Description:Field of the Invention
This invention relates to Computer Vision based Alert System to Protect Wild Animals from Vehicle Collision.
Background of the Invention
Wildlife and automobile crash are main alarms in the areas, where the road transient through the forests. To overcome this delinquent, the researchers has executed different procedures like reflectors, fences and warning signs. However, the main and exploratory analyses have settled these methods have not shown weighty impact on minimizing the wildlife-vehicle crashes. The other methods for adapting wildlife vehicle crash risk must be industrialized, such as dynamic driver warnings need to instrument to overcome this task and enhance the safety of both wild animals and human life. Due to fear of the vehicles, the wild animals are moving towards human settled and disturbing them and also causing loss to goods and human lives.
The sudden increase in automobile traffic jams through wildlife areas has diminished the risk of wildlife-vehicle accidents, posing a remarkable threat to biodiversity. Old fashion methods like reflectors, warning boards and fences often lost to account for the random movements of wildlife and also the speeds of automobiles. Innovative technologies provide a good substitute, particularly those that make use of vision-based systems and the Internet of Things (IoT). With the systems they may offer and receive real-time notifications and critical care, riders and drivers can behave sensibly to prevent collisions. A method like this not only attempts to save animals but also improves road safety by reducing the possibility of accidents caused by errant animal movements.
WO2021118675A1 Disclosed embodiments include technologies for improving safety mechanisms in computer assisted and/or automated driving (CA/AD) vehicles for protecting vulnerable road users (VRUs). Embodiments include various mechanisms to enable early Responsibility Sensitive Safety (RSS) checks for the CA/AD vehicles driving policy to protect in-danger VRUs. Embodiments also include controlled forwarding mechanisms to notify other CA/AD vehicles and roadside infrastructure when a potentially dangerous situation is detected. Other embodiments are described and/or claimed.
RESEARCH GAP:
• Enhanced Processing Capabilities: Deep learning algorithms can capably process and analyse huge datasets of visual data from cameras in forested areas.
• Reduced Latency with On-Site Processing: Integration with edge devices allows for on-the-spot processing of visual data, reducing latency and ensuring drivers receive instant warnings.
CN117104256A The application provides a vehicle risk event monitoring method, a vehicle risk event monitoring device, computer equipment and a storage medium, and belongs to the technical field of computers. The method comprises the following steps: acquiring a plurality of risk events in a preset duration in a vehicle running process, wherein the plurality of risk events comprise: at least one driver risk event and at least one vehicle risk event, at least two driver risk events, or at least two vehicle risk events; determining risk scores in a preset duration according to each driver risk event, the weight corresponding to each driver risk event, each vehicle risk event and the weight corresponding to each vehicle risk event; determining the risk level of the vehicle according to the risk score in the preset time period; and outputting corresponding alarm information according to the risk level of the vehicle. The application can achieve the effect of improving the intelligentization of vehicle risk judgment and being flexibly applicable to various transportation scenes.
RESEARCH GAP:
• High Accuracy in Wildlife Detection: These algorithms can identify plentiful species of wildlife with high accurateness, even under thought-provoking conditions such as low light or dense vegetation.
• Real-Time Animal Detection: The capacity to detect and distinguish animals in real-time allows for the appropriate production of alerts, avoiding accidents and saving lives.
None of the prior art indicate above either alone or in combination with one another disclose what the present invention has disclosed. This invention relates to Computer Vision based Alert System to Protect Wild Animals from Vehicle Collision.
SUMMARY OF THE INVENTION
This summary is provided to introduce a selection of concepts, in a simplified format, that are further described in the detailed description of the invention.
This summary is neither intended to identify key or essential inventive concepts of the invention and nor is it intended for determining the scope of the invention.
To further clarify advantages and features of the present invention, a more particular description of the invention will be rendered by reference to specific embodiments thereof, which is illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail with the accompanying drawings.
Deep learning, a subcategory of artificial intelligence, theatres a pivotal role in enhancing the aptitudes of vision and IoT-based alert systems premeditated to keep wild animals from vehicle accidents. By leveraging large datasets and cultured neural networks, deep learning algorithms can efficiently process and analyses visual data from cameras deployed in forested areas. These algorithms can be trained to identify numerous species of wildlife with high correctness, even in stimulating conditions such as low light or dense foliage. The skill to precisely detect and identify animals in real-time is critical for producing timely alerts that can prevent the collisions and save the lives of the both wild animals and humans. The mixture of deep learning and IoT not only enhances detection accuracy but also enables projecting analytics, where potential animal crossings can be calculated based on ancient data and current trends.
