Intelligent security and protection method based on edge calculation

Abstract

The invention mainly discloses an intelligent security and protection method based on edge calculation. The method comprises the following steps: acquiring a video collected by a camera; detecting the video to obtain a detection result; the script code and the configuration file uploaded by the user are received and stored, when it is judged that an abnormal event exists in the detection result, the script is preprocessed, an intermediate code is generated, the intermediate code is analyzed and executed in the sandbox environment, an external device API is called to send a trigger command to an external device, and the trigger command is used for triggering the external device to make a response. And the flexibility, the expandability and the safety of the system are improved.

Description

Technical Field

[0001] The present invention relates to the field of electronic information technology, and in particular to an intelligent security method based on edge computing. Background Art

[0002] With the development of Internet of Things technology, smart security systems are increasingly used in homes, offices, warehouses and other places. Traditional security systems rely on central servers for video analysis and decision-making, which has problems such as high data transmission delay and limited processing capacity. In order to solve these problems, edge computing technology has been introduced into smart security systems, which improves the system's response speed and processing capacity by performing real-time data analysis and decision-making on edge node devices.

[0003] Although existing edge computing security systems can achieve basic video surveillance and alarm functions, they are still insufficient in flexibility and scalability. For example, when new equipment needs to be added or existing logic needs to be modified, the entire system code often needs to be rewritten and deployed, which is not only time-consuming and labor-intensive, but may also affect the stability and security of the system. In addition, traditional systems usually only support a single type of alarm, and cannot flexibly select the most appropriate alarm according to specific circumstances. Summary of the invention

[0004] To solve at least one of the aforementioned technical problems, the present disclosure proposes in a first aspect an intelligent security method based on edge computing, comprising: obtaining a video captured by a camera; detecting the video to obtain a detection result; receiving and storing script codes and configuration files uploaded by the user, and when it is determined that the detection result has an abnormal event, preprocessing the script to generate an intermediate code, parsing and executing the intermediate code in a sandbox environment, and calling an external device API to send a trigger command to the external device, the trigger command being used to trigger the external device to respond.

[0005] Preferably, it also includes: creating and uploading a configuration file, the configuration file includes the properties and trigger conditions of each external device; the virtual machine parses the execution module to load the configuration file, and parses the properties and trigger conditions of the external device; according to the detected abnormality type, location information, time conditions and trigger conditions, the most suitable external device is selected; the external device API sends a trigger command to the selected external device, triggering the selected external device to respond.

[0006] Preferably, when a new external device is connected, the script is modified according to the interface document of the new external device; the modified script is uploaded to the edge node device; the updated script is preprocessed to generate an intermediate code adapted to the new external device; the intermediate code adapted to the new external device is parsed and executed; and the new external device is triggered to respond.

[0007] Preferably, the method further includes: the external device is an alarm, and an alarm to be responded to is selected from a plurality of alarms according to the detection result and current environmental data.

[0008] Preferably, the alarm communicates with the edge node device using a standard communication protocol.

[0009] The present disclosure proposes in a second aspect an intelligent security system based on edge computing, an intelligent security system based on edge computing, characterized in that it includes: a camera, used to obtain video captured by the camera; an AI edge node, used to detect the video and obtain the detection result; a programmable central control module, used to receive and store script codes and configuration files uploaded by users, when it is judged that there is an abnormal event in the detection result, pre-process the script, generate intermediate code, parse and execute the intermediate code in a sandbox environment, call the external device API to send a trigger command to the external device, and the trigger command is used to trigger the external device to respond.

[0010] Preferably, the programmable central control module includes a source code module, an encoder module, and a virtual machine parsing and execution module. The source code module is used to receive and store script codes uploaded by users; the encoder module is used to preprocess the scripts and generate intermediate codes; and the virtual machine parsing and execution module is used to parse and execute the intermediate codes in a sandbox environment.

[0011] Preferably, the source code module is written in Python and supports script uploading and storage.

[0012] Preferably, it is characterized in that it also includes an alarm, which is used to sound an alarm when an abnormal situation is detected, and there is at least one alarm.

[0013] The present disclosure proposes a computer-readable medium in a third aspect, wherein a computer program is stored in the computer-readable medium. The computer program is loaded and executed by a processing module to implement the steps of any of the above-mentioned methods.

