimizing injury risk while maintaining uninterrupted training sessions.
[0008] In some embodiments,the edge device is configured to detect a fall event by(i) analyzing a posture of thebatsman using a pose estimation technique, (ii) assessing depth information to determine a position of the batsman relative to a playing surface of the cricket lane, and (iii) identifying a fall when the posture of thebatsman deviates from an upright stance and the depth information indicates a proximity to the playing surface of the cricket lane.In some embodiments,upon detection of a fall event of the batsman, the edge device is configured toimmediately (i) stop the operation of the automated cricket bowling machine to prevent thedelivery of the ball, (ii) activate an audible alarm to alertnearby personnel to indicate the detection of the fall event of the batsman, and(iii) generate a notification indicating the detection of the fall event of the batsmanto a designated monitoring system or a device associated with an individual.
[0009] The system detects player falls or collapse events through a multi-model pipeline combining object detection, human pose estimation, and temporal action recognition. This allows for real-time identification of abnormal posture transitions or collapse patterns that may indicate injury or medical emergency. This real-time fall detection ensures that the automated cricket bowling machine halts automatically, and alerts are raised immediately, thereby enabling swift intervention by coaches or medical staff. This feature is important in mitigating the risk of concussion-related complications or unattended injuries during practice sessions.
[0010] In some embodiments,the edge device is configured to detect an obstruction or obstacle within a predefined safety zone of the cricket lane by(i) employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone, (ii) employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone, and (iii) determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone for a duration exceeding a predefined threshold.
[0011] In some embodiments,upon detection of an obstruction, the edge device is configured to(i) stop the operation of the automated cricket bowling machine to prevent thedelivery of the ball, (ii) generate a visual or an audible alert to inform players and coaching staff of the obstruction, and(iii) enable normal operation of the automated cricket bowling machine only after the obstruction is cleared and the predefined safety zone is confirmed to be free of unauthorized objects or individuals. The system ensures pitch area safety through the detection of foreign objects or obstructions within a designated danger zone in the cricket lane. Using the depth mapping, the system enables detection of loose balls, equipment, or unauthorized movement in real time. This feature enhances player safety by preventing accidental tripping, collisions, or misdirected deliveries, and avoids damage to the automated cricket bowling machine. Furthermore, the systemenables operators to dynamically define safety zones through configurable parameters, and allows adaptation to different pitch layouts or training scenarios.
[0012] In some embodiments, the edge devicegenerates an alert indicating the absence of required safety gear.The system ensures not just helmet detection but verifies proper wearing position using positional algorithms to avoid false positives from misplaced helmets. This improves player safety and system reliability by preventing play without proper gear.
[0013] In some embodiments, the edge device comprises a network interface that is configured totransmit data collected by the edge device to a cloud-based platform for storage and analysis, receive software updates and configuration settings from the cloud-based platform, andenable remote monitoring and control of the system via the cloud-based platform.
[0014] In some embodiments, the computer vision model utilizes a segmentation model that is optimized for detection of the one or more safety gears in cricket environments. The computer vision model is a deep learning model trained using a custom dataset specific to indoor or outdoor cricket environments.In some embodiments, the edge device is configured to log instances of missing safety gear for subsequent review and analysis. The edge device comprises a user interface for monitoring a status of the system and configuring operational parameters of the system.
[0015] In another aspect, acomputer-implemented method for automatically determining whether a batsman is wearing at least one required safety gear and controlling anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gear is provided. The method comprises receiving, by an edge device, one or more real-time images from one or morecameras, the one or more of camerasis configured to capture the one or morereal-time images of a cricket lane at one or more angles. The edge deviceis operatively connected to theone or morecameras and the automated cricketbowling machine. The method comprisesprocessing, by the edge device, the one or more real-time images to extract depth information for accurate object localization. The method comprisesapplying, by the edge device, a computer vision model to the plurality of real-time images that are processed to detect the presence and positioning ofone or more safety gears on the batsman.Applying the computer vision model comprises (i)identifying objects corresponding to the one or moresafety gears comprising a helmet, hand gloves, shoes, and leg pads, and (ii) determininga spatial relationship between the plurality of safety gears that are detected and the batsman by(a) detecting key points on the batsman's body, (b) associating the one or more safety gears that are detected with one or morespecific body regions comprising head, hands and legs of the batsman based on predefined positional criteria, and (c) verifying that each safety gear is properly worn by confirming that the helmet is on the head, thehand gloves are on the hands, and the leg pads are on the legs of the batsman. The method comprises automatically determining, based on the analysis of the spatial relationshipbetween the one or more safety gears that are detected and the batsman, whether the batsman is wearing all required safety gears, and automatically controlling, by the edge device, the operation of the automated cricket bowling machine to prevent the delivery of a ball if the batsman is not wearing at least one required safety gear.
[0016] The method offers significant advantages by enhancing player safety through real-time monitoring and intelligent control. Leveraging the cameras and the edge computing, the method uses advanced computer vision models to detect the presence and correct positioning of essential safety gearsuch as the helmets, the hand gloves, and the leg padsas well as identifying falls, obstructions, and dangerous objects within the pitch/cricket lane. If safety conditions are unmet, the methodproactively halts the automated cricket bowling machine, issues audio alerts, and notifies coaching staff, thereby reducing injury risks and liability. Its automation minimizes human error and ensures consistent safety enforcement, improving training efficiency. The system supports modular expansion, cloud integration, and remote monitoring. The system employs affordable hardware and software while helping prevent costly injuries and legal exposure, and makes it a robust and adaptive solution for indoor cricket safety.
[0017] In some embodiments, the method comprisesdetecting a fall event by(i) analyzing a posture of thebatsman using a pose estimation technique, (ii) assessing depth information to determine a position of the batsman relative to a playing surface of the cricket lane, and (iii) identifying a fall when the posture of thebatsman deviates from an upright stance and the depth information indicates a proximity to the playing surface of the cricket lane.
[0018] In some embodiments,upon detection of a fall event of the batsman, the method comprises (i)immediately stopping the operation of the automated cricket bowling machine to prevent thedelivery of the ball, (ii) activating an audible alarm to alertnearby personnel to indicate the detection of the fall event of the batsman, and(iii) generating a notification indicating the detection of the fall event of the batsmanto a designated monitoring system or a device associated with an individual.
[0019] In some embodiments, the method comprisesdetectingan obstruction or obstacle within a predefined safety zone of the cricket lane by(i) employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone, (ii) employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone, and (iii) determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone for a duration exceeding a predefined threshold.
[0020] In some embodiments, upon detection of an obstruction, the method comprises(i) stopping the operation of the automated cricket bowling machine to prevent thedelivery of the ball, (ii) generating a visual or an audible alert to inform players and coaching staff of the obstruction, and (iii) enabling normal operation of the automated cricket bowling machine only after the obstruction is cleared and the predefined safety zone is confirmed to be free of unauthorized objects or individuals.In some embodiments,the method comprisesgenerating an alert indicating the absence of required safety gear.
[0021] In some embodiments, the method comprisesproviding a network interface on the edge device that is configured to(i) transmit data collected by the edge device to a cloud-based platform for storage and analysis, (ii) receive software updates and configuration settings from the cloud-based platform, and(iii) enable remote monitoring and control of the system via the cloud-based platform.
[0022] In some embodiments,the computer vision model utilizes a segmentation model that is optimized for detection of the one or more safety gears in cricket environments. The computer vision model is a deep learning model trained using a custom dataset specific to indoor or outdoor cricket environments.
[0023] These and other aspects of the embodiments herein will be better appreciated and understood when considered in conjunction with the following description and the accompanying drawings. It should be understood, however, that the following descriptions, while indicating preferred embodiments and numerous specific details thereof, are given by way of illustration and not of limitation. Many changes and modifications may be made within the scope of the embodiments herein without departing from the spirit thereof, and the embodiments herein include all such modifications.

