Description:4. DESCRIPTION
TECHNICAL FIELD
[001] The disclosed subject matter relates generally to mithun, cattle, buffalo identification detection. More particularly, the present disclosure relates to an artificial intelligence-based bovid identification system using muzzle recognition and natural biometrics.

BACKGROUND
[002] The livestock industry has an increasing interest in identifying individual animals for precision management of individuals and tracking animal movement. Individual cattle identification is a crucial component of cattle traceability, which provides insights into tracking disease trajectories, ascertaining ownership, and managing cattle production and distribution. Contact and non-contact methods are commonly used for cattle identification, with branding being a common contact method to establish ownership. Common contact methods include ear notching, earmarking, tagging, and branding. While these methods identify individual cattle, they require human effort to recognize and locate the animals, which can be laborious and time-consuming.

[003] Radiofrequency identification (RFID) systems offer a non-contact method for cattle identification. However, these systems have drawbacks, including stress on cattle and the potential for damage once attached to the animals. Contact methods may also lead to short- and long-term complications in the integrity of cattle ears or other anatomical body parts. In addition, the sensors (e.g., tags, transponders) with animal identification can fade, be damaged, and be lost due to cattle interference, movement, and environmental exposure.

[004] Alternatively, contactless identification methods using unique animal biometric features offer a solution to eliminate human disturbance to the animals. Similar to human biometrics, such as faces and fingerprints, animals also have biometrics for individual identifiers. Identifying biometric livestock markers includes DNA pairing, autoimmune antibody matching, iris scanning, retinal imaging, coat pattern recognition, muzzle identification, and facial recognition. Among these methods, muzzle identification is a relatively low-cost and simple method that has recently received increasing research interest. Muzzle pattern is a dermatoglyphic cattle trait equivalent to human fingerprints. Cattle Muzzle Identification worked on biometrics of cattle using images and collected muzzle patterns. By identifying cattle muzzle patterns, non-invasive and unique methods for cattle identification and tracking are possible, including validation with advancements in machine learning modeling.

[005] For muzzle-based identification, the conventional digital image processing algorithms (e.g., scale-invariant feature transform and box-counting fractal dimension models) were used to identify individual cattle automatically. These methods typically matched features, including color, texture, shape, and edge, among different muzzle images and achieved high identification accuracy with small image sets and controlled conditions. However, the method performance may be challenged by inconsistent illumination and background, variable muzzle shapes and sizes, similar appearances of the same animal at different times, missing parts or occlusions on a muzzle, and low resolution.

[006] The problems associated with current technologies used for identifying and tracking individual mithun, cattle and buffalo include the strenuous process of revealing the biometric pattern of the muzzle, the accurate identification and storage of muzzle patterns in a registration and identification dataset, accurate object detection and segmentation of mithun, cattle and buffalo muzzles, and the need for faster image processing and classification. Hence, there is a need to develop an artificial intelligence-based bovid (mithun, cattle, and buffalo) identification system using muzzle recognition and natural biometrics.

[007] In the light of the aforementioned discussion, there exists a need for a certain system with novel methodologies that would overcome the above-mentioned disadvantages.

SUMMARY
[008] The following presents a simplified summary of the disclosure in order to provide a basic understanding of the reader. This summary is not an extensive overview of the disclosure and it does not identify key/critical elements of the invention or delineate the scope of the invention. Its sole purpose is to present some concepts disclosed herein in a simplified form as a prelude to the more detailed description that is presented later.

[009] An objective of the present disclosure is directed towards an AI-based system that uses image processing and machine learning techniques to identify and track individual mithun, cattle, and buffalo based on their unique muzzle patterns.

[0010] Another objective of the present disclosure is directed towards a system that uses high-resolution images of the muzzle to capture the distinctive biological characteristics that are unique to each bovid.

[0011] Another objective of the present disclosure is directed towards a system that processes the images through an AI algorithm to generate a unique identification code that can be stored in a server or a database and retrieved for easy identification and tracking in future interactions.

[0012] Another objective of the present disclosure is directed towards a system that provides several advantages over traditional methods of identification, including non-invasiveness, accuracy, and efficiency.

[0013] Another objective of the present disclosure is directed towards a system that eliminates the need for physical tagging or branding, reducing the risk of animal injury or stress.

[0014] Another objective of the present disclosure is directed towards a system that enables use in real-time and provides reliable identification and tracking, even in large herds.

