Smart police assistance, crime observation and prevention system (spacops)

Abstract

A smart policing system (150) includes multiple Internet Protocol (IP) cameras (100) with edge computing capabilities, installed at designated hotspots and powered via a power and connectivity panel (200) and a Power over Ethernet (PoE) Injector (300). The cameras (100) transmit video data to a cloud-based Artificial Intelligence (AI) system (400). The AI system (400) features a video analytics module (410) for detecting and classifying traffic violations and crimes, digital cloud storage (420) for recording analysed data, a video query module (430) for processing natural language queries, an AI model (440) for translating queries into attributes and identifying relevant AI models, and an alert mechanism (450) for issuing notifications through emails, web, and mobile applications (500). The edge computing capabilities of the IP cameras (100) perform frame filtering using proprietary AI algorithms (470) to reduce data transmission to the cloud-based AI system (400)

Description

[0001] SMART POLICE ASSISTANCE, CRIME OBSERVATION AND PREVENTION SYSTEM (SPACOPS)

[0002] TECHNICAL FIELD

[0003] The present invention relates to smart surveillance systems more specifically, a Smart Police Assistance, Crime Observation and Prevention System specifically designed for Police to detect traffic violations and crimes along with raising near real-time alerts about the traffic violations and crimes, on emails, web, and mobile notifications for police authorities. This system helps the Police to monitor and take action accordingly.

[0004] BACKGROUND

[0005] Background description includes information that may be useful in understanding the present invention. It is not an admission that any of the information provided herein is prior art or relevant to the presently disclosed invention, or that any publication specifically or implicitly referenced is prior art.

[0006] In contemporary law enforcement practices, closed-circuit television (CCTV) cameras are extensively employed for monitoring public areas, traffic intersections, and other critical zones to ensure safety and security. While these cameras serve as a fundamental tool for surveillance, they present several significant limitations: a) Raw Footage Provision: The primary function of traditional CCTV cameras is to capture and record raw video footage. This footage, although valuable, does not inherently include any analytical capabilities to detect criminal activities or traffic violations. As a result, the footage requires further processing to derive actionable insights. b) Lack of Automated Alerts: Current CCTV systems do not possess the capability to autonomously identify and report incidents of crime or traffic violations. The absence of intelligent detection mechanisms means that the footage must be continuously monitored by human operators to identify any suspicious activities or breaches of traffic laws. c) Need for Manual Surveillance: The effectiveness of CCTV cameras is heavily reliant on manual surveillance. Law enforcement personnel must monitor the video feeds in real time to detect and respond to incidents. This process is labor-intensive and necessitates constant vigilance. d) Exhaustion and Fatigue: Continuous manual surveillance can be extremely tiresome and mentally exhausting for police staff. The prolonged periods of attention required to monitor multiple camera feeds can lead to decreased efficiency and an increased likelihood of missed incidents due to human error.

[0007] These drawbacks highlight the limitations of current CCTV systems in providing proactive and efficient surveillance. Addressing these challenges necessitates the development of more advanced, automated solutions that can enhance the effectiveness of surveillance operations.

[0008] To overcome these limitations, we require an intelligent smart policing or monitoring system that can generate near real-time alerts for traffic violations and criminal activities through the use of a proprietary algorithm. This advanced system is expected to offer several key improvements over traditional CCTV setups, for example, automated detection, real-time alerts, reduced manual effort, and enhanced surveillance efficiency.

[0009] In short, the required method and system is expected to address the inherent limitations of traditional CCTV cameras by incorporating advanced detection algorithms and realtime alert mechanisms, thus providing a more effective and efficient solution for modern law enforcement challenges.

[0010] SUMMARY OF THE INVENTION

[0011] The following presents a simplified summary of the subject matter in order to provide a basic understanding of some aspects of subject matter embodiments. This summary is not an extensive overview of the subject matter. It is not intended to identify key / critical elements of the embodiments or to delineate the scope of the subject matter. Its sole purpose is to present some concepts of the subject matter in a simplified form as a prelude to the more detailed description that is presented later.

[0012] A method for smart policing using a smart policing system addresses the limitations of traditional CCTV cameras by incorporating advanced detection algorithms and real-time alert mechanisms. This method involves installing a network of IP cameras with edge computing capabilities at strategic locations to monitor traffic violations and criminal activities. These cameras are continuously powered and connected through a specialized panel, using PoE Injectors for power and internet connectivity via Ethernet cables. The video data captured is transmitted to a cloud-based Al system for processing and analysis. The Al system, controlled by at least one processor, includes a video analytics module for detecting and classifying incidents, a storage system for recording data, a video query module for processing natural language queries, an Al model for translating queries into attributes, and an alert mechanism for issuing notifications through various channels.

[0013] In an embodiment, the smart policing method identifies Al models that correspond to specific attributes from video footage, such as vehicle color, type, make, clothing, and physical appearance. This ensures relevant Al models enhance the accuracy and efficiency of incident detection. Proprietary Al algorithms used for edge-level frame filtering involve advanced techniques, including PSNR, average difference values, similarity scores, and mean square errors. These metrics assess image quality and frame consistency, transmitting relevant frames to the cloud-based Al system for further analysis while discarding non-relevant frames to optimize bandwidth and resources.