Furthermore, deep learning can enable the expansion of more innovative features within the alert system, such as characteristic between different types of animals and their individual risk levels. For instance, the system could give prioritize alerts for the larger animals like deer, monkey or elephants and etc., which can pose a greater threat to both vehicles and themselves. Deep learning models can be used to analyses the effectiveness of the alert system by the help of monitoring collision rates and animal movement path, producing valuable information for continuous improvement. By integrating deep learning into vision and IoT-based alert systems, we can create a more intelligent, responsive, and effective solution for mitigating wildlife-vehicle collisions and promoting safer co-occurrence between humans and wildlife.
BRIEF DESCRIPTION OF THE DRAWINGS
The illustrated embodiments of the subject matter will be understood by reference to the drawings, wherein like parts are designated by like numerals throughout. The following description is intended only by way of example, and simply illustrates certain selected embodiments of devices, systems, and methods that are consistent with the subject matter as claimed herein, wherein:
FIGURE 1: OVERALL DEVICE ARCHITECTURE
FIGURE 2: WEBSERVER AND DASHBOARD
The figures depict embodiments of the present subject matter for the purposes of illustration only. A person skilled in the art will easily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.
DETAILED DESCRIPTION OF THE INVENTION
The detailed description of various exemplary embodiments of the disclosure is described herein with reference to the accompanying drawings. It should be noted that the embodiments are described herein in such details as to clearly communicate the disclosure. However, the amount of details provided herein is not intended to limit the anticipated variations of embodiments; on the contrary, the intention is to cover all modifications, equivalents, and alternatives falling within the scope of the present disclosure as defined by the appended claims.
It is also to be understood that various arrangements may be devised that, although not explicitly described or shown herein, embody the principles of the present disclosure. Moreover, all statements herein reciting principles, aspects, and embodiments of the present disclosure, as well as specific examples, are intended to encompass equivalents thereof.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments. As used herein, the singular forms “a",” “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms “comprises,” “comprising,” “includes” and/or “including,” when used herein, specify the presence of stated features, integers, steps, operations, elements and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and/or groups thereof.
It should also be noted that in some alternative implementations, the functions/acts noted may occur out of the order noted in the figures. For example, two figures shown in succession may, in fact, be executed concurrently or may sometimes be executed in the reverse order, depending upon the functionality/acts involved.
In addition, the descriptions of "first", "second", “third”, and the like in the present invention are used for the purpose of description only, and are not to be construed as indicating or implying their relative importance or implicitly indicating the number of technical features indicated. Thus, features defining "first" and "second" may include at least one of the features, either explicitly or implicitly.
Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which example embodiments belong. It will be further understood that terms, e.g., those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.
Deep learning, a subcategory of artificial intelligence, theatres a pivotal role in enhancing the aptitudes of vision and IoT-based alert systems premeditated to keep wild animals from vehicle accidents. By leveraging large datasets and cultured neural networks, deep learning algorithms can efficiently process and analyses visual data from cameras deployed in forested areas. These algorithms can be trained to identify numerous species of wildlife with high correctness, even in stimulating conditions such as low light or dense foliage. The skill to precisely detect and identify animals in real-time is critical for producing timely alerts that can prevent the collisions and save the lives of the both wild animals and humans. The mixture of deep learning and IoT not only enhances detection accuracy but also enables projecting analytics, where potential animal crossings can be calculated based on ancient data and current trends.
Furthermore, deep learning can enable the expansion of more innovative features within the alert system, such as characteristic between different types of animals and their individual risk levels. For instance, the system could give prioritize alerts for the larger animals like deer, monkey or elephants and etc., which can pose a greater threat to both vehicles and themselves. Deep learning models can be used to analyses the effectiveness of the alert system by the help of monitoring collision rates and animal movement path, producing valuable information for continuous improvement. By integrating deep learning into vision and IoT-based alert systems, we can create a more intelligent, responsive, and effective solution for mitigating wildlife-vehicle collisions and promoting safer co-occurrence between humans and wildlife.
1. Training and Development
3.1 Dataset Collection: This preliminary step involves gathering a all-inclusive dataset including images and videos of various animals in unlike environments and circumstances. This can include material captured during different times of the day (daytime and nighttime), under various climate conditions (rain, fog, snow), and in different types of greenery (dense forests, open fields, roadside bushes). The wide dataset is critical for training the deep learning model, ensuring it can identify and discriminate between various species of wildlife accurately. By taking a wide range of situations, the model becomes tougher and more capable of performing well in real-world scenarios. This step often involves firm with wildlife specialists to ensure the dataset is representative of the animals that are most likely to be come across.