[0015] Another objective of the present disclosure is directed towards a system that offers a significant improvement in the identification and tracking of individual mithun, cattle, and buffalo and helps improve the overall management and care of these animals.

[0016] Another objective of the present disclosure is directed towards a system that supports edge computing through a computing device, which reduces concerns about latency and performance for accurate identification of the muzzle region and head during image capture.

[0017] Another objective of the present disclosure is to provide a system that ensures fast and efficient identification through the use of edge computing, even in areas with poor internet connectivity or limited computing resources.

[0018] Another objective of the present disclosure is directed towards a system that provides a reliable and efficient identification method for mithun, cattle, and buffalo based on their unique muzzle pattern.

[0019] Another objective of the present disclosure is directed towards a system that improves a novel identification system that utilizes various machine learning and image processing techniques to enable the real-time identification of mithun, cattle, and buffalo.

[0020] Another objective of the present disclosure is directed towards a system that utilizes deep learning methods to accurately locate the muzzle portion of the bovid and extract its pattern for post-classification.

[0021] Another objective of the present disclosure is directed towards a system that enables distinct and reliable identification of mithun, cattle, and buffalo based on their muzzle pattern, which helps in effective livestock management.

[0022] In an embodiment of the present disclosure, an artificial intelligence-based bovid identification system, includes an image capturing module is configured to enable a user to perform at least one of: upload one or more images of one or more bovids on a computing device, open a camera of the computing device and capture one or more images of the one or more bovids; initiate the search process by opening the camera, the image capturing module is configured to send the one or more images of the one or more bovids to a cloud server over a network.

[0023] In another embodiment of the present disclosure, the cloud server includes an image processing module is configured to receive the one or more images from the computing device, the image processing module includes an image collection module is configured to collect the one or more images and delivers them to a muzzle identification module.

[0024] In another embodiment of the present disclosure, the muzzle identification module is configured to draw one or more patterns using one or more machine learning techniques and identify a muzzle part of the bovid using one or more deep learning models, the muzzle identification module is configured to send one or more muzzle-identified images to a pattern extraction module.

[0025] In another embodiment of the present disclosure, the pattern extraction module is configured to extract one or more muzzle patterns from the one or more muzzle-identified images using an image-processing technique or a pre-processing technique, the pattern extraction module is configured to send one or more extracted muzzle patterns to a feature identification module.

[0026] In another embodiment of the present disclosure, the feature identification module is configured to identify one or more beads and ridges in the one or more extracted muzzle patterns, the feature identification module is configured to send one or more identified beads and ridges to a feature matching module, the feature matching module is configured to define one or more key points and match the one or more key points in the one or more identified beads and ridges with one or more master images.

[0027] In another embodiment of the present disclosure, the feature matching module is configured to send one or more matched key points with the one or more master images to a bovid identification module, the bovid identification module is configured to identify an individual bovid upon a successful match of the one or more key points using one or more deep learning models thereby representing bovid information in livestock management.

[0028] In another embodiment of the present disclosure, the image capturing module is configured to enable the user to capture one or more videos of the one or more bovids on the computing device.

[0029] In another embodiment of the present disclosure, the image processing module is configured to identify the specific component of an object in the one or more images, specifically the muzzle.

[0030] In another embodiment of the present disclosure, the image processing module is configured to identify various attributes of the muzzle, including its size, orientation, and scale, without causing any delays or latency in the identification process.
[0031] In another embodiment of the present disclosure, the image processing module includes one or more machine learning and image processing techniques to identify the individual bovid based on real-time images and video capture.

[0032] In another embodiment of the present disclosure, the one or more deep learning models comprise both cosine and scale-invariant feature transform algorithms.

[0033] In another embodiment of the present disclosure, the image processing technique comprises a Gabor filter and strongly supports machine learning model training.

[0034] In another embodiment of the present disclosure, the pattern identification module is configured to perform the classification of one or more extracted muzzle patterns and is skeletonized, and thinned so as to extract a morphological operation that is used to remove selected foreground pixels from one or more binary images.

[0035] In another embodiment of the present disclosure, the one or more extracted muzzle patterns are further classified using a support vector machine (SVM).

[0036] In another embodiment of the present disclosure, the image processing module comprises one or more post-processing techniques are configured to capture at least one of: orientation; size; and scale; and to match using the one or more deep learning models.

[0037] To improve the muzzle identification we might use image deblurring and Image super resolution using deep learning. Using SRCNN and PyTorch deep learning library will increase the clarity and resolution of muzzle. We might also use fft_deblur method of Fourier Transformation to increase muzzle visual.