[0014] In an embodiment, the video analytics module automatically detects and classify various incidents, including number plate recognition, helmet-less riders, drivers without seat belts, vehicles in no-parking zones, public brawls, accidents, chain snatching, and identifying vehicle types or colors and persons based on natural language descriptions. This ensures comprehensive detection capabilities for traffic violations and criminal activities. The method includes a relevance checker within the cloud-based Al system to sequentially process frames, determining their relevance to incidents using a master algorithm. This ensures only pertinent frames are analyzed and stored. The video query module translates natural language queries into attribute -based searches within crime and traffic databases, facilitating specific searches like detecting persons with particular apparel, multiple riders on two-wheelers, vehicles running red lights, and searching by vehicle attributes. This allows quick and accurate retrieval of relevant information. In an embodiment, the method ensures continuous operation of IP cameras through grid electricity or a solar setup, viable even in areas with limited power infrastructure. The power and connectivity panel includes an SMPS to convert voltage and an Internet Router for connectivity via SIM cards, fiber optics, and broadband. An LED light is also included for illumination, enhancing footage capture in low-light conditions. The method transmits information from the cloud-based Al system to web and mobile applications, offering user authentication, real-time dashboards, map views of camera locations, alert notifications, natural language video queries, image uploads for tracking, statistical reporting, graphical data representations, and recent offense tracking. These tools help law enforcement monitor, analyze, and respond to incidents effectively.

[0015] In an embodiment, the smart policing system is software-defined, allowing for easy upgrades and expansion with new technologies and functionalities. The method includes generating and transmitting alert notifications with detailed information, such as location and images of detected incidents, through various channels to ensure timely action by authorities. Attribute-based searches in the video query module are customized for detecting specific scenarios like particular apparel combinations, multiple riders on two- wheelers, traffic signal violations, and specific vehicle attributes. This ensures precise and relevant searches for identifying and tracking suspects and vehicles.

[0016] In an embodiment, the cloud-based Al system operates in real-time using proprietary Al algorithms in the video analytics module and query translation for prompt alerting and response. The video query module processes text and voice-based queries, providing intuitive and efficient tools for data access and analysis. Edge devices within IP cameras use proprietary filtering algorithms to transmit only relevant frames to the cloud-based Al system, optimizing bandwidth and processing efficiency. The system handles various crimes and traffic violations, including public disturbances, theft, speed infractions, and safety regulation non-compliance, enhancing public safety and maintaining order.

[0017] In short, the SMART Police Assistance, Crime Observation and Prevention System (SPACOPS) is a computer vision-based Artificial Intelligence (Al) system used to assist Police to detect and identify Traffic Violations and Crimes at potential Hotspots of the city and take appropriate and timely actions. The SPACOPS consists of both hardware and software parts. The hardware part consists of cameras with Internet Protocol (IP). A power and connectivity panel are required to power the camera and provide internet connectivity through SIM or broadband. The system can use either grid power or solar power accordingly to power the camera and LED Light. The images taken by the camera are transmitted to the cloud. In the cloud, there is a proprietary algorithm (470) that checks the image data for traffic violations and crimes. The algorithm (470) can easily detect chain snatching, brawls, street fighting, crimes against women and children, etc. The algorithm (470) has the special feature of video query to understand natural language queries to extract precise information from the video footage data in near real-time. It can be used to identify any vehicle or person based on the description provided like the colour of the vehicle, type, or colour of clothes of the person or accessory the person is wearing, etc. by simply asking statements like “identify a person with long brown coat and hat”. The algorithm (470) can also identify and detect persons or vehicles from an uploaded photograph. The system has an active alert system if traffic violations or crimes are observed then alerts in terms of emails, web, and mobile app notifications are released for the Police. The Web Application and Mobile Application are part of the overall SPACOPS providing near real-time alerts, history of alerts, statistics, map view, etc. SPACOPS is a software-defined system, so no hardware changes will be required at later stages.

[0018] BRIEF DESCRIPTION OF FIGURES

[0019] The foregoing and further objects, features and advantages of the present subject matter will become apparent from the following description of exemplary embodiments with reference to the accompanying drawings, wherein like numerals are used to represent like elements.

[0020] It is to be noted, however, that the appended drawings along with the reference numerals illustrate only typical embodiments of the present subject matter, and are therefore, not to be considered for limiting its scope, for the subject matter may admit to other equally effective embodiments. Figure 1 exemplarily shows a schematic overview of the smart policing system or the SMART Police Assistance, Crime Observation and Prevention System (SPACOPS), as an example embodiment of the present disclosure.

[0021] Figure 2A exemplarily shows a Block Diagram showing different components of the smart policing system, as an example embodiment of the present disclosure.

[0022] Figure 2B exemplarily shows a Block Diagram showing interaction of the cloud-based Al systems and the cameras of the smart policing system, as an example embodiment of the present disclosure.

[0023] Figure 2C exemplarily shows inside power and connectivity panel of the smart policing system, as an example embodiment of the present disclosure.

[0024] Figure 3 exemplarily shows reduction in transmitted frames at the edge, as an example embodiment of the present disclosure.

[0025] Figure 4 exemplarily shows Flow chart of ML Algorithm in cloud, as an example embodiment of the present disclosure.

[0026] Figure 5 exemplarily shows Ad-hoc video query system of the smart policing system, as an example embodiment of the present disclosure.

[0027] Figure 6 exemplarily shows dimensions of pole and positions of cameras, lights, and power and connectivity panel. , as an example embodiment of the present disclosure.