[0014] Some technical effects of the present disclosure are: an intelligent security method based on edge computing, including: obtaining video collected by a camera; detecting the video to obtain the detection result; receiving and storing the script code and configuration file uploaded by the user, and when it is judged that there is an abnormal event in the detection result, preprocessing the script, generating intermediate code, parsing and executing the intermediate code in a sandbox environment, calling the external device API to send a trigger command to the external device, and the trigger command is used to trigger the external device to respond. Improve the flexibility, scalability and security of the system. . BRIEF DESCRIPTION OF THE DRAWINGS

[0015] To better understand the technical solution of the present disclosure, you may refer to the following drawings for auxiliary explanation of the prior art or embodiments. These drawings will selectively display the products or methods involved in the prior art or some embodiments of the present disclosure. The basic information of these drawings is as follows: Figure 1 The present invention is a flowchart of an embodiment of an intelligent security method based on edge computing disclosed herein. DETAILED DESCRIPTION

[0016] The technical means or technical effects involved in the present disclosure will be further described below. Obviously, the embodiments (or implementation methods) provided are only some of the implementation methods covered by the present disclosure, but not all of them. Based on the embodiments in the present disclosure and the explicit or implicit descriptions of the figures and texts, all other embodiments that can be obtained by those skilled in the art without making creative efforts will be within the scope of protection requested by the present disclosure.

[0017] The present invention provides an intelligent security method based on edge computing, and the above method is applied to an intelligent security system based on edge computing, which includes a camera, an AI edge node, a source code module, an encoder module, and a virtual machine parsing execution module. When it is necessary to connect some different types of external devices to the system, the external device here can be an alarm, etc. The method provided by this application does not require rewriting and deploying the entire system code. Instead, through the collaborative work of these modules, flexible script programming, efficient device control, and dynamic selection of the most suitable alarm based on factors such as abnormality type, location information, and time conditions are realized, thereby improving the flexibility, scalability, and security of the system.

[0018] The camera of the system is used to collect real-time video and transmit the video stream to the AI ​​edge node through the RTSP (Real-Time Streaming Protocol) protocol. The AI ​​edge node is a high-performance computing device responsible for video analysis, anomaly detection and decision logic. It can run deep learning models to identify abnormal behaviors. The AI ​​edge node includes a programmable central control module, which includes a source code module, an encoder module, and a virtual machine parsing and execution module. The source code module is used to receive and store script codes uploaded by users; the encoder module is used to preprocess the scripts and generate intermediate codes; the virtual machine parsing and execution module is used to parse and execute intermediate codes in a sandbox environment. The source code module can store script codes and configuration files uploaded by users and supports Python or other programming languages. The encoder module preprocesses the scripts, including syntax checking, optimization, and compiling them into intermediate codes. The virtual machine parsing and execution module is responsible for parsing and executing intermediate codes in a safe sandbox environment to ensure the safe execution of the script. The alarm cluster contains various types of alarms (such as sound alarms, lighting controllers, SMS notifications, etc.), which can trigger different alarms according to specific needs. The source code module is written in Python and supports script uploading and storage.

[0019] AI edge nodes can efficiently process video data and run complex deep learning models. The hardware configuration of AI edge nodes includes processors: select high-performance embedded computing platforms, such as NVIDIA Jetson series (such as JetsonNano, Jetson Xavier NX), Intel NUC, etc. These devices have powerful GPU acceleration capabilities and are suitable for running image processing and machine learning tasks. Memory: Equipped with at least 4GB of RAM to ensure that the system can remain smooth during multitasking. Storage: Provide sufficient internal storage space (such as eMMC or SSD) for installing the operating system, saving video clips and detection logs. It is recommended to use a storage medium with a capacity of 64GB or larger. Network interface: Built-in Gigabit Ethernet port or Wi-Fi module, support fast and stable network connection, and facilitate receiving video streams and other control signals. Scalability: Consider possible future functional expansion needs, and choose devices with rich interfaces such as USB and HDMI to facilitate the connection of additional sensors or other external devices.

[0020] The software environment of the AI ​​edge node determines what types of tasks it can perform and its performance: Operating system: Usually a Linux distribution (such as Ubuntu) is used, which can provide good stability and security, and has extensive community support and technical resources. Development framework: Install necessary dependency libraries and development toolkits (SDKs), such as Python, TensorFlow, PyTorch, OpenCV, etc. These tools provide strong support for video decoding, image processing, and deep learning model deployment. Containerization technology: Using containerization technologies such as Docker can further improve the isolation and portability of the system and simplify the application deployment process, especially in a multi-device environment.