BRIEF DESCRIPTION OF THE DRAWINGS
[0024] The embodiments herein will be better understood from the following detailed description with reference to the drawings, in which:
[0025] FIG. 1illustrates a system view of a system that automatically determines whether a batsman is wearing at least one required safety gear and controls anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gearaccording to an embodiment herein;
[0026] FIG. 2 illustrates a block diagram of an edge device of the system of FIG. 1 that automatically determines whether a batsman is wearing at least one required safety gear and controls anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gearaccording to an embodiment herein;
[0027] FIG. 3 illustrates a method of automatically determining whether a batsman is wearing at least one required safety gear and controlling anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gearaccording to an embodiment herein;
[0028] FIG. 4illustrates an exemplary view of the system of FIG. 1 thatis implemented in a cricket lane and automatically determines whether a batsman is wearing at least one required safety gear and controls anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gear according to an embodiment herein;
[0029] FIG. 5 illustrates an exemplary view of the system of FIG. 1 that detects whether a batsman 414 is wearing a helmetusing an object detection techniqueand a helmet detection techniquethat are implemented on the edge device according to an embodiment herein;
[0030] FIG. 6 illustrates a block diagram of the edge device of FIG. 1 that automatically controls anautomated cricket bowling machine to prevent delivery of a ball according to an embodiment herein; and
[0031] FIG. 7 illustrates an exemplary view of the edge deviceof FIG. 1 that identifiesan unauthorized objector individual within apredefined safety zoneby employingan object detection techniqueaccording to an embodiment herein; and
[0032] FIG. 8 is a representative hardware environment for practicing the embodiments herein with respect to FIG. 1 through 7 in accordance with the embodiments herein.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
[0033] The embodiments herein and the various features and advantageous details thereof are explained more fully with reference to the non-limiting embodiments that are illustrated in the accompanying drawings and detailed in the following description. Descriptions of well-known components and processing techniques are omitted so as to not unnecessarily obscure the embodiments herein. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.
[0034] As mentioned, there remains a need for a system and a method that automatically determines whether a batsman is wearing at least one required safety gear and controls anautomated cricket bowling machine to prevent delivery of a ball if the batsman is not wearing the required safety gear.Referring now to the drawings, and more particularly to FIGS. 1 through 8, where similar reference characters denote corresponding features consistently throughout the figures, there are shown preferred embodiments.
[0035] A cricket pitch is a rectangular strip located at the center of the field. It typically measures 22 yards (20.12 meters) in length and 10 feet (3.05 meters) in width. It is made of hard-packed clay covered with short grass. The pitch includes stumps (wickets) at each end. It is flat and serves as the focal point of the game, where bowlers deliver the ball and batsman aim to protect their wickets while scoring runs.
[0036] Bowlers use different lengths strategically depending on the situation. A full-length delivery or yorker lands very close to the batsman on the cricket pitch, usually near their feet, and is designed to target the base of the stumps. A good length balllands about 6–8 meters in front of the batsman on the cricket pitch, and creates maximum uncertainty and forcing the batsman to choose between playing forward or back. Short-length deliveries land in the middle of the cricket pitch, and causes the ball to bounce higher to prompt a back-foot response from the batsman. A bouncer is an even shorter delivery that rises to chest or head height and is used for intimidation or to induce errors. The half-volley lands just in front of the batsman and makes it ideal for aggressive, attacking strokes.
[0037] There are several fundamental differences between cricket bowling and baseball pitching. First, cricket bowlers keep their arms straight during delivery, whereas baseball pitchers bend their elbows. In cricket, the ball is deliberately bounced on the pitch before reaching the batsman. Cricket bowlers generate speed using a run-up that can exceed 30 meters, and the movement is created through seam positioning, i.e., the shiny surface of the ball (which aids swing), and interaction with the pitch. Cricket includes a wide variety of bowling styles such as fast bowling, medium-pace, off-spin, and leg-spin. A single cricket ball is used for many overs, sometimes up to 80 overs, and its condition mainly impacts play. Moreover, Cricket bowlers can deliver from a wide area relative to the wicket, thereby providing tactical versatility. Cricket bowling demands continuous strategic adjustments in line, length, and pace to exploit both the batsman’s weaknesses and the pitch conditions.
[0038] The nature of the pitch plays animportant role in the effectiveness of bowling. A green pitch, which has more grass coverage, benefits fast bowlers by offering extra bounce, quicker movement, and more seam deviation during a match, especially during the first day of atest match. In contrast, dry or dusty pitches tend to favor spin bowlers, as the ball grips the surface and turns sharply. Cracks that develop over time cause unpredictable bounce and deviation, and make it challenging for the batsman. Such conditions often arise in subcontinental venues like India, Pakistan, and Sri Lanka (for example, during the later days of a Test match).
[0039] As matches progress, the pitch deterioration becomes more pronounced. Day one usually supports seamers, days two and three tend to favor batsmen, and days four and five become a spinner’s paradise due to increasing wear and tear. Some teams even "doctor" the pitch to suit their strengths by watering for swing, under-rolling for variable bounce, or over-rolling for flat batting tracks. The dynamic nature of pitch conditions requires teams to adjust bowling strategies and player selection accordingly. Thus, reading the pitch is animportant skill for captains and bowlers alike.
[0040] Indoor cricket surfaces, typically synthetic, introduce unique factors that affect bowling performance. For fast bowlers, bouncers gain consistency and may bounce higher due to uniform surfaces, making their timing more predictable. Yorkers also benefit from a skiddy surface that remains effective throughout the match and is challenging for batsmen to read. Cutters and seamers may find the seam movement more predictable but occasionally sharp, depending on how the seam grips the synthetic material. Side-cutters perform well due to consistent traction.
[0041] Spin bowlers face distinct challenges indoors. Off-spinners or finger spinners experience less natural grip, which requires them to impart more revolutions and focus on pace variations. Wrist spinners, such as leg-spinners, may find their googlies and flippers particularly potent, although they may get less grip overall. Flighted deliveries gain an edge due to the absence of wind, thereby allowing for more control and deception. However, drift tends to be reduced and may be compensated for with enhanced spin and subtle variation.
[0042] Indoor conditions also influence ball deterioration. Unlike outdoor surfaces, synthetic pitches are generally less abrasive, which results in slower wear and affects swing potential. Back-of-the-hand slower balls become more deceptive due to surface predictability. Pace variations are more important than surface exploitation, which makes short-pitched deliveries with subtle changes in speed especially valuable. Climate-controlled environments introduce consistent temperature and humidity, sometimes enhancing conventional swing and enabling bowlers to refine muscle memory for particular variations. The predictability of indoor surfaces places emphasis on a bowler’s technical skill, which requires precise manipulation of grip, wrist position, and release to outwit batsmen.
[0043] FIG. 1 illustrates a system 100 that automatically determines whether a batsman is wearing at least one required safety gear and controls anautomated cricket bowling machine 106 to prevent delivery of a ball if the batsman is not wearing the required safety gear according to an embodiment herein. The system 100 includes an edge device 102 operatively connected to one or morecameras104A-N and the automated cricketbowling machine 106. The edge device102 includesa processor 108 that receivesone or more real-time images from theone or morecameras 104A-N. The one or more cameras104A-N is configured to capture the one or morereal-time images of a cricket lane at one or more angles. The one or more cameras104A-N may includea stereo camera thatcaptures real-time stereo images of the cricket lane to provide depth information important for accurate object localization.The one or more cameras 104A-N may include a LiDAR sensor, a radar, or an optical tracking systemthatcaptures real-time images of the cricket lane to provide depth information important for accurate object localization.