[0028] Figure 7 exemplarily shows dimensions of pole and positions of cameras, lights, and power and connectivity panel with solar setup, as an example embodiment of the present disclosure.

[0029] DETAILED DESCRIPTION Illustrative examples of the subject matter will now be disclosed. In the interest of clarity, not all features of an actual implementation are described in this specification. It will be appreciated that in the development of any such actual implementation, numerous implementation- specific decisions may be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which will vary from one implementation to another. Moreover, it will be appreciated that such a development effort, even if complex and time-consuming, would be a routine undertaking for those of ordinary skill in the art having the benefit of this disclosure.

[0030] Exemplary embodiments now will be described with reference to the accompanying drawings. The disclosure may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey its scope to those skilled in the art. The terminology used in the detailed description of the particular exemplary embodiments illustrated in the accompanying drawings is not intended to be limiting. In the drawings, like numbers refer to like elements.

[0031] It is to be noted, however, that the reference numerals used herein illustrate only typical embodiments of the present subject matter, and are therefore, not to be considered for limiting its scope, for the subject matter may admit to other equally effective embodiments.

[0032] The specification may refer to “an”, “one” or “some” embodiment(s) in several locations. This does not necessarily imply that each such reference is to the same embodiment(s), or that the feature only applies to a single embodiment. Single features of different embodiments may also be combined to provide other embodiments.

[0033] As used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless expressly stated otherwise. It will be further understood that the terms “includes”, “comprises”, “including” and / or “comprising” when used in this specification, 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 will be understood that when an element is referred to as being “connected” or “coupled” to another element, it can be directly connected or coupled to the other element or intervening elements may be present. Furthermore, “connected” or “coupled” as used herein may include operatively connected or coupled. As used herein, the term “and / or” includes any and all combinations and arrangements of one or more of the associated listed items.

[0034] Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skills in the art to which this disclosure pertains. It will be further understood that terms, such as 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.

[0035] The figures depict a simplified structure only showing some elements and functional entities, all being logical units whose implementation may differ from what is shown. The connections shown are logical connections; the actual physical connections may be different. It is apparent to a person skilled in the art that the structure may also comprise other functions and structures.

[0036] Also, all logical units described and depicted in the figures include the software and / or hardware components required for the unit to function. Further, each unit may comprise within itself one or more components which are implicitly understood. These components may be operatively coupled to each other and be configured to communicate with each other to perform the function of the said unit.

[0037] DEFINITIONS:

[0038] As used herein, “Smart Policing System” or “the Smart Police Assistance, Crime Observation and Prevention System” (SPACOPS) refers to an integrated surveillance and monitoring system designed to enhance law enforcement capabilities. It uses a network of cameras and cloud-based Al to detect and respond to traffic violations and criminal activities in real-time.

[0039] As used herein, “Internet Protocol (IP) Cameras” refer to cameras equipped with edge computing capabilities, allowing them to process and analyze video data locally before transmitting relevant information to the cloud. These cameras are installed at strategic locations to monitor high-traffic and high-crime areas.

[0040] As used herein, “Power and Connectivity Panel” refers to a unit that provides power and internet connectivity to the IP cameras. It ensures that the cameras are continuously operational and connected to the network for data transmission.

[0041] As used herein, “Power over Ethernet (PoE) Injector” refers to a device that supplies both power and data to the IP cameras through Ethernet cables. This simplifies the installation process by eliminating the need for separate power sources.

[0042] As used herein, “Cloud-Based Artificial Intelligence (Al) System” refers to the central processing unit of the smart policing system that receives video data from the IP cameras. It analyzes the data using advanced Al algorithms to detect and classify incidents.

[0043] As used herein, “Video Analytics Module” refers to a component of the Al system responsible for automatically detecting and classifying traffic violations and crimes based on the analyzed video data.

[0044] As used herein, “Digital cloud storage” refers to storage solutions within a cloud system used to record and organize analyzed data into respective crime and traffic databases.

[0045] As used herein, “Video Query Module” refers to a feature of the Al system that processes natural language queries, allowing users to search the stored data using spoken or typed queries.

[0046] As used herein, “Al Model” refers to the part of the Al system that translates queries into specific attributes and identifies the corresponding Al models for accurate data retrieval and analysis.

[0047] As used herein, “Alert Mechanism” refers to a notification system within the Al system that issues alerts via emails, web, and mobile applications based on identified events from the video analytics module. As used herein, “Proprietary Al Algorithms” refer to custom-developed algorithms used by the IP cameras for frame filtering to reduce data transmission to the cloud-based Al system by ensuring only relevant frames are sent.

[0048] As used herein, “Density Plot of Peak Signal -to-Noise Ratio (PSNR)” refers to a statistical tool used by the frame filtering algorithms to measure the quality of video frames by comparing the signal strength to background noise.

[0049] As used herein, “Density Plot of Average Difference Values” refers to a statistical method used to assess the change between successive video frames, helping to identify significant movements or actions.

[0050] As used herein, “Density Plot of Similarity Scores” refers to a plot that compares current frames with reference frames to find matches, ensuring relevant frames are transmitted for further analysis.

[0051] As used herein, “Density Plot of Mean Square Errors” refers to a metric used to quantify the deviation between video frames, helping to discard non-relevant frames.

[0052] As used herein, “Reframe Relevance Checker” refers to a component of the Al system that sequentially processes frames to determine their relevance to traffic violations or crimes using a master algorithm.