[0038] To predict the missing part of muzzle we might use BG-NMF model , it helps the system to identify the animal if any part of the muzzle capture is missing or any food/ dust particle on muzzle while capturing resulting low accuracy. To overcome this problem BG-NMF model helps in predicting the missing data.

BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In the following, numerous specific details are set forth to provide a thorough description of various embodiments. Certain embodiments may be practiced without these specific details or with some variations in detail. In some instances, certain features are described in less detail so as not to obscure other aspects. The level of detail associated with each of the elements or features should not be construed to qualify the novelty or importance of one feature over the others.

[0040] FIG. 1 is an example diagram depicting an artificial intelligence-based bovid identification system using muzzle recognition and natural biometrics, in accordance with one or more exemplary embodiments.

[0041] FIG. 2 is an example diagram depicting a schematic representation of the image processing module, in accordance with one or more exemplary embodiments.

[0042] FIG. 3 is an example diagram depicting a schematic representation of a system for identifying an individual bovid through muzzle patterns, in accordance with one or more exemplary embodiments.

[0043] FIG. 4 is an example of a flow diagram depicting a method for identifying an individual bovid using muzzle recognition and natural biometrics, in accordance with one or more exemplary embodiments.

[0044] FIG. 5 is a block diagram illustrating the details of a digital processing system in which various aspects of the present disclosure are operative by execution of appropriate software instructions.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS
[0045] It is to be understood that the present disclosure is not limited in its application to the details of construction and the arrangement of components set forth in the following description or illustrated in the drawings. The present disclosure is capable of other embodiments and of being practiced or of being carried out in various ways. Also, it is to be understood that the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting.

[0046] The use of “including”, “comprising” or “having” and variations thereof herein is meant to encompass the items listed thereafter and equivalents thereof as well as additional items. The terms “a” and “an” herein do not denote a limitation of quantity, but rather denote the presence of at least one of the referenced item. Further, the use of terms “first”, “second”, and “third”, and the like, herein do not denote any order, quantity, or importance, but rather are used to distinguish one element from another.

[0047] Referring to FIG. 1, FIG. 1 is an example diagram 100 depicting an artificial intelligence-based bovid identification system using muzzle recognition and natural biometrics, in accordance with one or more exemplary embodiments. The diagram 100 includes a computing device 102, an image capturing module 104, a real-time bovid identification module 106, a network 108, a cloud server 110, and an image processing module 112. The computing device 102 may be operated by a user. The user may include, but not limited to, a farmer, an owner, and the like. The computing device 102 includes the image capturing module 104, and the real-time bovid identification module 106, and the cloud server 108 includes the image processing module 112. The image capturing module 104 may be configured to enable the user to upload the images of the bovids stored in the memory of the computing device 102; to open a camera of the computing device 102, and enable to capture the images of the one or more bovids; to initiate the search process by opening the camera on the computing device 102.

[0048] The real-time bovid identification module 106 may be configured to receive the images and transfers the images to the cloud server over the network 108. The real-time bovid identification module 106 may be configured to enable the user to identify the type of bovid, and bovid information in real-time using muzzle recognition and natural biometrics. The bovid may include, but not limited to, the family of mammals that includes a diverse group of cloven-hoofed, ruminant animals such as mithun, cattle, buffalo, sheep, goats, bison, antelopes, and the like. The bovid information may include, but not limited to, a registration number, a registration form, blood, body fluid, or other biological components, videos or images of nose prints, DNA analysis report, muzzle patterns, and the like.

[0049] The network 108 may include, but is not limited to, an Ethernet, a wireless local area network (WLAN), or a wide area network (WAN), a Bluetooth low energy network, a ZigBee network, a Controller Area Network (CAN bus), a WIFI communication network, e.g., the wireless high-speed Internet, or a combination of networks, a cellular service such as a 4G (e.g., LTE, mobile WiMAX) or 5G cellular data service, an RFID module, an NFC module, wired cables, such as the world-wide-web based Internet, or other types of networks may include Transport Control Protocol/Internet Protocol (TCP/IP) or device addresses (e.g., network-based MAC addresses, or those provided in a proprietary networking protocol, such as Modbus TCP, or by using appropriate data feeds to obtain data from various web services, including retrieving XML data from an HTTP address, then traversing the XML for a particular node) and the like without limiting the scope of the present disclosure.