[0044] The processor 108 processes the real-time images to extract depth information for accurate object localization. The processor 108 applies a computer vision model 110 to the processedreal-time imagesto detect the presence and positioning ofone or more safety gears on the batsman. The edge device 102 may be a high-performance computing device, such as an NVIDIA GPU, that processes the real-time images/stereo images and runs the computer vision model 110. Theedge device 102 interfaces with the automated cricket bowling machine 106 and other safety devices to trigger appropriate actions. The computer vision model 110 is trained to identify objects corresponding to the one or moresafety gears comprising a helmet, hand gloves, shoes, and/or leg pads. The computer vision model110 determinesa spatial relationship between the one or more safety gears that are detected and the batsman by (a) detecting key points on the batsman's body, (b) associating the safety gears that are detected with one or morespecific body regions comprising head, hands and legs of the batsman based on predefined positional criteria, and (c) verifying that each safety gear is properly worn by confirming that the helmet is on the head, thehand gloves are on the hands, and the leg pads are on the legs of the batsman. The computer vision model 110 utilizes deep learning techniques that are optimized for real-time inference using TensorRT.
[0045] The processor 108 automatically determines, based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman, whether the batsman is wearing all required safety gears, and automatically controls the operation of the automated cricket bowling machine 106 to prevent the delivery of a ball if the batsman is not wearing at least one required safety gear.
[0046] The system100 enforces safety compliance by verifying whether a batsman is wearing essential protective safety gear, such as the helmet, thehand gloves, or the leg pads, before allowing ball delivery. This automated decision-making by the system 100 eliminates human intervention and ensures that player protection is prerequisite to commencing play. By integrating image capture with depth estimation and object localization through the computer vision model 110, the system 100 provides high accuracy in detecting safety gear presence on the batsman and placement in real-time, thereby minimizing injury risk while maintaining uninterrupted training sessions.
[0047] In some embodiments,the edge device102 is configured to detect a fall event byanalyzing a posture of thebatsman using a pose estimation technique, assessing depth information to determine a position of the batsman relative to a playing surface of the cricket lane, and identifying a fall when the posture of thebatsman deviates from an upright stance and the depth information indicates a proximity to the playing surface of the cricket lane. In some embodiments,upon detection of a fall event of the batsman, the edge device 102 is configured toimmediately stop the operation of the automated cricket bowling machine 106 to prevent thedelivery of the ball. Further, the edge device 102 is configured toactivate an audible alarm to alertnearby personnel to indicate the detection of the fall event of the batsman, and generate a notification indicating the detection of the fall event of the batsmanto a designated monitoring system or a device associated with an individual.
[0048] The system 100 detects player falls or collapse events through a multi-model pipeline combining object detection, human pose estimation, and temporal action recognition. This allows for real-time identification of abnormal posture transitions or collapse patterns that may indicate injury or a medical emergency. This real-time fall detection ensures that the automated cricket bowling machine halts automatically, and alerts are raised immediately, thereby enabling swift intervention by coaches or medical staff. This feature is important in mitigating the risk of concussion-related complications or unattended injuries during practice sessions.
[0049] The system 100 combines multiple models to detect falls, including YOLOv11, AlphaPose for pose estimation, and a spatio-temporal Graph Convolutional Network (ST-GCN) for action recognition. The system 100 is optimized for NVIDIA GPUs using TensorRT for faster inference to ensure rapid and accurate fall detection in real-time applications.
[0050] In some embodiments, the edge device102 is configured to detect an obstruction or obstacle within a predefined safety zone of the cricket lane by employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone, employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone, and determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone for a duration exceeding a predefined threshold.
[0051] In some embodiments,upon detection of an obstruction, the edge device102 is configured to stop the operation of the automated cricket bowling machine 106 to prevent thedelivery of the ball. The edge device102 may generate a visual or an audible alert to inform players and coaching staff of the obstruction. The edge device102 may enable normal operation of the automated cricket bowling machine 106 only after the obstruction is cleared and the predefined safety zone is confirmed to be free of unauthorized objects or individuals. The system 100 ensures pitch area safety through the detection of foreign objects or obstructions within a designated danger zone in the cricket lane. Using the depth mapping, the system 100 enables detection of loose balls, equipment, or unauthorized movement in real time. This feature enhances player safety by preventing accidental tripping, collisions, or misdirected deliveries, and avoids damage to the automated cricket bowling machine 106. Furthermore, the system100 enables operators to dynamically define safety zones through configurable parameters, and allows adaptation to different pitch layouts or training scenarios.
[0052] The system 100 utilizes the YOLOv11 segmentation model, which is fine-tuned on a custom hand-labeled dataset specific to indoor cricket practice environments. The fine-tuning ensures the model accurately detects safety-critical objects in this niche setting. The YOLOv11 segmentation model is optimized for NVIDIA GPUs to achieve high-performance inference using TensorRT to enable real-time object detection with enhanced efficiency.
[0053] In some embodiments, the edge device102 generates an alert indicating the absence of required safety gear.The system 100 ensures not just helmet detection but verifies proper wearing position using positional algorithms to avoid false positives from misplaced helmets. This improves player safety and system reliability by preventing play without proper safety gear.
[0054] In some embodiments, the edge device102 includes a network interface that is configured totransmit data collected by the edge device102 to a cloud-based platform for storage and analysis, and receive software updates and configuration settings from the cloud-based platform. The network interfaceenables remote monitoring and control of the system 100 via the cloud-based platform. The network interface enables data transmission to a cloud-based platform for storage, analysis, and remote monitoring. It also updates configuration settings from the cloud.
[0055] In some embodiments, the computer vision model 110 utilizes a segmentation model that is optimized for the detection ofthe safety gearsof the batsman in cricket environments. The computer vision model 110 is a deep learning model trained using a custom dataset specific to indoor or outdoor cricket environments. In some embodiments, the edge device102 is configured to log instances of missing safety gear for subsequent review and analysis. The edge device102 includesa user interface for monitoring a status of the system 100 and configuring operational parameters of the system 100. The edge device 102 may provide real-time monitoring and configuration capabilities, andallow coaches and staff to oversee system status and adjust operational parameters.
[0056] The system 100 enables remote monitoring and control of multiple lanes from a central location.The system 100 is enabled with data storage, analysis, and system updates through cloud platforms. The system 100 is capable of detecting the any unwanted objects in the pitch or multiple batsman/users in the pitch and automatically stops the automated cricket bowling machine.
[0057] FIG. 2 illustrates a block diagram of the edge device 102 of the system 100 of FIG. 1 that automatically determines whether a batsman is wearing at least one required safety gear and controls anautomated cricket bowling machine 106 to prevent delivery of a ball if the batsman is not wearing the at least one required safety gear according to an embodiment herein.The edge device 102 is operatively connected to one or more cameras104A-N and anautomated cricketbowling machine 106. The edge device102 includes a database 200, a real-time image capturing module 202, a real-time image receiving module 204, a real-time image processing module 206, a safety gear detection module 208, a spatial relationshipdetermination module 210 and aballdelivery prevention module 212.