[0021] like Figure 1 The present invention provides an intelligent security method based on edge computing, comprising: S10: Obtaining the video collected by the camera; S20: Detect the video and obtain the detection result; S30: Receive and store the script code and configuration file uploaded by the user. When it is determined that there is an abnormal event in the detection result, pre-process the script, generate intermediate code, parse and execute the intermediate code in the sandbox environment, call the external device API to send a trigger command to the external device, and the trigger command is used to trigger the external device to respond.

[0022] The intelligent security system of the present invention adopts a distributed edge computing architecture to improve the response speed and processing capability of the system while ensuring the security and privacy of data.

[0023] S10: Obtaining the video collected by the camera; Use multiple cameras for monitoring, collect video streams in real time through the cameras, and transmit the video streams to the AI ​​edge node through the network (such as RTSP protocol). Obtain real-time video streams from cameras through protocols such as RTSP to ensure the stability and low latency of data transmission. Decode the received compressed video stream into a sequence of image frames for subsequent processing. This step may involve operations such as adjusting resolution, cropping the region of interest (ROI), and denoising to optimize the quality of data input to the AI ​​model.

[0024] S20: Detect the video and obtain the detection result; The AI ​​model on the AI ​​edge node performs real-time analysis of the received video stream to detect intruders or other anomalies. When performing anomaly detection and analysis, pre-trained deep learning models such as YOLOv5, SSD, FasterR-CNN, etc. are pre-loaded for tasks such as object detection and behavior recognition. The loaded model is applied to each frame of the image to perform object detection, classification, or behavior recognition tasks. The model outputs the location of the detection box, the category label, and the corresponding confidence score.

[0025] The AI ​​edge node then analyzes the video. The AI ​​edge node can use the AI ​​model to detect abnormal situations in the video. Abnormal situations include pedestrian intrusion, garbage dumping, and other abnormal events in various scenarios. Various conventional methods can be used to detect and judge abnormal events in the video, and there is no restriction again. When it is judged that there is an abnormal event in the video, the AI ​​edge node will call the script in the programmable central control module to decide the next step. When an external device needs to be connected, the script can be temporarily uploaded by the user. The script content can include operations such as triggering an alarm, turning on the on-site lights, and sending notifications. Combine multiple frames of information for time series analysis, such as tracking the trajectory of moving objects and judging specific behavior patterns (such as wandering, running). This method helps to reduce false alarms and improve detection accuracy. Set trigger conditions based on the detection results. For example, when an intruder is detected, if there are high-confidence targets in several consecutive frames, or the target appears in a specific area, it is considered that an intrusion event has occurred. Based on the final screened detection results, decide whether to trigger the alarm. For example, if the intrusion behavior is detected at night and at the entrance, select the sound alarm with the highest priority to sound the alarm. Send instructions to selected alarms through API or SDK to ensure that commands can be issued quickly and accurately. Record the timestamp, content and results of each detection to facilitate subsequent auditing and troubleshooting. Log data can also be used to analyze system performance indicators such as false alarm rate, missed alarm rate, etc.

[0026] The AI ​​edge node is one of the core components of the intelligent security system. It is located between the data source (such as the camera) and the cloud service. It is responsible for real-time processing and analysis of the video stream from the camera, performing anomaly detection, and triggering corresponding alarm measures based on the detection results.

[0027] S30: Receive and store the script code and configuration file uploaded by the user. When it is determined that there is an abnormal event in the detection result, pre-process the script, generate intermediate code, parse and execute the intermediate code in the sandbox environment, call the external device API to send a trigger command to the external device, and the trigger command is used to trigger the external device to respond.

[0028] When an AI edge node needs to connect a new external device or replace some already connected external devices, the user first needs to write a simple script to define the actions when an abnormal situation is detected. The user creates and uploads a configuration file to define the properties and trigger conditions for each newly added external device. The script and configuration file are uploaded to the AI ​​edge node device through the network interface. The source code module of the AI ​​edge node device receives the script and configuration file and stores them.

[0029] Create and upload a configuration file, which includes the properties and trigger conditions of each external device; the virtual machine parses the execution module to load the configuration file and parses the properties and trigger conditions of the external device; selects the most suitable external device based on the detected abnormality type, location information, time conditions and trigger conditions; the external device API sends a trigger command to the selected external device to trigger the selected external device to respond. When connecting a new external device, modify the script according to the interface document of the new external device; upload the modified script to the edge node device; preprocess the updated script to generate intermediate code adapted to the new external device; parse and execute the intermediate code adapted to the new external device; trigger the new external device to respond.