[0050] Although the computing device 102 shown in FIG. 1, is an embodiment of the system 100 may support any number of computing devices. The system 100 may support only one computing device (102). The computing device 102 may include, but are not limited to, a desktop computer, a personal mobile computing device such as a tablet computer, a laptop computer, or a netbook computer, a smartphone, a server, an augmented reality device, a virtual reality device, a digital media player, a piece of home entertainment equipment, backend servers hosting the database and other software, and the like. Each computing device 102 supported by the system 100 is realized as a computer-implemented or computer-based device having the hardware or firmware, software, and/or processing logic needed to carry out the intelligent messaging techniques and computer-implemented methodologies described in more detail herein.

[0051] The real-time bovid identification module 106 may be downloaded from the cloud server 110. For example, the real-time bovid identification module 106 may be any suitable application downloaded from, GOOGLE PLAY® (for Google Android devices), Apple Inc.'s APP STORE® (for Apple devices, or any other suitable database). In some embodiments, the real-time bovid identification module 106 may be software, firmware, or hardware that is integrated into the computing device 102. The real-time bovid identification module 106 which is accessed as mobile applications, web applications, software that offers the functionality of accessing mobile applications, and viewing/processing of interactive pages, for example, are implemented in the computing device 102 as will be apparent to one skilled in the relevant arts by reading the disclosure provided herein.

[0052] The real-time bovid identification module 106 may be configured to identify the mithun, cattle, and buffalo based on the unique pattern of their muzzle using machine learning and image processing techniques. The real-time bovid identification module 106 may be configured to enable the user to capture the image on the computing device 102 for real-time detection of the bovid, thereby eliminating cloud latency and offering better efficiency. The real-time bovid identification module 106 may be configured to obtain the bovid information from the multiple integrated modules (as shown in FIG.2) on the cloud server 110. The cloud server 110 may be configured to perform object detection and post-processing, respectively. The post-processing techniques on the cloud server 110 involve pre-processing the image using CLAHE for contrast level correction and Gabor filtration for feature extraction like muzzle texture identification. The muzzle pattern is then skeletonized and thinned for final classification using a support vector machine or other classifiers. The image processing module 112 on the cloud server 112 may be effective in capturing muzzle patterns of mithun, cattle, and buffalo at any orientation, size, and scale. The image processing module 112 may be configured to provide an accurate and efficient way of identifying mithun, cattle, and buffalo, making it suitable for livestock management and other relevant applications.

[0053] The image processing module 112 may be programmed with the Machine learning classification models (e.g., support vector machine, K-nearest neighbor, and decision tree) and may be embedded with image processing-based feature extractors (e.g., Weber local descriptor) to boost the performance of muzzle identification further. Furthermore, the image processing module 112 may be programmed with the Deep learning models to capture spatial and temporal dependencies of images/videos through the use of shared-weight filters and may be trained end-to-end without the strenuous hand-crafted design of feature extractors, empowering the models to adaptively discover the underlying class-specific patterns and the most discriminative features automatically.

[0054] The cloud server 110 may be configured to partition the muzzle dataset into training, validation, and testing subsets for training and testing the deep learning models. The system 100 may be configured to calculate the accuracy of each model during each epoch of training using the validation dataset and after training using the testing dataset to determine model performance for the overall classification. The system 100 uses data augmentation to create synthesized images and increase limited datasets for training deep learning models. The system 100 may be configured to adopt four augmentation strategies based on raw image limitations, namely, horizontal flipping, brightness modification, randomized rotation, and blurring.

[0055] The image processing module 112 includes a total of 50+ models used for deep learning image classification models comparatively evaluated to determine the optimal models for identifying individual bovids with the cropped and resized muzzle images. These models are derived from image classification models like AlexNet, DenseNet, DPN, EfficientNet, Inception, MnasNet, MobileNet, RegNet, ResNet, ResNeXt, ShuffleNet, SqueezeNet, VGG, Wide ResNet and GoogLeNet.

[0056] Transfer learning may be deployed during training, with which models may be pre-trained with a large dataset, ImageNet. In contrast, only the fully connected layers of the models may be fine-tuned with the current dataset for custom classification. This strategy improves training efficiency without compromising inference performance. The mithun, cattle, and buffalo muzzle dataset was randomly partitioned and reshuffled into three subsets: 65% for training, 15% for validation, and 20% for testing. Image pixel intensities per color channel were normalized to the range of [0,1] for enhanced image recognition performance.