[0058] The real-time image capturing module 202 capturesone or more real-time images of a cricket lane at one or more angles using the one or more cameras104A-N. The one or more cameras 104A-N may include at least one stereo camera thatcaptures real-time stereo images of the cricket lane to provide depth information important for accurate object localization.The real-time image receiving module 204 receives thereal-time images from thereal-time image capturing module 202.The real-time image processing module 206 processes the real-time images to extract depth information for accurate object localization.
[0059] The safety gear detection module 208 detects the presence and positioning of one or more safety gears on the batsman by applying a computer vision model 110 to the processedreal-time.The edge device 102 may process the real-time images/stereo images and runs the computer vision model 110.
[0060] The spatial relationshipdetermination module 210 determinesa spatial relationship between the one or more safety gears that are detected and the batsman by(i) detecting key points on the batsman's body, (ii) associating the one or more safety gears that are detected with one or more specific body regions comprising head, hands and legs of the batsman based on predefined positional criteria, and(iii) verifying that each safety gear is properly worn by confirming that the helmet is on the head, thehand gloves are on the hands, and the leg pads are on the legs of the batsman. The spatial relationshipdetermination module 210 utilizes deep learning techniques that are optimized for real-time inference using TensorRT.
[0061] In some embodiments, the deep learning techniques includes Convolutional Neural Networks (CNNs), Object Detection Models, Pose Estimation Models, Graph Neural Networks (GNNs), or Transformer Networks. The customized CNNs may be trained on a dataset specifically featuring cricket environments, players, and safety gear, learning to identify relevant visual features such as shapes, textures, and colors of helmets, pads, gloves, and other equipment. Optimization with TensorRT may involve fusing layers, reducing precision, and selecting the most efficient kernels to ensure rapid feature extraction from video frames for real-time processing.
[0062] In some embodiments, the Object Detection Models are specifically trained to detect and precisely locate bounding boxes around various cricket safety gear (e.g. helmets, pads, gloves, etc.) and potentially players within diverse cricket scenarios (e.g. different lighting, camera angles, indoor/outdoor settings). TensorRT optimization may accelerate the inference speed of these object detection models to enable the quick identification and localization of safety gear in each frame.In some embodiments, for cricket, thePose Estimation models may be customized to accurately estimate the pose of players, and focus on key joints relevant to wearing or holding safety gear. By understanding the player's body position, the system can more accurately determine whether gear is being worn correctly or is in the player's possession. TensorRT may optimize these Pose Estimation models for real-time performance, thereby enabling the system to track player movements and gear attachment fluidly.
[0063] In some embodiments, after detecting players and gear with other models, the GNN may be used to model their relationships; nodes represent detected objects, and learned edges indicate spatial or functional connections (e.g., a player wearing a helmet). This allows for reasoning about the scene structure and validating safety compliance based on object relationships. While the GNNs process the output of faster models, TensorRT can optimize parts of the graph processing or feature extraction that feed the GNN. In some embodiments, the vision transformers or similar attention-based models may be fine-tuned on the cricket dataset to understand complex spatial arrangements and contexts within the scene. They could learn nuanced relationships between players and scattered or worn gear. Optimizing these complex architectures with TensorRT is crucial for reducing their computational cost and achieving the low latency required for real-time safety monitoring.
[0064] The balldelivery prevention module 212 preventsthe delivery of a ball if the batsman is not wearing at least one required safety gearand automatically controls the operation of the automated cricket bowling machine 106 by automatically determining, based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman, whether the batsman is wearing all required safety gears.
[0065] The system100 proactively enforces safety compliance by verifying whether a batsman is wearing essential protective gear such as the helmet, thehand gloves, or the leg pads before allowing ball delivery. This automated decision-making by the system 100 eliminates human intervention and ensures that player protection is prerequisite to commencing play. By integrating image capture with depth estimation and object localization through the computer vision model 110, the system 100 provides high accuracy in detecting safety gear presence and placement in real-time, thus minimizing injury risk while maintaining uninterrupted training sessions.
[0066] FIG. 3 illustrates a method of automatically determining whether a batsman is wearing at least one required safety gear and controlling anautomated cricket bowling machine 106 to prevent delivery of a ball if the batsman is not wearing the required safety gearaccording to an embodiment herein. At step 302,one or more real-time images arereceived from one or more cameras 104A-N by an edge device 102. The one or more of cameras104A-N is configured to capture the one or more real-time images of a cricket lane at one or more angles. The edge device102 is operatively connected to theone or more cameras104A-N and the automated cricketbowling machine 106. At step 304,the real-time images areprocessed, by the edge device 102, to extract depth information for accurate object localization.
[0067] At step 306,a computer vision model 110 is applied, by the edge device 102, to the processedreal-time imagesto detect the presence and positioning of one or more safety gears on the batsman.The computer vision model 110 is appliedby (i) identifying objects corresponding to the one or more safety gears comprising a helmet, hand gloves, shoes, and leg pads, and (ii) determininga spatial relationship between the plurality of safety gears that are detected and the batsman by (a) detecting key points on the batsman's body, (b) associating the one or more safety gears that are detected with one or more specific body regions comprising head, hands and legs of the batsman based on predefined positional criteria, and (c) verifying that each safety gear is properly worn by confirming that the helmet is on the head, thehand gloves are on the hands, and the leg pads are on the legs of the batsman.
[0068] At step 308,the edge device 102 automatically determines whether the batsman is wearing all required safety gears, based on the analysis of the spatial relationshipbetween the one or more safety gears that are detected and the batsman.At step 310,the operation of the automated cricket bowling machine 106 is automatically controlled, by the edge device 102, to prevent the delivery of a ball if the batsman is not wearing at least one required safety gear.
[0069] The method enhances player safety through real-time monitoring and intelligent control. By leveraging the cameras 104A-N and the edge computing, the method uses advanced computer vision models to detect the presence and correct positioning of essential safety gearsuch as the helmets, the hand gloves, and the leg padsas well as identifying falls, obstructions, and dangerous objects within the pitch/cricket lane. If safety conditions are unmet, the method proactively halts the automated cricket bowling machine 106, issues audio alerts, and notifies coaching staff, thereby reducing injury risks and liability. Its automation minimizes human error and ensures consistent safety enforcement, improving training efficiency. The methodsupports modular expansion, cloud integration, and remote monitoring.The methodemploys affordable hardware and software while helping prevent costly injuries and legal exposure, and makes it a robust and adaptive solution for indoor cricket safety.
[0070] In some embodiments, the edge device 102 detects a fall event by (i) analyzing a posture of thebatsman using a pose estimation technique, (ii) assessing depth information to determine a position of the batsman relative to a playing surface of the cricket lane, and (iii) identifying a fall when the posture of thebatsman deviates from an upright stance and the depth information indicates a proximity to the playing surface of the cricket lane.
[0071] In some embodiments,upon detection of a fall event of the batsman, the edge device 102 (i) immediately stops the operation of the automated cricket bowling machine to prevent thedelivery of the ball, (ii) activates an audible alarm to alertnearby personnel to indicate the detection of the fall event of the batsman, and (iii) generates a notification indicating the detection of the fall event of the batsmanto a designated monitoring system or a device associated with an individual.