[0030] When the system is running, the encoder module obtains the script file from the source code module, performs syntax checking and optimization on the script, and generates intermediate code. The virtual machine parsing and execution module obtains the intermediate code from the encoder module, and parses and executes the intermediate code in the sandbox environment. The most suitable external device is selected based on the detection results and current environmental data. The API of the selected external device is called according to the script content to trigger the external device. When an abnormal situation is detected, the API of the selected external device is called separately to trigger an alarm or other response. At the same time, other devices (such as lighting controllers) may also be called to increase the safety of the site.

[0031] When the system is connected to multiple external devices, such as multiple alarms, the alarm can be selected dynamically. The configuration file is a configuration file created and uploaded by the user in JSON or YAML format, which defines the properties of each alarm (such as type, location, priority, etc.) and trigger conditions (such as abnormality type, time conditions, etc.). Then read the configuration file, the virtual machine parses the execution module to load the configuration file, and parses the alarm properties and trigger conditions. Then perform logical judgment and select the most suitable alarm based on the detected abnormality type, location information, time conditions and other factors. Generate a trigger command and send the trigger command to the selected alarm device through the alarm API to trigger the alarm sound or light.

[0032] Specifically, the camera: use IP camera and support RTSP protocol. AI edge node: use high-performance computing devices such as NVIDIA Jetson series or Intel NUC. Source code module: written in Python language, support script upload and storage. Encoder module: use Python's ast module for syntax checking and code optimization. Virtual machine parsing execution module: use Python's exec function to execute intermediate code in a sandbox environment. Alarm: use standard communication protocols (such as HTTP, TCP / IP) to communicate with edge node devices.

[0033] Dynamically select the specific process of alarms and collect information: Detection results: determine the type of abnormal situation detected (such as intrusion, fire, smoke, etc.), identify the specific location where the abnormality occurred, and the confidence score of the AI ​​model for the abnormality detection result. Current environmental data: including the current time period, weather conditions, on-site personnel status, etc. Load configuration files: Attribute definition: the attributes of each alarm, such as type (sound, light, SMS notification, etc.), location, priority, etc. Trigger conditions: define the conditions for triggering each alarm, including abnormality type, time range, specific location, etc. Parse configuration files: Load configuration: load predefined configuration files from source code modules. Parse conditions: parse the trigger conditions in the configuration files based on the detection results and current environmental data. Logical judgment: Matching conditions: check whether the detection results meet the trigger conditions of a certain alarm. Priority sorting: if multiple alarms meet the conditions, select the best option based on priority sorting. Execute response measures: Call API: once the most suitable alarm is selected, trigger the corresponding alarm or perform other predefined actions through API calls. Logging: record the response measures taken and their results for subsequent audits and improvements. The external device is an alarm, which selects the alarm to respond from multiple alarms based on the detection results and current environmental data. The alarm communicates with the edge node device using a standard communication protocol.

[0034] By introducing configuration files and logical judgment, the present invention realizes flexible script programming, efficient device control, and dynamic selection of the most suitable alarm according to factors such as abnormal type, location information, and time conditions, thereby improving the flexibility, scalability, and security of the system. Users can easily expand the functions of the system by writing simple scripts without modifying the original code. In addition, by performing real-time data analysis and decision-making on edge node devices, the response speed and processing power of the system are greatly improved.

[0035] The present invention discloses an intelligent security system based on edge computing and a dynamic alarm selection method thereof, including a camera, an AI edge node, a source code module, an encoder module, a virtual machine parsing execution module and a plurality of different types of alarms. Through the collaborative work of these modules, flexible script programming, efficient device control and dynamic selection of the most suitable alarm according to factors such as abnormal type, location information, and time conditions are realized, thereby improving the flexibility, scalability and security of the system. Users can easily expand the functions of the system by writing simple scripts without modifying the original code. In addition, by performing real-time data analysis and decision-making on edge node devices, the response speed and processing power of the system are greatly improved.