3.2 Training Using Roboflow: the dataset which is collected, is used to train the deep learning model Roboflow, it provides tools for annotating data, augmenting images, and developing models for computer vision tasks. For the training in push the labeled images to the roboflow, the pre-defined algorithm, stats and feature learn the details form the images like the pattern the features making cluster of the images for the better classification of different animals. In the training part the model learns to classify different animals and understand their activities. By extending Roboflow’s abilities, the trained model can be fine-tuned by adding some more dataset of the images to enhance its accuracy and efficiency, making it adept at distinguishing wildlife in various settings. This step may involve several iterations to refine the model, ensuring it performs well even in challenging conditions such as poor lighting or partial obstruction of animals.
Transferring Model to Raspberry Pi: when the model is done training, the next step is to import the model to the Raspberry Pi, a Pi is a small and cheap computer from the competition which does not require any further instruction to run the program it just need electrical power that can be easily arranged. The Raspberry has built in software and module which are require to run the program and can handle all the operation with ease and it can handle data input from the vision node independently. The portability and cost-effectiveness of the Raspberry Pi make it ideal for deployment in remote areas where the chances of incidents are higher and we can get the area information from the collision reported by communities.
Use The Raspberry Pi to Perform: After loading the model, the Raspberry Pi can now be used to perform its intended function in the real-world. It is designed to continuously capture visual information and implement surveillance algorithms. This includes integration of Raspberry Pi with vision node incorporating a stable source of power and defense against outdoor elements. This phase ensures that the system operates within the field constantly watching out for animal presence. For quick alerts to avoid wildlife-vehicle conflicts, it requires fast real-time processing capacity on the part of Raspberry Pi. Its capability for self-operation and low-power consumption makes it suitable for long-term deployment in harsh environments.
Deployment and Functionality
Vision Node: A vision node has been made up from a strategically placed camera set at an area where there are chances of wildlife crossing over roads. It captures images to check any animal movement taking place. The cameras often used are high resolution and have night vision capabilities that allow them to monitor efficiently all round the clock. It is this component that does data primary capture activity first. By keeping glance over this range always, every animal motion near roadways can be detected and recorded by vision nodes. The vision node must be robust, weatherproof, and capable of operating under various environmental conditions to provide reliable data.
Central System: The mainframe is the central computing part where data from the vision node is analyzed. It gets visual information, analyses it and runs a deep learning model to find out animals. Besides, this may be more powerful computer or server handling data from several vision nodes at once. The function of the central system is to accurately identify animals and determine if an alert is needed. Central data processing ensures that detection has high precision and reliability. Also, it acts as a center for aggregating data hence supporting overall monitoring and analyzing animal movements over time.
Customized LoRa: To communicate between the central system and other components in the alarm system such as display boards and control room, LoRa (Long Range) communication technology is employed. LoRa technology is self-selected due to its long distance low power communication that works best in rural or large areas. Signals can be transmitted through long distances by using LoRa technology which provides reliable long range communication method that allows for the same. Especially when traditional means of communication are no longer effective in far-flung or vast regions use of Lora becomes crucial because it guarantees that alerts can still travel considerable distances even when such are not present on such parts of a country’s geographical space like cell coverage zones only upon all these above described factors considered lora will guarantee connectivity even if cellular network could not be reached in some places
Display Board: Display boards are placed on roadways to alert drivers using the road. This receives alerts from the central system through LoRa communication. They have bright LED displays that make them visible even during bad weather or in bright sunlight. By displaying in real-time alerts about the presence of wildlife near the roadways, the display board warns drivers to slow down or take necessary precautions before any mis-happening. This proactive measure will helps in reducing the likelihood of collisions tremendously. The display boards can be programmed to show different types of alerts based on the detected animal, providing specific warnings for higher-risk animals for example green color NORMAL for relax situation and red color ALERT! for danger situation.
Control Room: The control room aids as the monitoring and administration center for the alert system. It receives alert signals and visual data from the dominant system, allowing staff to oversee and reply to alerts. The control room is armed with monitors, communication tools, and data analysis software to manage the system efficiently. The control room confirms that there is human misunderstanding and intervention when needed. Staff can monitor the effectiveness of the system, manage alerts messages, and coordinate with relevant responsible authorities to take action if essential. This centralized management allows for real-time decision-making and ensures that any technical issues can be promptly addressed
Vision Node: The vision node has cameras on it that are meant to record images from places where animals are likely to cross roads. The cameras are strategically placed to monitor these critical areas continuously. Capturing photos and identifying any animal movement near the road is the vision node's main job. This is achieved through real-time monitoring, which ensures that any detected movement is promptly processed. The vision node must work under different light and weather conditions for it to operate 24 hours a day.