[0072] In some embodiments, the edge device 102 detectsan obstruction or obstacle within a predefined safety zone of the cricket lane by (i) employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone, (ii) employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone, and (iii) determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone for a duration exceeding a predefined threshold.
[0073] In some embodiments, upon detection of an obstruction, the edge device 102(i) stops the operation of the automated cricket bowling machine 106 to prevent thedelivery of the ball, (ii) generates a visual or an audible alert to inform players and coaching staff of the obstruction, and (iii) enables normal operation of the automated cricket bowling machine 106 only after the obstruction is cleared and the predefined safety zone is confirmed to be free of unauthorized objects or individuals. In some embodiments,the method comprisesgenerating an alert indicating the absence of required safety gear.
[0074] In some embodiments, the edge device 102 providesa network interface that is configured to (i) transmit data collected by the edge device102 to a cloud-based platform for storage and analysis, (ii) receive software updates and configuration settings from the cloud-based platform, and (iii) enable remote monitoring and control of the system 100 via the cloud-based platform.
[0075] In some embodiments,the computer vision model 110 utilizes a segmentation model that is optimized for the detection of one or more safety gears in cricket environments. The computer vision model 110 is a deep learning model trained using a custom dataset specific to indoor or outdoor cricket environments.
[0076] FIG. 4 illustrates an exemplary view of the system of FIG. 1 thatis implemented in a cricket lane and automatically determines whether a batsman 414 is wearing at least one required safety gear and controls anautomated cricket bowling machine to prevent delivery of a ball if the batsman 414 is not wearing the required safety gear according to an embodiment herein.The system 100 includes an edge device 102 operatively connected to one or more cameras104A-N and the automated cricketbowling machine 106. The edge device102 includes a processor 108 that receives one or more real-time images from theone or more cameras 104A-N. The one or more cameras104A-N areconfigured to capture the real-time images of a cricket lane at one or more angles. The one or more cameras 104A-N may include a stereo camera thatcaptures real-time stereo images of the cricket lane to provide depth information important for accurate object localization.
[0077] The processor 108 processes the real-time images to extract depth information for accurate object localization. The processor 108 applies a computer vision model 110 to processedthe real-time imagesto detect the presence and positioning of one or more safety gears on the batsman 414. The edge device 102 may be a high-performance computing device, such as an NVIDIA GPU, that processes the real-time images/stereo images and runs the computer vision model 110. Theedge device 102 interfaces with the automated cricket bowling machine 106 and other safety devices to trigger appropriate actions. The computer vision model 110 is trained to identify objects corresponding to the safety gears comprising a helmet 402, hand gloves 404, shoes 408, and leg pads 406. The computer vision model 110 then determinesa spatial relationship between the one or more safety gears that are detected and the batsman 414 by (a) detecting key points on the batsman's body, (b) associating the one or more safety gears that are detected with one or more specific body regions comprising head, hands and legs of the batsman 414 based on predefined positional criteria, and (c) verifying that each safety gear is properly worn by confirming that the helmet 402 is on the head, thehand gloves 404 are on the hands, and the leg pads 406 are on the legs of the batsman 414. The computer vision model 110 utilizes deep learning techniques that are optimized for real-time inference using TensorRT.
[0078] The processor 108 automatically determines, based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman 414, whether the batsman 414 is wearing all required safety gears, and automatically controls the operation of the automated cricket bowling machine 106 to prevent the delivery of a ball if the batsman 414 is not wearing at least one required safety gear.
[0079] FIG. 5 illustrates an exemplary view ofdetectingwhether a batsman 414 is wearing a helmetusing anobject detection techniqueand a helmet detection techniquethat are implemented on the edge device 102 of FIG. 1 according to an embodiment herein.The edge device102 receives one or more real-time images of a cricket lanefrom one or more cameras 104A-N (e.g., a camera feed) or a database (e.g. custom dataset input). The edge device102 processes the real-time images of the cricket lane using a frame normalization technique and a dimension adjustment technique. The edge device102 applies a computer vision model 110 on the processed real-time imagesto detect the presence and positioning of a helmet402 on the batsman 414. The computer vision model 110 may be trained to identify objects corresponding to the safety gears comprisingthe helmet 402, hand gloves 404, shoes 408, and leg pads 406. The computer vision model 110 then determinesa spatial relationship between the helmet that isdetected and the batsman 414, using an object detection technique and a helmet detection technique,by (a) detecting key points on the batsman's body, (b) associating the helmet that is detected with head of the batsman 414 based on predefined positional criteria, and (c) verifying that helmet 402 is properly worn by confirming that the helmet 402 is on the head of the batsman 414.
[0080] In some embodiments, the edge device102 generates an alert indicating the absence of the helmet 402.The edge device 102ensures not just detectshelmet 402 but verifies proper wearing position using positional algorithms to avoid false positives from misplaced helmets. This improves player safety and system reliability by preventing play without proper helmet.
[0081] The edge device 102 may employs apose estimation techniquethat continuously monitors a player's/batsman’s batting stance within an indoor cricket practice environment. When a batsman 414 takes position, the pose estimation techniqueimmediately identifies important postural elements. A key safety application of this pose estimation techniqueis helmet detection verification. Even when a YOLO segmentation model successfully identifies a helmet 402 within the practice area, the pose estimation techniqueverifies proper helmet placement by correlating the helmet's position with the detected head key points of the player/batsman 414. The edge device 102 detects potentially hazardous situations, such as a player 414 batting without a helmet 402 or an improperly secured helmet 402 that doesn't align with the head's key points. It triggers real-time safety alerts through visual indicators. The TensorRT-optimized architecture generates these alerts with minimal latency (e.g., under 16ms), providing immediate feedback to coaches and players. This critical safety feature ensures compliance with protective equipment requirements even in fast-paced training scenarios where traditional object detection alone might not verify proper equipment usage.
[0082] FIG. 6 illustrates a block diagram of the edge device 102 of FIG. 1 that automatically controls anautomated cricket bowling machine 106 to prevent delivery of a ball according to an embodiment herein.The edge device102 automatically determines, based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman 414, whether the batsman 414 is wearing all required safety gears, and automatically controls the operation of the automated cricket bowling machine 106 to prevent the delivery of a ball if the batsman 414 is not wearing at least one required safety gear.
[0083] The edge device 102 performs comprehensive control over the automated cricketbowling machine 106 through an advanced control system that manages multiple critical components. The control system interfaces with four primary mechanical elements: a horizontal rotation mechanism (pan), a vertical angle adjustment mechanism (tilt), and two high-precision spinner motors that control ball rotation. These components harmonize the ball's trajectory, spin characteristics, and delivery position. The edge device 102 continuously monitors theoperational status of the automated cricketbowling machine 106, and maintains real-time awareness of its connection state, power conditions, and overall readiness through an intelligent monitoring system.
[0084] The capability of the edge device 102 is to execute ball delivery sequences enhances the control system. This includes precise control over the timing between ball releases, motor rotation speeds, and specialized delivery patterns that can simulate various bowling scenarios. The control system incorporates comprehensive safety protocols, verifying proper machine connection, power status, and operational readiness before executing commands. The edge device 102 maintains detailed communication logs of all control signals sent to and received from the automated cricketbowling machine 106, enabling precise monitoring and adjustment of the parameters of the automated cricketbowling machine 106. This two-way communication ensures accurate and reliable control over every aspect of the ball delivery system, from initial setup to final release.