[0053] As used herein, “Switch Mode Power Supply (SMPS)” refers to a component of the power and connectivity panel that converts high-voltage AC power to lower- voltage DC power suitable for the IP cameras.

[0054] As used herein, “Internet Router” refers to a device within the power and connectivity panel that provides internet connectivity to the IP cameras via various options such as SIM card, fiber, or broadband.

[0055] As used herein, “Web and Mobile Applications” refers to user interfaces that provide functionalities such as user authentication, real-time dashboards, map views, alert notifications, natural language queries, and statistical reporting for effective monitoring and incident response.

[0056] As an overview before description of Figures 1-7, the smart policing system (150) or the SMART Police Assistance, Crime Observation and Prevention System (SPACOPS) is a computer vision-based Artificial Intelligence (Al) system comprising internet protocolbased cameras that can transmit video feeds and images to the Cloud through the internet provided by the Power and Connectivity panel. The images transmitted to the Cloud are analyzed through the proprietary algorithm (470). The Al Algorithm (470) detects vehicle number plates, vehicle type or color, riders without helmets, drivers not wearing seat belts, wrong parking or vehicles parked at No Parking, identification of accidents, public brawls, street fighting, chain snatching, robbery, crimes against women and children, and search and identification of person or vehicle on a lookout from reference photograph. SPACOPS has a unique video query feature that helps in detecting vehicles and persons through natural language queries like “Identify the person with muffler and brown bag”. After detection alerts are raised for the Police to take action. Figure 1 provides an overview of the SPACOPS.

[0057] Referring to Figures 1, 2A, 2B, and 3, a smart policing system or the Smart Police Assistance, Crime Observation and Prevention System [SPACOPS] (150) disclosed here comprises a plurality of Internet Protocol (IP) cameras (100) equipped with edge computing capabilities. These IP cameras (100) are strategically installed at designated hotspots, which are identified based on historical data and current analysis of high-traffic and high-crime areas. Each camera (100) is mounted on poles or building facades and powered through either direct grid electricity or an alternative solar setup. The SPACOPS (150) includes a power and connectivity panel (200) configured to supply both power and internet connectivity to the IP cameras (100), ensuring uninterrupted operation. Additionally, a Power over Ethernet (PoE) Injector (300) is utilized to deliver power and data to the IP cameras (100) via Ethernet cables, simplifying the installation process and reducing the need for separate power supplies.

[0058] Referring to Figures 1, 2A, 2B, and 3, at the heart of the SPACOPS (150) is a cloud-based Artificial Intelligence (Al) system (400) controlled by at least one processor (250) as shown in Figure 2B, which is designed to receive and process video data from the IP cameras (100). This Al system (400) comprises several key components: a video analytics module (410) for the automatic detection and classification of traffic violations and crimes; digital cloud storage (420) for recording and organizing analysed data into respective crime and traffic databases; a video query module (430) for processing natural language queries against the stored analysed data; an Al model (440) to translate queries into specific attributes (Figure 5) and identify corresponding Al models; and an alert mechanism (450) configured to issue notifications via emails, web, and mobile applications (500) based on identified events from the video analytics module (410). The edge computing capabilities of the IP cameras (100) involve proprietary Al algorithms (470) that perform frame filtering, reducing the amount of data transmitted to the cloudbased Al system (400) and thus optimizing bandwidth usage.

[0059] Referring to Figures 2B and 5, the SPACOPS (150) also includes an Al model (440) that identifies relevant models corresponding to specific attributes such as the colour of a vehicle, the type or make of a vehicle, the clothing of a person, and the overall appearance of a person. This attribute-based identification is crucial for accurate categorization and analysis of visual data. For instance, if a crime report mentions a red car or a person in a blue jacket, the Al model (440) can filter through video footage to locate and highlight these specific attributes, enhancing the efficiency and accuracy of incident detection and response.

[0060] Referring to Figure 3, the SPACOPS (150) also employs proprietary Al algorithms (470) at the edge for frame filtering. These Al algorithms (470) use various density plots to assess the quality and relevance of video frames. For instance, a Peak Signal-to-Noise Ratio (PSNR) plot (310) measures the signal quality against background noise, ensuring that only clear and useful frames are processed. Similarly, an average difference values plot (320) assesses the change between successive frames to identify significant movements or actions. A similarity scores plot (330) compares current frames with reference frames to find matches, while a mean square errors plot (340) quantifies the deviation between frames. By utilizing these metrics, the system ensures that only relevant frames are transmitted to the cloud-based Al system (400), discarding non- relevant frames and thereby optimizing data processing and transmission.

[0061] Referring to Figure 2B, the video analytics module (410) of the SPACOPS (150) is equipped with sophisticated capabilities for automatic detection and classification of various incidents. This includes number plate recognition for identifying vehicles, detecting helmet-less riders to enforce safety regulations, identifying drivers not wearing seat belts, spotting vehicles parked in no-parking zones, and recognizing public brawls and street fights. Additionally, the system can detect accidents, chain snatching, and robbery, and can identify vehicle types or colours. It can also identify persons based on descriptions provided in natural language, such as "a man in a red shirt" or "a woman with a yellow handbag," making the system highly effective in real-time monitoring and incident response.