[0057] In another embodiment of the present invention, the image processing module 112 may include the machine learning classification models (e.g., support the vector machine, K-nearest neighbor, and decision tree) embedded with image processing-based feature extractors (e.g., Weber local descriptor) to boost the performance of muzzle identification further. The image processing module 112 may be programmed with the deep learning technique and is a data-driven method for computer vision applications in animal production. The image processing module 112 with the deep learning technique may be configured to capture spatial and temporal dependencies of images/videos through the use of shared-weight filters. The image processing module 112 may be trained end-to-end without the strenuous hand-crafted design of feature extractors, empowering the models to adaptively discover the underlying class-specific patterns and the most discriminative features automatically.

[0058] The raw images contained unnecessary anatomical parts (e.g., face, eye, and body), particularly for classification purposes. To reduce classification interference and highlight muzzle visual features, the bovid face area is rotated to align horizontally, after which the muzzle area may be manually cropped. Extremely blurry, incomplete, or feed-covered muzzle images were removed to maintain dataset quality for model training, validation, and testing. Approximately 300 bovid images are required, for each bovid qualified images of 10-30 are required. On total, 4000 to 6000 images would be required, Expected Accuracy 80% to 98%. Image Dimension: Cropped muzzle images in the dataset should be 300x300 pixels with high resolution.

[0059] In another embodiment of the present invention, the system and method disclose a non-invasive and unique mithun, cattle, and buffalo identification through a muzzle pattern. The method includes identifying the muzzle pattern as a mithun, cattle, and buffalo dermatoglyphic trait equivalent to human fingerprints and capturing high-definition mithun, cattle, and buffalo images to identify the muzzle part and define beads and ridges. The method involves pre-processing raw images by rotating the mithun, cattle, and buffalo face area so that the muzzle area aligns horizontally and then manually cropping the muzzle area to reduce interference and highlight muzzle visual features. The muzzle pattern is then extracted from the raw image using pre-processing techniques and identifying ridges and beads in the extracted muzzle pattern. Key points are defined and matched using feature descriptors such as SIFT or ORB (Oriented Fast & Rotated Brief) with the master image to identify the mithun, cattle, and buffalo. Digital image processing algorithms are employed to identify individual mithun, cattle, and buffalo based on muzzle images automatically.

[0060] The method further includes embedding machine learning classification models with image processing-based feature extractors to improve mithun, cattle, buffalo identification performance. Deep learning models are trained for mithun, cattle, and buffalo identification through muzzle using transfer learning and image augmentation techniques. Transfer learning is deployed during training by pre-training models on a large dataset, ImageNet, and fine-tuning them with the current mithun, cattle, buffalo muzzle dataset for custom classification. Evaluating over 50 deep learning image classification models for identifying individual mithun, beef cattle, and buffalo with cropped and resized muzzle images. The mithun, cattle, and buffalo muzzle dataset is partitioned into training, validation, and testing subsets for training and testing the deep learning models. Accuracy is calculated for each model during each epoch of training using the validation dataset and after training using the testing dataset to determine model performance for the overall classification. Data augmentation creates synthesized images and increases limited datasets for training deep learning models. Adopting four augmentation strategies based on raw image limitations, namely, horizontal flipping, brightness modification, randomized rotation, and blurring. Identifying the mithun, cattle, and buffalo on the successful matching of key points using the image processing technique.

[0061] The system allows for real-time capture and live identification of bovids in three ways: through uploading image on the computing device 102, 2) through the opening camera and capturing image, and 3) by opening the camera, the system requests to initiate the search process to identify the bovid. The system provides accurate, reliable, and efficient mithun, cattle, and buffalo identification through muzzle patterns. The system 100 may be configured to enable the user to capture and identify the head and muzzle portion in real-time. The system 100 may be configured to support edge computing through the mobile end in identification so that the latency and performance is not a concern for accurate identification of muzzle region and head during capture.

[0062] Embodiments of the present invention provide a system and method for identifying mithun, cattle, and buffalo based on their muzzle patterns. The system utilizes various techniques to detect, pre-process, and classify the muzzle pattern of the bovid. In particular, the system employs real-time detection using edge computing, pre-processing with appropriate machine learning models, and post-processing techniques such as CLAHE and Gabor filtration for feature enhancement support. With the enhanced filtration and feature matching show the majority of key points in the matching algorithm, such as the Oriented FAST and Rotated BRIEF (ORB) algorithm and speed up robust feature (SURF) algorithm. The extracted muzzle pattern is further classified using a support vector machine (SVM) or other classifiers. The system 100 may accurately identify mithun, cattle, and buffalo based on their muzzle pattern even when the animals are captured at different orientations, sizes, and scales. The identification process is efficient and reliable, making it suitable for use in livestock management and other related fields.