LoRa Communication: LoRa communication is used to send data from the vision node to the server. It is mostly applied in remote and large-scale monitoring sites due to its long-range capability with low power consumption. Through LoRa communication, the vision node ensures efficient and reliable transmission of data captured by it to the central server. Even though at a distance, this wireless communication method keeps up connection between field tools and CPU
Server: The server performs as a central processing unit for the whole system. It gets data that comes from vision node via LoRa communication after which it processes this information and uses it to show animal movements. Server’s role is analyzing incoming visual data with sophisticated algorithms for tracking animal movements. This profiled information guides decision-making within an organization while also promoting cost-saving opportunities, this will enable one to understand and respond to wildlife activities. Thence, thus it facilitates data processing capabilities for a huge amount of data that can be efficiently processed for complex analysis.
Dashboard: This is the dashboard on which the processed data are shown and visualizations created. It gives information about animal movements in an easy-to-understand way with real time updates. The dashboard presents detailed information about animals near highways derived from the processed data. It is meant to be user-friendly to allow for quick interpretations. The dashboard should be able to provide latest reports thereby enabling users make informed choices regarding prevention of wildlife-vehicle collisions.
ADVANTAGES OF THE INVENTION
Continuous Learning and Improvement: Deep learning models become more robust and adaptable if they can continuously learn even in the face of new data.
Predictive Analytics: A blend of deep learning and IoT leads to predictive analytics that enable calculation or estimation of potential animal crossings using historical data as well as current trends.
Differentiation Between Animal Types: Deep Learning can differentiate between species of animals and their individual risk levels. Such alerts would then prioritize larger animals such as deer or elephants.
Effectiveness Analysis: By looking at the collision rates and the ways animals move, deep learning models are able to analyze whether or not they are effective in alert systems; thereby giving feedback for its improvement
Innovative Feature Development: Deep learning promotes the development and introduction of new features within an alert system that improve on its efficiency as well as interactivity.
Promoting Safer Co-Existence: Enhancing safe co-existence between humans and wildlife is one of the key advantages of integrating deep learning with vision and IoT-based alert systems in mitigating wildlife-vehicle interactions.
, Claims:1. A Computer Vision based Alert System to Protect Wild Animals from Vehicle Collision comprises a wildlife detection system, a deep learning model, a microcontroller, a vision node, a central system, a Lo-Ra Technology, a display board, wherein, provide integrated wildlife detection system employs a deep learning (CNN) model trained on diverse datasets and transferred to a microcontroller for field deployment.
2. The system as claimed in claim 1, wherein provides Vision nodes capture data, central systems analyse it for animal presence, and LoRa technology enables long-range communication for alerts.
3. The system as claimed in claim 1, wherein Display boards warns drivers of wildlife near roadways, reducing collision risks, while control rooms oversee system functionality and facilitate timely intervention.
4. The system as claimed in claim 1, wherein the control room aids as the monitoring and administration center for the alert system; which receives alert signals and visual data from the dominant system, allowing staff to oversee and reply to alerts.
5. The system as claimed in claim 1, wherein the control room is armed with monitors, communication tools, and data analysis software to manage the system efficiently; and said control room confirms that there is human misunderstanding and intervention when needed; and this centralized management allows for real-time decision-making and ensures that any technical issues can be promptly addressed.
6. The system as claimed in claim 1, wherein the vision node has cameras on it that are meant to record images from places where animals are likely to cross roads; and the cameras are strategically placed to monitor these critical areas continuously. Capturing photos and identifying any animal movement near the road is the vision node's main job; which is achieved through real-time monitoring, which ensures that any detected movement is promptly processed; and the vision node works under different light and weather conditions for it to operate 24 hours a day.
7. The system as claimed in claim 1, wherein LoRa communication is used to send data from the vision node to the server; and applied in remote and large-scale monitoring sites due to its long-range capability with low power consumption; wherein through LoRa communication, the vision node ensures efficient and reliable transmission of data captured by it to the central server; and even though at a distance, this wireless communication method keeps up connection between field tools and CPU.
8. The system as claimed in claim 1, wherein the server performs as a central processing unit for the whole system; and it gets data that comes from vision node via LoRa communication after which it processes this information and uses it to show animal movements; and it facilitates data processing capabilities for a huge amount of data that can be efficiently processed for complex analysis.
9. The system as claimed in claim 1, wherein the dashboard gives information about animal movements in an easy-to-understand way with real time updates; and the dashboard presents detailed information about animals near highways derived from the processed data. It is meant to be user-friendly to allow for quick interpretations; and the dashboard is able to provide latest reports thereby enabling users make informed choices regarding prevention of wildlife-vehicle collisions.