[0062] Referring to Figures 2B and 5, the cloud-based Al system (400) in the SPACOPS (150) further comprises a reframe relevance checker (460). This component sequentially processes video frames to determine their relevance to traffic violations or crimes using a master algorithm. The master algorithm is designed to evaluate each frame's context and content, ensuring that only pertinent frames are retained for further analysis. This step enhances the system's efficiency by focusing resources on relevant data, thereby improving the accuracy and speed of incident detection and reporting. Referring to Figure 2B, the video query module (430) in the SPACOPS (150) translates natural language queries into attribute-based searches within the crime or traffic databases. This facilitates specific searches, allowing users to detect persons with particular apparel, identify vehicles with multiple riders, and recognize instances where vehicles run red lights. The module also enables searches for specific vehicles by colour and registration details, making it easier to track and locate suspects or vehicles involved in incidents. For example, a user can input a query like "show me all red cars that ran a red light yesterday," and the system will process this request and retrieve the relevant footage.

[0063] Referring to Figure 1, 2A, and 2B, In the SPACOPS (150), the IP cameras (100) are powered through either grid electricity or a solar setup, depending on the availability and feasibility at each installation site. The power and connectivity panel (200) that supplies power and internet connectivity to the IP cameras (100) includes a Switch Mode Power Supply (SMPS) (210) to convert 220 V-AC to a lower VDC suitable for the cameras (100). An Internet Router (220) is also part of the panel, providing internet connectivity via a SIM card, fibre, orbroadband connection. This setup ensures that the cameras (100) are always connected and operational, enabling continuous video monitoring and data transmission to the cloud-based Al system (400). The power and connectivity panel (200) in the SPACOPS (150) further includes an LED light (230) for illumination purposes, which is powered through the same panel. This LED light (230) ensures that the IP cameras (100) have adequate lighting for video capture in low-light conditions, such as during the night or in poorly lit areas. This feature enhances the quality of the video footage, making it easier to detect and analyse incidents accurately.

[0064] Information from the cloud-based Al system (400) in the SPACOPS (150) is transmitted to web and mobile applications (500). These applications (500) provide a range of functionalities, including user authentication and login, real-time dashboards, and map views indicating camera locations. Users receive alert notifications for traffic violations and crimes and performs natural language video queries. The applications (500) also allow for the uploading of images for tracking suspects or vehicles, offer statistical reporting and graphical representations of crime and traffic data, and track recent offenses or violations. These features provide law enforcement agencies with powerful tools for monitoring and responding to incidents effectively.

[0065] The SPACOPS (150) is designed to be software-defined, allowing for the addition of new capabilities without the need for hardware modifications. This flexibility makes the system highly adaptable and scalable, ensuring that it can evolve to meet changing policing needs and incorporate new technologies as they become available. For example, if a new type of video analytics is developed, it can be integrated into the system through a software update, enhancing its functionality without requiring physical changes to the hardware. The SPACOPS (150) generates and transmits alert notifications that include detailed information such as the location and images of detected traffic violations or crimes. These notifications are sent via emails, web, and mobile applications (500), ensuring that relevant authorities receive timely and actionable information. For example, an alert might include the exact location of a traffic violation, images of the vehicle involved, and a brief description of the incident, enabling quick and effective response by law enforcement.

[0066] Referring to Figure 5, the attribute-based searches within the SPACOPS (150) are customized to detect specific apparel combinations, the presence of multiple riders on a two-wheeler, vehicles violating traffic signals, and particular vehicle attributes including colour, make, and registration details. This customization enhances the system's ability to identify and analyse specific incidents accurately. For example, if a user searches for "blue cars with two occupants," the system will filter and display only the relevant footage, making it easier to investigate and respond to incidents.

[0067] The cloud-based Al system (400) in the SPACOPS (150) utilizes proprietary Al algorithms (470) in the video analytics module (410) and query translation to operate in real-time. This real-time operation facilitates prompt alerting and response, ensuring that authorities can act quickly to address incidents. The video query module (430) is capable of understanding and processing both text and voice-based natural language queries, providing versatile user interaction. For instance, an officer can verbally ask the system to "show footage of red cars in the past hour," and the system will process and display the relevant video clips.

[0068] The IP cameras (100) in the SPACOPS (150) comprise an edge device that employs a proprietary filtering algorithm. This algorithm ensures that only relevant frames are transmitted to the cloud-based Al system (400), optimizing bandwidth usage and enhancing overall system efficiency. By processing data at the edge, the system reduces the load on the central cloud infrastructure and speeds up the detection and analysis of incidents. The SPACOPS (150) is equipped to handle a wide variety of crimes and traffic violations, including but not limited to public disturbances, theft, vehicle speed infractions, and non-compliance with safety regulations such as helmet or seat belt usage. This comprehensive coverage makes the system a robust tool for maintaining public safety and order. For example, the system can detect a speeding vehicle, a person snatching a chain, or a driver.

[0069] The SPACOPS (150) comprises a computer vision-based Artificial Intelligence (Al) system that aids in policing activities. The SPACOPS (150) includes both hardware and software components, following a three-step approach: 1. Installation of Equipment: Internet Protocol (IP)-based cameras (100) with edge capabilities are set up at potential hotspots with available communication networks.

[0070] 2. Al and Video Analytics: Artificial Intelligence in the cloud processes video queries using natural language, image detection, and video analytics for tasks such as automatic number plate detection, identifying riders without helmets, drivers not wearing seat belts, detecting wrong parking or vehicles in no-parking zones, identifying accidents, public brawls, street fights, chain snatching, robbery, crimes against women and children, and searching for specific vehicle types or colours, and identifying persons based on provided descriptions.