[0085] FIG. 7 illustrates an exemplary view of the edge device102of FIG. 1 that identifiesan unauthorized objector individual within the predefined safety zoneby employingan object detection technique to according to an embodiment herein.The edge device102 detectsan obstruction or obstacle within a predefined safety zone 410 of the cricket lane by (i) employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone 410, (ii) employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone 410, and (iii) determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone 410 for a duration exceeding a predefined threshold. Upon detection of an obstruction, the edge device102 is configured to stop the operation of the automated cricket bowling machine 106 to prevent thedelivery of the ball. The edge device102 is configured togenerate a visual or an audible alert to inform players and coaching staff of the obstruction, and enable normal operation of the automated cricket bowling machine 106 only after the obstruction is cleared and the predefined safety zone 410 is confirmed to be free of unauthorized objects or individuals. The edge device102 is configured toensure pitch area safety through the detection of foreign objects or obstructions within a designated danger zone in the cricket lane. Using the depth mapping, the edge device102 enables detection of loose balls, equipment, or unauthorized movement in real time. This feature enhances player safety by preventing accidental tripping, collisions, or misdirected deliveries, and avoids damage to the automated cricket bowling machine 106. Furthermore, the edge device102 enables operators to dynamically define safety zones through configurable parameters, and allows adaptation to different pitch layouts or training scenarios.
[0086] FIG. 7 alsoillustrates key components of a camera-based object detection model that is implemented on the edge device 102. The exemplary viewincludesa cricket practice area with clearly defined danger zones, automated cricketbowling machine 106 with an integrated camera 104 mounted on top, camera field of view covering the entire cricket practice area, and the detection model has three primary components: (i) object detection module that processes the video feed, (ii) classification module that determines if detected objects (e.g. an object, a ball or a person) are unauthorized, (iii) a tracking module that tags the detected object with an ID and tracks the detected object using the ID in the subsequent frames, and (iv) safety protocol module that initiates appropriate responses.
[0087] The process flow of the edge device 102 includes (i) continuous monitoring through the camera 104 mounted on the automated cricketbowling machine 106, (ii) real-time detection and classification of objects/individuals, (iii) automatic safety protocol activation when unauthorized presence is detected. The edge device 102 demonstrates its effectiveness through the illustrated example where a cricket ball and an unauthorized player are detected within the danger zone, triggering a stop signal to the automated cricketbowling machine 106 before any safety incidents occur.
[0088] The edge device 102 employs an object detection framework fine-tuned specifically for indoor cricket practice environments. The detection model is trained on a proprietary dataset encompassing various safety-critical elements of cricket training facilities. This specialized training enables the system 100 to accurately identify and track multiple object classes simultaneously, including cricket balls, players, and other potential obstructions.
[0089] For unauthorized object or individual detection, the edge device100 utilizes a camera-based approach: The high-resolution camera 104 mounted on the automated cricketbowling machine 106 provides a comprehensive view of the practice area. The detection model processes the video feed in real time and creates dynamic safety zone mapping with virtual boundaries defining authorized and unauthorized areas.The system 100 performs real-time classification and threat assessment when any object (such as a cricket ball) or individual is detected within a designated danger zone.
[0090] If an unauthorized presence is confirmed, the edge device100 triggers appropriate safety protocols, including immediate cessation of ball delivery from the automated cricketbowling machine 106.For example, suppose a player 414 moves into a designated danger zone while a delivery sequence is in progress. In that case, the camera 104 captures this movement, the detection system identifies it as an unauthorized presence, classifies it as a safety risk, and automatically prevents the automated cricketbowling machine 106 from releasing additional balls until the zone is clear. Similarly, if cricket balls remain in specific danger areas of the pitch, which could create hazardous playing conditions, and the system 100 will delay subsequent deliveries until the pitch is safe.
[0091] The object detection framework is optimized for performance using hardware acceleration techniques that enable real-time processing with minimal latency, ensuring timely safety responses in dynamic training environments without requiring additional external sensors.
[0092] A representative hardware environment for practicing the embodiments herein is depicted in FIG. 8, with reference to FIGS. 1 through 7. This schematic drawing illustrates a hardware configuration of a server/computer system/computing device in accordance with the embodiments herein.The system includes at least one processing device CPU 10 that may be interconnected via system bus 14 to various devices such as a random access memory (RAM) 12, read-only memory (ROM) 16, and an input/output (I/O) adapter 18. The I/O adapter 18 can connect to peripheral devices, such as disk units 38 and program storage devices 40 that are readable by the system. The system can read the inventive instructions on the program storage devices 40 and follow these instructions to execute the methodology of the embodiments herein. The system further includes a user interface adapter 22 that connects a keyboard 28, mouse 30, speaker 32, microphone 34, and/or other user interface devices such as a touch screen device (not shown) to the bus 14 to gather user input. Additionally, a communication adapter 20 connects the bus 14 to a data processing network 42, and a display adapter 24 connects the bus 14 to a display device 26, which provides a graphical user interface (GUI) 36 of the output data in accordance with the embodiments herein, or which may be embodied as an output device such as a monitor, printer, or transmitter, for example.
[0093] The foregoing description of the specific embodiments will so fully reveal the general nature of the embodiments herein that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without departing from the generic concept, and, therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. Therefore, while the embodiments herein have been described in terms of preferred embodiments, those skilled in the art will recognize that the embodiments herein can be practiced with modification within the scope.
, Claims:I/We claim:
1.A system (100) for automatically determining whether a batsman (414) is wearing at least one required safety gear and controlling anautomated cricketbowling machine (106) to prevent delivery of a ball if the batsman (414) is not wearing the at least one required safety gear, wherein the system (100) comprises:
an edge device (102) operatively connected to aplurality of cameras (104A-N) and the automated cricketbowling machine (106), wherein the edge device (102) comprises:
a processor (108); and
a non-transitory memory storing instructions that, when executed by the processor (108), cause the edge device (102) to:
receive a plurality of real-time images from theplurality of cameras (104A-N), wherein the plurality of cameras (104A-N)is configured to capture the plurality of real-time images of a cricket lane at a plurality of angles;
process the plurality of real-time images to extract depth information for accurate object localization;
apply a computer vision model (110) to the plurality of real-time images that are processed to detect the presence and positioning of a plurality of safety gears on the batsman (414), wherein the computer vision model (110) is trained to
(i) identify objects corresponding to the plurality of safety gears comprising a helmet (402), hand gloves (404), shoes (408), and leg pads (406);
(ii) determine a spatial relationship between the plurality of safety gears that are detected and the batsman (414) by:
detecting key points on the batsman's body;
associating the plurality of safety gears that are detected with a plurality of specific body regions comprising head, hands and legs of the batsman (414)based on predefined positional criteria; and
verifying that each safety gear is properly worn by confirming that the helmet (402) is on the head, thehand gloves (404) are on the hands, and the leg pads (406) are on the legs of the batsman (414);
automatically determine, based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman (414), whether the batsman (414) is wearing all required safety gears; and
automatically control the operation of the automated cricket bowling machine (106) to prevent the delivery of a ball if the batsman (414) is not wearing at least one required safety gear.