[0063] Referring to FIG. 2 is an example diagram 200 depicting a schematic representation of the image processing module, in accordance with one or more exemplary embodiments. The diagram 200 depicts the image processing module 112. The image processing module 112 includes a bus 201, an image collection module 202, a muzzle identification module 204, a pattern extraction module 206, a feature identification module 208, a feature matching module 210, and a bovid identification module 212. The bus 201 may include a path that permits communication among the modules of the image processing module 112. The term "module" is used broadly herein and generally refers to a program resident in memory of the cloud server 108.

[0064] The image collection module 202 may be configured to receive the raw high-definition images of the bovid captured through the camera on the computing device 102. The muzzle identification module 204 may be configured to draw the pattern and identify the muzzle part of the bovid in the raw high-definition images. The pattern extraction module 206 may be configured to extract the muzzle pattern from the raw high-definition images using the pre-processing technique. The pattern extraction module 206 may also be represented as a pre-processing module. The pre-processing technique is strongly supported that encourages the machine learning model appropriate training in identifying the bovid. The machine learning techniques use both cosine and scale-invariant feature transform algorithms. In addition, the pre-processing technique may use a Gabor filter. The feature identification module 208 may be configured to identify the beads and ridges in the extracted muzzle pattern. The feature matching module 210 may be configured to define key points and match the key points with the master images to identify the bovid. The bovid identification module 212 may be configured to identify the bovid information upon a successful match of the key points using the deep learning model.

[0065] Referring to FIG. 3 is an example diagram 300 depicting a schematic representation of a system for identifying an individual bovid through muzzle patterns, in accordance with one or more exemplary embodiments. The diagram 300 includes the image collection module 202, the muzzle identification module 204, the pattern extraction module 206, the feature identification module 208, the feature matching module 210, and the bovid identification module 212.

[0066] The image capturing module 104 on the computing device 102 may be configured to enable the user to perform at least one of: upload the images on the computing device 102; open a camera of the computing device 102 and capture the images; initiate the search process by opening the camera. The image capturing module 104 is configured to send the images to the cloud server 110 over the network 108. The cloud server 108 includes the image processing module 112, and is configured to receive the images from the computing device 102 over the network 108. The image processing module on the cloud server 110 includes the image collection module 202 may be configured to collect the images and delivers the collected images to the muzzle identification module 204. The muzzle identification module 204 may be configured to draw patterns using the machine learning techniques and identify the muzzle part of the bovid using the deep learning models. The muzzle identification module 204 may be configured to send the muzzle-identified images to the pattern extraction module 206.

[0067] The pattern extraction module 206 may be configured to extract the muzzle patterns from the muzzle-identified images using the image-processing technique and/or the pre-processing technique. The pattern extraction module 206 may be configured to send the extracted muzzle patterns to the feature identification module 208. The feature identification module 208 may be configured to identify the beads and ridges in the extracted muzzle patterns. The feature identification module 208 may be configured to send the identified beads and ridges to the feature matching module 210. The feature matching module 210 may be configured to define the key points and match the key points in the identified beads and ridges with the master images. The feature matching module 210 may be configured to send the matched key points with the master images to the bovid identification module 212. The bovid identification module 212 may be configured to identify an individual bovid and bovid information upon the successful match of the key points using the deep learning models, thereby representing the bovid information in livestock management. The bovid information may include, but not limited to, a registration number, a registration form, blood, body fluid, or other biological components, videos or images of nose prints, DNA analysis report, muzzle patterns, and the like.

[0068] Referring to FIG. 4 is an example of flow diagram 400 depicting a method for identifying an individual bovid using muzzle recognition and natural biometrics, in accordance with one or more exemplary embodiments. The method 400 may be carried out in the context of the details of FIG. 1, FIG. 2, and FIG. 3. However, the method 400 may also be carried out in any desired environment. Further, the aforementioned definitions may equally apply to the description below.

[0069] The method commences at step 402, receiving the raw images of the bovid from the computing device by the image collection module. At step 404, drawing the pattern using the machine learning techniques and identifying the muzzle part of the bovid using the deep learning models by the muzzle identification module. At step 406, extracting the muzzle patterns from the muzzle-identified images by the pattern extraction module using the image-processing technique or the pre-processing technique, which comprises a Gabor filter and strongly supports machine learning model training. At step 408, identifying the beads and ridges in the extracted muzzle patterns by the feature identification module. At step 410, defining the key points and matching the key points in the identified beads and ridges with the master images by the feature matching module. At step 412, identifying an individual bovid upon the successful match of the key points by the bovid identification module using the deep learning model that uses both cosine and scale-invariant feature transform algorithms. At step 414, sending bovid information to the computing device from the cloud server by the image processing module.