[0071] 3. Alerts: The SPACOPS (150) raises alerts through emails and mobile app notifications (230) for crimes or traffic violations.

[0072] Figures 2A and 2C illustrate the system's mechanism. The camera (100) is powered by either grid electricity or a solar setup (400). Being an IP-based camera (100), it requires internet connectivity to transmit image data to the cloud (400), provided by the Power and Connectivity Panel (200). This power and connectivity panel (200) includes a Switch Mode Power Supply (SMPS) (210), Internet Router (220), and Power over Ethernet (POE) Injector (300). The SMPS (210) reduces the 220 VAC supply to the lower VDC required by electronic systems like the Internet Router (220). The Internet Router (220) provides connectivity via a SIM card, fiber, or broadband, while the POE Injector (300) supplies power and internet connectivity to the camera through an Ethernet port. The panel also connects to an LED light (230).

[0073] The camera (100) houses an edge device that reduces the number of frames transmitted to the cloud, as shown in Figure 3. Frames that correlate with reference frames are transmitted using approaches such as the density plot of Peak SNR, average difference value, similarity scores, and mean square errors. A proprietary algorithm filters and transmits relevant frames, discarding others to reduce data transmission.

[0074] Referring to Figure 4, in the cloud (400), incoming frames are processed by a proprietary Al algorithm (470). Frames are checked for relevance and classified as either crime or traffic infringements. If a crime is detected, it is further classified (e.g., public brawl, robbery, accident, street fight, chain snatching, and crimes against women and children). For traffic infringements, it identifies issues like no helmet, wrong parking, vehicle speeding, traffic light violations, or zebra crossing violations. Recognized incidents are stored in the crime or traffic database. The system's video query feature, explained in Figure 6, uses a proprietary model to convert text or voice queries into attributes (Figure 5), extracting precise information from large video footage databases in near real-time. Queries can include:

[0075] • "Detect persons with a hat and muffler"

[0076] • "Show me all instances where two-wheelers have three riders"

[0077] • "Identify instances of vehicles running red lights"

[0078] • "Find a white car with a registration number ending in 0445"

[0079] • "Find a red car with a Delhi registration number"

[0080] • "Show a two-wheeler where the rider has no helmet"

[0081] Furthermore, the Al system is capable of understanding and interpreting queries made in various Indian languages, such as Hindi, Gujarati, Marathi, Tamil, and others. Users can communicate with the Al in their preferred regional language, and the Al will accurately process and respond to these queries based on the context and nuances of each language." This emphasizes the Al's multilingual capabilities, making it clear that users can interact with the system in different Indian languages, and the Al will respond appropriately.

[0082] Alerts generated in the Cloud-based Artificial Intelligence (Al) system (400) are sent via emails, web, and mobile app notifications, providing police with location and image details of traffic violations or crimes. The hardware includes IP cameras mounted on poles 12 feet above ground or building infrastructure. These 5-megapixel cameras can transmit data to the cloud via the internet, powered by grid or solar power. Corresponding designs are shown in Figures 6 and 7.

[0083] The software includes cloud infrastructure, custom Al algorithms (470) with video query features, web applications and mobile applications (500) for both Android and iOS. The applications offer functionalities such as user login, dashboards, a map view of all cameras specifying locations, alerts for traffic violations or crimes, natural language video queries, uploading pictures, tracking suspects or vehicles, statistical information on crimes / traffic violations, graphical representations, and details on recent offenses / violations. The entire system is software-defined, allowing for capability additions without hardware changes.

[0084] Advantages of the SPACOPS (150) or SPACOPS:

[0085] 1. Comprehensive Monitoring System: Integrates cameras, solar power / electrical connections, power and connectivity panels, and LED lights for robust surveillance and crime prevention.

[0086] 2. Advanced Connectivity: Utilizes IP -based cameras with edge and internet capabilities for seamless image transmission to the cloud.

[0087] 3. Sustainable Power: Features solar power units with solar panels, inverters, and batteries for continuous operation.

[0088] 4. Efficient Infrastructure: Includes a power and connectivity panel with SMPS, distribution channels, an internet router with SIM capabilities, antennas, and a PoE injector.

[0089] 5. Enhanced Visibility: Equipped with high-powered LED lights (100W or more) for better illumination.

[0090] 6. Data Optimization: Uses a proprietary edge algorithm to reduce the amount of data transmitted to the cloud.

[0091] 7. Intelligent Video Analysis: Employs a machine-learning algorithm with natural language video query capabilities, especially Indian languages, for near real-time detection of crimes and traffic violations.

[0092] 8. Accurate Identification: Features a proprietary algorithm for identifying vehicles or persons based on uploaded images.

[0093] 9. Real-Time Alerts: Provides electronic alerts via email, web, and mobile notifications for immediate response.

[0094] 10. User-Friendly Interface: Offers a web application for police use, raising alerts for detected violations and crimes, with priority mobile notifications that require acknowledgment. 11. Detailed Alerts: Mobile application delivers comprehensive information, including violation / crime picture, location, and camera ID, aiding police in swift action.

[0095] It will be understood that each block of the block diagrams can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions / acts specified in the flowchart and / or block diagram block or blocks.