[0036] In the second aspect, the present disclosure proposes a device for controlling a display device based on gestures, which is characterized by comprising: an acquisition unit for acquiring the skeleton node information of the operator and the skeleton node information of the operating hand in the image; a matching unit for generating a feature vector describing the operator's hand posture according to the skeleton node information of the operating hand, matching the feature vector describing the operator's hand posture with the gesture model in the gesture model library, and dividing the image into a frame image with gesture category and a frame image without gesture category according to the matching result, and obtaining the gesture category information in the frame image with gesture category. A judgment unit for acquiring the gesture category information and the position information of the hand in the frame image without gesture category according to the skeleton node information of the operator and the gesture category information of the previous and next frames of the frame image without gesture category; a generation unit for generating instructions for controlling the display device according to the gesture category information and the position information of the hand in the frame image with gesture category and the frame image without gesture category.

[0037] The present disclosure proposes a computer-readable medium in a third aspect, wherein a computer program is stored in the computer-readable medium, and the computer program is loaded and executed by a processing module to implement the steps of the acquisition method. It can be understood by those skilled in the art that all or part of the steps in the embodiment can be implemented by instructing related hardware through a computer program, and the program can be stored in a computer-readable medium, and the readable medium can include various media that can store program codes, such as a flash drive, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk.

[0038] Within the scope of the knowledge and ability level of those skilled in the art, the various embodiments or technical features mentioned herein may be combined with each other as other optional embodiments without conflict. These limited number of optional embodiments, which are not listed one by one and are formed by combining a limited number of technical features, still fall within the technical scope disclosed in the present disclosure and can be understood or inferred by those skilled in the art in combination with the drawings and the above text.

[0039] In addition, the description of most embodiments is based on different focuses. If you need to further understand what is not described in detail, you can refer to the relevant content of the prior art, other relevant descriptions in this article or the purpose of the invention for reasonable reasoning.

[0040] It is emphasized again that the embodiments listed above are typical and preferred embodiments of the present disclosure, which are only used to explain and interpret the technical solutions of the present disclosure in detail to facilitate the understanding of the readers, and are not used to limit the scope or application of the present disclosure. Any technical solutions obtained by modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the scope of the present disclosure.

Claims

1. An intelligent security method based on edge computing, characterized in that: include: Get the video captured by the camera; Detect the video and obtain the detection result; Receive and store script codes and configuration files uploaded by users. When it is determined that there are abnormal events in the detection results, preprocess the script, generate intermediate code, parse and execute the intermediate code in the sandbox environment, and call the external device API to send a trigger command to the external device. The trigger command is used to trigger the external device to respond.

2. The method according to claim 1, characterized in that: Also includes: Create and upload a configuration file, which includes the properties of each external device and the trigger conditions; The virtual machine parses the execution module to load the configuration file and parses the properties and trigger conditions of the external device; Select the most suitable external device based on the detected anomaly type, location information, time conditions and trigger conditions; The external device API sends a trigger command to the selected external device, triggering the selected external device to respond.

3. The method according to claim 1, characterized in that When a new external device is connected, modify the script according to the interface documentation of the new external device; Upload the modified script to the edge node device; Preprocess the updated script to generate intermediate code adapted to the new external device; Parse and execute intermediate codes adapted to new external devices; Trigger a new external device to respond.

4. The method according to claim 1, characterized in that: Also includes: The external device is an alarm, and an alarm to be responded to is selected from multiple alarms based on the detection results and current environmental data.

5. The method according to claim 4, characterized in that The alarm communicates with the edge node device using standard communication protocols.

6. An intelligent security system based on edge computing, characterized in that: include: Camera, used to obtain the video collected by the camera; AI edge node, used to detect videos and obtain detection results; The programmable central control module is used to receive and store script codes and configuration files uploaded by users. When it is judged that there are abnormal events in the detection results, the script is preprocessed to generate intermediate code, which is parsed and executed in the sandbox environment. The external device API is called to send a trigger command to the external device. The trigger command is used to trigger the external device to respond.

7. The system according to claim 6, characterized in that The programmable central control module includes a source code module, an encoder module, and a virtual machine parsing and execution module. The source code module is used to receive and store script codes uploaded by users; the encoder module is used to pre-process the scripts and generate intermediate codes; The virtual machine parsing and execution module is used to parse and execute the intermediate code in the sandbox environment.

8. The system according to claim 6, characterized in that The source code module is written in Python and supports script uploading and storage.

9. The system according to claim 6, characterized in that It also includes an alarm, which is used to sound an alarm when an abnormal situation is detected, and there is at least one alarm.

10. A computer readable medium, characterized in that: The computer readable medium stores a computer program, which is loaded and executed by the processing module to implement the steps of any one of the methods described in claims 1 to 5.

Images