[0070] Referring to FIG. 5, FIG. 5 is a block diagram illustrating the details of digital processing system 500 in which various aspects of the present disclosure are operative by execution of appropriate software instructions. Digital processing system 500 may correspond to the computing device 102 (or any other system in which the various features disclosed above can be implemented).

[0071] Digital processing system 500 may contain one or more processors such as a central processing unit (CPU) 510, random access memory (RAM) 520, secondary memory 530, graphics controller 560, display unit 570, network interface 580, an input interface 590. All the components except display unit 570 may communicate with each other over communication path 550, which may contain several buses as is well known in the relevant arts. The components of Figure 5 are described below in further detail.

[0072] CPU 510 may execute instructions stored in RAM 520 to provide several features of the present disclosure. CPU 510 may contain multiple processing units, with each processing unit potentially being designed for a specific task. Alternatively, CPU 510 may contain only a single general-purpose processing unit.

[0073] RAM 520 may receive instructions from secondary memory 530 using communication path 550. RAM 520 is shown currently containing software instructions, such as those used in threads and stacks, constituting shared environment 525 and/or user programs 526. Shared environment 525 includes operating systems, device drivers, virtual machines, etc., which provide a (common) run time environment for execution of user programs 526.

[0074] Graphics controller 560 generates display signals (e.g., in RGB format) to display unit 570 based on data/instructions received from CPU 510. Display unit 570 contains a display screen to display the images defined by the display signals. Input interface 590 may correspond to a keyboard and a pointing device (e.g., touch-pad, mouse) and may be used to provide inputs. Network interface 580 provides connectivity to a network (e.g., using Internet Protocol), and may be used to communicate with other systems (such as those shown in Figure 1, a network 108) connected to the network 108.

[0075] Secondary memory 530 may contain hard drive 535, flash memory 536, and removable storage drive 537. Secondary memory 530 may store the data software instructions (e.g., for performing the actions noted above with respect to the Figures), which enables digital processing system 500 to provide several features in accordance with the present disclosure.

[0076] Some or all of the data and instructions may be provided on the removable storage unit 540, and the data and instructions may be read and provided by removable storage drive 537 to CPU 510. Floppy drive, magnetic tape drive, CD-ROM drive, DVD Drive, Flash memory, a removable memory chip (PCMCIA Card, EEPROM) are examples of such removable storage drive 537.

[0077] The removable storage unit 540 may be implemented using medium and storage format compatible with removable storage drive 537 such that removable storage drive 537 can read the data and instructions. Thus, removable storage unit 540 includes a computer-readable (storage) medium having stored therein computer software and/or data. However, the computer (or machine, in general) readable medium can be in other forms (e.g., non-removable, random access, etc.).

[0078] In this document, the term "computer program product" is used to generally refer to the removable storage unit 540 or hard disk installed in hard drive 535. These computer program products are means for providing software to digital processing system 500. CPU 510 may retrieve the software instructions, and execute the instructions to provide various features of the present disclosure described above.

[0079] The term "storage media/medium" as used herein refers to any non-transitory media that store data and/or instructions that cause a machine to operate in a specific fashion. Such storage media may comprise non-volatile media and/or volatile media. Non-volatile media includes, for example, optical disks, magnetic disks, or solid-state drives, such as storage memory 530. Volatile media includes dynamic memory, such as RAM 520. Common forms of storage media include, for example, a floppy disk, a flexible disk, hard disk, solid-state drive, magnetic tape, or any other magnetic data storage medium, a CD-ROM, any other optical data storage medium, any physical medium with patterns of holes, a RAM, a PROM, and EPROM, a FLASH-EPROM, NVRAM, any other memory chip or cartridge.

[0080] Storage media is distinct from but may be used in conjunction with transmission media. Transmission media participates in transferring information between storage media. For example, transmission media includes coaxial cables, copper wire and fiber optics, including the wires that comprise bus 550. Transmission media can also take the form of acoustic or light waves, such as those generated during radio-wave and infra-red data communications.

[0081] Reference throughout this specification to “one embodiment”, “an embodiment”, or similar language means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Thus, appearances of the phrases “in one embodiment”, “in an embodiment” and similar language throughout this specification may, but do not necessarily, all refer to the same embodiment.