2.The system (100)as claimed inclaim 1, wherein the edge device (102) is configured to detect a fall event by
analyzing a posture of thebatsman (414) using a pose estimation technique;
assessing depth information to determine a position of the batsman (414) relative to a playing surface of the cricket lane; and
identifying a fall when the posture of thebatsman (414) deviates from an upright stance and the depth information indicates a proximity to the playing surface of the cricket lane.

3.The system (100)as claimed inclaim 2, wherein upon detection of a fall event of the batsman (414), the edge device (102) is configured to
immediately stop the operation of the automated cricket bowling machine (106) to prevent thedelivery of the ball;
activate an audible alarm to alertnearby personnel to indicate the detection of the fall event of the batsman (414); and
generate a notification indicating the detection of the fall event of the batsman (414)to a designated monitoring system or a device associated with an individual.

4.The system (100)as claimed inclaim 1, wherein the edge device (102) is configured to detect an obstruction or obstacle within a predefined safety zone (410)of the cricket lane by
employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone (410);
employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone (410); and
determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone (410) for a duration exceeding a predefined threshold.

5.The system (100)as claimed inclaim 4, wherein upon detection of an obstruction, the edge device (102) is configured to
stop the operation of the automated cricket bowling machine (106) to prevent thedelivery of the ball;
generate a visual or an audible alert to inform players and coaching staff of the obstruction; and
enable normal operation of the automated cricket bowling machine (106) only after the obstruction is cleared and the predefined safety zone (410) is confirmed to be free of unauthorized objects or individuals.