[0096] In the drawings and specification, there have been disclosed exemplary embodiments of the invention. Although specific terms are employed, they are used in a generic and descriptive sense only and not for purposes of limitation of the scope of the invention.

[0097] The foregoing description, for purposes of explanation, used specific nomenclature to provide a thorough understanding of the disclosure. However, it will be apparent to one skilled in the art that the specific details are not required in order to practice the systems and methods described herein. The foregoing descriptions of specific examples are presented for purposes of illustration and description. They are not intended to be exhaustive of or to limit this disclosure to the precise forms described. Obviously, many modifications and variations are possible in view of the above teachings. The examples are shown and described in order to best explain the principles of this disclosure and practical applications, to thereby enable others skilled in the art to best utilize this disclosure and various examples with various modifications as are suited to the particular use contemplated.

Claims

1. We Claim:

1. A smart policing system (150) comprising: a plurality of Internet Protocol (IP) cameras (100) equipped with edge computing capabilities, wherein the IP cameras (100) are installed at designated hotspots and powered; a power and connectivity panel (200) configured to supply power and internet connectivity to the IP cameras (100); a Power over Ethernet (PoE) Injector (300) to supply power and internet to the IP cameras (100) via Ethernet; a cloud-based Artificial Intelligence (Al) system (400) controlled by at least one processor (250) to receive video data from the IP cameras (100), wherein the Al system (400) includes: a video analytics module (410) for automatic detection and classification of traffic violations and crimes, digital cloud storage (420) for recording analyzed data in respective crime and traffic databases; a video query module (430) for processing natural language queries that are compared with the stored analyzed data, an Al model (440) to translate queries into attributes, and to identify Al models corresponding to the attributes; an alert mechanism (450) configured to issue notifications via emails, web, and mobile applications (500) based on identified events from the video analytics module (410); wherein the edge computing capabilities of the IP cameras (100) perform frame filtering using proprietary Al algorithms (470) to reduce data transmission to the cloud-based Al system (400).

2. The smart policing system (150) as claimed in claim 1, wherein the identification of the Al models (440) corresponding to the attributes are based, but not limited to, the colour of a vehicle, type or make of the vehicle, clothes of a person, and the appearance of the person.

3. The smart policing system (150) as claimed in claim 1, wherein the proprietary Al algorithms (470) employed for frame filtering at the edge comprises: a density plot of Peak Signal-to-Noise Ratio (PSNR) (310), a density plot of average difference values (320), a density plot of similarity scores (330), a density plot of mean square errors (340), and wherein frames correlating to reference frames based on these metrics are transmitted to the cloud-based Al system (400), and non-relevant frames are discarded.

4. The smart policing system (150) as claimed in claim 1, wherein the video analytics module (410) includes, but not limited to, automatic detection and classification of: number plate recognition, helmet-less riders, drivers not wearing seat belts, vehicles parked in no-parking zones, public brawls and street fighting, accidents, chain snatching, robbery, identification of vehicle types or colors, and identification of persons based on provided descriptions in natural language.

5. The smart policing system (150) as claimed in claim 1, wherein the cloud-based Al system (400) further comprises a reframe relevance checker (460) that sequentially processes frames to determine their relevance to traffic violations or crimes using a master algorithm.

6. The smart policing system (150) as claimed in claim 1, wherein the video query module (430) translates natural language queries into attribute-based searches within the crime or traffic databases, facilitating specific searches comprising: detection of persons with particular apparel, identification of vehicles with multiple riders, identification of instances of vehicles running red lights, and search for specific vehicles by color and registration details.

7. The smart policing system (150) as claimed in claim 1, wherein the IP cameras (100) are powered through one of grid electricity and a solar setup, and wherein the power and connectivity panel (200) that supplies power and internet connectivity to the IP cameras (100) comprises a Switch Mode Power Supply (SMPS) (210) to convert 220 VAC to lower VDC, and an Internet Router (220) for providing internet connectivity via one of SIM card, fiber, and broadband.

8. The smart policing system (150) as claimed in claim 1, wherein the power and connectivity panel (200) comprises an LED light (230) for illumination purposes, powered through the power and connectivity panel (200)9. The smart policing system (150) as claimed in claim 1, wherein information from the cloud-based Al system (400) is transmitted to the web and mobile applications (500) that provide functionalities comprising: user authentication and login, real-time dashboards and map views indicating camera locations, alert notifications for traffic violations and crimes, natural language video query, uploading of images for tracking suspects or vehicles, statistical reporting and graphical representations of crime and traffic data, and recent offense or violation tracking.

10. The smart policing system (150) as claimed in claim 1 is software-defined, allowing for the addition of new capabilities without requiring hardware modifications.

11. The smart policing system (150) as claimed in claim 1 generates and transmits alert notifications with detailed information including the location and images of detected traffic violations or crimes.

12. The smart policing system (150) as claimed in claim 6, wherein the attributebased searches are customized to detect: specific apparel combinations, the presence of multiple riders on a two-wheeler, vehicles violating traffic signals, and particular vehicle attributes including color, make, and registration details.

13. The smart policing system (150) as claimed in claim 1, wherein the cloud-based Al system (400) uses proprietary Al algorithms (470) in the video analytics module (410) and query translation to operate in real-time to facilitate prompt alerting and response, and wherein the video query module (430) is capable of understanding and processing both text and voice-based natural language queries.