[0082] Although the present disclosure has been described in terms of certain preferred embodiments and illustrations thereof, other embodiments and modifications to preferred embodiments may be possible that are within the principles and spirit of the invention. The above descriptions and figures are therefore to be regarded as illustrative and not restrictive.

[0083] Thus the scope of the present disclosure is defined by the appended claims and includes both combinations and sub-combinations of the various features described hereinabove as well as variations and modifications thereof, which would occur to persons skilled in the art upon reading the foregoing description. , Claims:1. An artificial intelligence-based bovid identification system using muzzle recognition and natural biometrics, comprising:

a computing device comprises an image capturing module configured to enable a user to perform at least one of: upload one or more images of one or more bovids on the computing device; open a camera of the computing device and capture one or more images of the one or more bovids; initiate the search process by opening the camera, whereby the computing device comprises a real-time bovid identification module configured to receive the one or more images from the image capturing module and deliver to a cloud server over a network;

the cloud server comprises an image processing module configured to receive the one or more images from the computing device, whereby the image processing module comprises an image collection module configured to collect the one or more images and delivers them to a muzzle identification module, wherein the muzzle identification module configured to draw one or more patterns using one or more machine learning techniques and identify a muzzle part of the bovid using one or more deep learning models, the muzzle identification module configured to send one or more muzzle portion in identified images to a pattern extraction module;

the pattern extraction module configured to extract one or more muzzle patterns from the one or more muzzle-identified images using an image-processing technique, whereby the pattern extraction module configured to send one or more extracted muzzle patterns to a feature identification module, wherein the feature identification module configured to identify one or more beads and ridges in the one or more extracted muzzle patterns;

the feature identification module configured to send one or more identified beads and ridges to a feature matching module, wherein the feature matching module configured to define one or more key points and match the one or more key points in the one or more identified beads and ridges with one or more master images; and

the feature matching module configured to send one or more matched key points with the one or more master images to a bovid identification module, wherein the bovid identification module configured to identify an individual bovid and bovid information upon a successful match of the one or more key points using one or more deep learning models thereby representing the bovid information on the computing device in livestock management.

2. The system as claimed in claim 1, wherein the image capturing module is configured to enable the user to capture one or more videos of the one or more bovids on the computing device.

3. The system as claimed in claim 1, wherein the image processing module is configured to identify specific components of an object in the one or more images, specifically the muzzle.

4. The system as claimed in claim 1, wherein the image processing module is configured to identify various attributes of the muzzle, including its size, orientation, and scale, without causing any delays or latency in the identification process.

5. The system as claimed in claim 1, wherein the image processing module comprises one or more machine learning and image processing techniques to identify the individual bovid based on real-time images and video capture.

6. The system as claimed in claim 1, wherein the one or more deep learning models comprise both cosine and scale-invariant feature transform algorithms.

7. The system as claimed in claim 1, wherein the pre-processing techniques comprise a Gabor filter and strongly supports machine learning model training.

8. The system as claimed in claim 1, wherein the pattern identification module is configured to perform classification of the one or more extracted muzzle patterns and are skeletonized, thinned so as to extract a morphological operation that is used to remove selected foreground pixels from one or more binary images.

9. The system as claimed in claim 1, wherein the one or more extracted muzzle patterns are further classified using a support vector machine (SVM).

10. The system as claimed in claim 1, wherein the image processing module comprises one or more post-processing techniques are configured to capture at least one of: orientation, size; and scale; and to match using the one or more deep learning models.

11. A method for identifying bovid using muzzle recognition and natural biometrics, comprising:

receiving one or more images of one or more bovids from a computing device by an image collection module;

drawing one or more patterns using one or more machine learning techniques and identifying a muzzle part of the bovid using one or more deep learning models by a muzzle identification module;

extracting one or more muzzle patterns from one or more muzzle-identified images by a pattern extraction module using an image-processing technique, the image-processing technique comprises a Gabor filter and strongly supports machine learning model training;

identifying one or more beads and ridges in one or more extracted muzzle patterns by a feature identification module;

defining one or more key points and matching the one or more key points in one or more identified beads and ridges with one or more master images by a feature matching module;

identifying an individual bovid upon a successful match of the one or more key points by a bovid identification module using a deep learning model that uses both cosine and scale-invariant feature transform algorithms; and

sending bovid information to the computing device from a cloud server by an image processing module.