6.The system (100)as claimed inclaim 1, wherein the edge device (102)generates an alert indicating the absence of required safety gear.

7.The system (100)as claimed inclaim 1, wherein the edge device (102) comprises a network interface that is configured to
transmit data collected by the edge device (102) to a cloud-based platform for storage and analysis;
receive software updates and configuration settings from the cloud-based platform; and
enable remote monitoring and control of the system (100) via the cloud-based platform.

8.The system (100)as claimed inclaim 1, wherein the computer vision model (110) utilizes a segmentation model that is optimized for detection of the plurality of safety gears in cricket environments, wherein the computer vision model (110) is a deep learning model trained using a custom dataset specific to indoor or outdoor cricket environments.

9.The system (100)as claimed inclaim 1, wherein the edge device (102) is configured to log instances of missing safety gear for subsequent review and analysis, wherein the edge device (102) comprises a user interface for monitoring a status of the system (100) and configuring operational parameters of the system (100).

10.A computer-implemented method for automatically determining whether a batsman (414) is wearing at least one required safety gear and controlling anautomated cricket bowling machine (106) to prevent delivery of a ball if the batsman (414) is not wearing the at least one required safety gear, wherein the method comprises:
receiving, by an edge device (102), a plurality of real-time images from aplurality of cameras (104A-N), wherein the plurality of cameras (104A-N)is configured to capture the plurality of real-time images of a cricket lane at a plurality of angles, wherein theedge device (102)is operatively connected to theplurality of cameras (104A-N) and the automated cricketbowling machine (106);
processing, by the edge device (102), the plurality of real-time images to extract depth information for accurate object localization;
applying, by the edge device (102), a computer vision model (110) to the plurality of real-time images that are processed to detect the presence and positioning of a plurality of safety gears on the batsman (414), wherein applying the computer vision model (110) comprises:
identifying objects corresponding to the plurality of safety gears comprising a helmet (402), hand gloves (404), shoes (408), and leg pads (406);and
determininga spatial relationship between the plurality of safety gears that are detected and the batsman (414) by
detecting key points on the batsman's body;
associating the plurality of safety gears that are detected with a plurality of specific body regions comprising head, hands and legs of the batsman (414)based on predefined positional criteria; and
verifying that each safety gear is properly worn by confirming that the helmet (402) is on the head, thehand gloves (404) are on the hands, and the leg pads (406) are on the legs of the batsman (414);
automatically determining, by the edge device (102), based on the analysis of the spatial relationshipbetween the plurality of safety gears that are detected and the batsman (414), whether the batsman (414) is wearing all required safety gears; and
automatically controlling, by the edge device (102), the operation of the automated cricket bowling machine (106) to prevent the delivery of a ball if the batsman (414) is not wearing at least one required safety gear.

11.The computer-implemented method as claimed in claim 10, wherein the method comprisesdetecting a fall event by
analyzing a posture of thebatsman (414) using a pose estimation technique;
assessing depth information to determine a position of the batsman (414) relative to a playing surface of the cricket lane; and
identifying a fall when the posture of thebatsman (414) deviates from an upright stance and the depth information indicates a proximity to the playing surface of the cricket lane.

12.The computer-implemented method as claimed inclaim 11, wherein, upon detection of a fall event of the batsman (414), the method comprises
immediately stopping the operation of the automated cricket bowling machine (106) to prevent thedelivery of the ball;
activating an audible alarm to alertnearby personnel to indicate the detection of the fall event of the batsman (414); and
generating a notification indicating the detection of the fall event of the batsman (414)to a designated monitoring system or a device associated with an individual.

13.The computer-implemented method as claimed inclaim 10, wherein the method comprisesdetectingan obstruction or obstacle within a predefined safety zone (410)of the cricket lane by
employingan object detection technique to identify an unauthorized objector individual within the predefined safety zone (410);
employing depth information to determine the exact location and size of the obstructionwithin the predefined safety zone (410); and
determining that an obstruction is present when an unauthorized object or individual is detected within the predefined safety zone (410) for a duration exceeding a predefined threshold.

14.The computer-implemented method as claimed inclaim 13, wherein, upon detection of an obstruction, the method comprises
stopping the operation of the automated cricket bowling machine (106) to prevent thedelivery of the ball;
generating a visual or an audible alert to inform players and coaching staff of the obstruction; and
enabling normal operation of the automated cricket bowling machine (106) only after the obstruction is cleared and the predefined safety zone (410) is confirmed to be free of unauthorized objects or individuals.

15.The computer-implemented method as claimed inclaim 10, wherein the method comprisesgenerating an alert indicating the absence of required safety gear.

16.The computer-implemented method as claimed inclaim 10, wherein the method comprisesproviding a network interface on the edge device (102) that is configured to
transmit data collected by the edge device (102) to a cloud-based platform for storage and analysis;
receive software updates and configuration settings from the cloud-based platform; and
enable remote monitoring and control of the system (100) via the cloud-based platform.

17.The computer-implemented method as claimed inclaim 10, wherein the computer vision model (110) utilizes a segmentation model that is optimized for detection of the plurality of safety gears in cricket environments, wherein the computer vision model (110) is a deep learning model trained using a custom dataset specific to indoor or outdoor cricket environments.

Dated this May 21st, 2025

Arjun KarthikBala
(IN/PA 1021)
Agent for Applicant