14. The smart policing system (150) as claimed in claim 1, wherein the IP cameras (100) comprise an edge device that uses a proprietary filtering algorithm to ensure only relevant frames are transmitted to the cloud-based Al system (400), thereby optimizing bandwidth usage.

15. The smart policing system (150) as claimed in claim 1 is equipped to handle a variety of crimes and traffic violations, including but not limited to: public disturbances, theft, vehicle speed infractions, and non-compliance with safety regulations such as helmet or seat belt usage.

16. A method for smart policing, comprising:• installing a plurality of Internet Protocol (IP) cameras (100) equipped with edge computing capabilities at designated hotspots and powering them;• supplying power and internet connectivity to the IP cameras (100) via a power and connectivity panel (200);• providing power and internet to the IP cameras (100) through a Power over Ethernet (PoE) Injector (300);• receiving video data from the IP cameras (100) into a cloud-based Artificial Intelligence (Al) system (400) controlled by at least one processor (250), which includes: o automatically detecting and classifying traffic violations and crimes using a video analytics module (410); o recording analyzed data in respective crime and traffic databases using digital cloud storage (420); o processing natural language queries that are compared with the stored analyzed data using a video query module (430);o translating queries into attributes and identifying Al models corresponding to the attributes using an Al model (440); o issuing notifications via emails, web, and mobile applications (500) based on identified events from the video analytics module (410) using an alert mechanism (450); and• performing frame filtering using proprietary Al algorithms (470) at the edge computing capabilities of the IP cameras (100) to reduce data transmission to the cloud-based Al system (400).

17. The method for smart policing as claimed in claim 16, further comprising identifying Al models (440) corresponding to attributes based on, but not limited to, the color of a vehicle, type or make of the vehicle, clothes of a person, and the appearance of the person.

18. The method for smart policing as claimed in claim 16, wherein the proprietary Al algorithms (470) employed for frame filtering at the edge comprise:• generating a density plot of Peak Signal-to-Noise Ratio (PSNR) (310);• generating a density plot of average difference values (320);• generating a density plot of similarity scores (330);• generating a density plot of mean square errors (340);• transmitting frames correlating to reference frames based on these metrics to the cloud-based Al system (400) and discarding non-relevant frames.

19. The method for smart policing as claimed in claim 16, wherein the video analytics module (410) includes, but is not limited to, automatically detecting and classifying: number plate recognition, helmet-less riders, drivers not wearing seat belts, vehicles parked in no-parking zones, public brawls and street fighting, accidents, chain snatching, robbery, crimes related to women and children, identification of vehicle types or colors, and identification of persons based on provided descriptions in natural language.

20. The method for smart policing as claimed in claim 16, comprising processing frames sequentially to determine their relevance to traffic violations or crimes using a reframe relevance checker (460) and a master algorithm in the cloud-based Al system (400).

21. The method for smart policing as claimed in claim 16, wherein the video query module (430) translates natural language queries into attribute-based searches within the crime or traffic databases, facilitating specific searches comprising:• detecting persons with particular apparel;• identifying vehicles with multiple riders;• identifying instances of vehicles running red lights; and• searching for specific vehicles by color and registration details.

22. The method for smart policing as claimed in claim 16, comprising powering the IP cameras (100) through one of grid electricity and a solar setup, and supplying power and internet connectivity to the IP cameras (100) via a power and connectivity panel (200) that includes a Switch Mode Power Supply (SMPS) (210) to convert 220 VAC to lower VDC, and an Internet Router (220) for providing internet connectivity via one of SIM card, fiber, and broadband.

23. The method for smart policing as claimed in claim 16, comprising illuminating designated hotspots using an LED light (230) powered through the power and connectivity panel (200).

24. The method for smart policing as claimed in claim 16, comprising transmitting information from the cloud-based Al system (400) to web and mobile applications (500) that provide functionalities comprising: user authentication and login, real-time dashboards and map views indicating camera locations, alert notifications for traffic violations and crimes, natural language video queries, uploading images for tracking suspects or vehicles, statistical reporting and graphical representations of crime and traffic data, and recent offense or violation tracking.

25. The method for smart policing as claimed in claim 16, wherein the system is software-defined, allowing for the addition of new capabilities without requiring hardware modifications.

26. The method for smart policing as claimed in claim 16, comprising generating and transmitting alert notifications with detailed information including the location and images of detected traffic violations or crimes.

27. The method for smart policing as claimed in claim 21, wherein the attributebased searches are customized to detect: specific apparel combinations, the presence of multiple riders on a two-wheeler, vehicles violating traffic signals, and particular vehicle attributes including color, make, and registration details.

28. The method for smart policing as claimed in claim 16, wherein the cloud-based Al system (400) uses proprietary Al algorithms (470) in the video analytics module (410) and query translation to operate in real-time to facilitate prompt alerting and response, and wherein the video query module (430) is capable of understanding and processing both text and voice-based natural language queries.

29. The method for smart policing as claimed in claim 16, comprising using a proprietary filtering algorithm at the edge device within the IP cameras (100) to ensure only relevant frames are transmitted to the cloud-based Al system (400), thereby optimizing bandwidth usage.

30. The method for smart policing as claimed in claim 16, comprising handling a variety of crimes and traffic violations, including but not limited to: public disturbances, theft, vehicle speed infractions, and non-compliance with safety regulations such as helmet or seat belt usage.

Images