An active alarm information generation method and system based on AI video analysis

Abstract

The present invention discloses an active warning information generation method and system based on AI video analysis, belonging to the technical field of video surveillance. Specifically, it includes: creating a fireworks detection task, selecting a corresponding video analysis algorithm and drawing an identification area; adjusting the depth of field of the camera lens to a shallow depth of field to blur the identification area and marking the blurred light spots of known light sources; when unmarked blurred light spots appear in the identification area, identifying the color temperature value and the maximum diameter of the light spot of the blurred light spot, marking the blurred light spot whose color temperature value and light spot meet the set conditions as a pending light spot, and marking the minimum circumscribed rectangle area of the pending light spot as the target area; adjusting the depth of field of the camera lens to a deep depth of field, collecting an image frame again, cropping the image block corresponding to the target area, inputting the image block into a preset image recognition model, and outputting the recognition result of the image block. When the recognition result is fireworks, trigger an alarm push, otherwise continue to monitor.

Description

Technical Field

[0001] The present invention relates to the technical field of video surveillance, and particularly relates to an active alarm information generation method and system based on AI video analysis. Background Art

[0002] With the rapid development of artificial intelligence technology, especially the continuous improvement of deep learning methods and performance, technologies such as computer vision, image processing, video structuring, and data analysis are also constantly improving. These technological advancements provide a solid technical foundation for the application of AI video analysis. In practical applications, traditional video surveillance systems often face problems such as "difficult extraction and search", unreliable target feature retrieval, large errors in image search and comparison, and lack of effective methods for in-depth mining of video image information. These problems limit the application value of video surveillance systems, while AI video analysis can solve these problems and improve the efficiency and accuracy of video surveillance.

[0003] During the process of AI video analysis, the system needs to extract useful feature information from the entire photo for subsequent classification or recognition tasks. However, the number of features contained in the entire photo is large and complex, which leads to a large computing power requirement for image recognition. Moreover, the entire photo may contain a large amount of irrelevant information, which will interfere with the recognition result and thus reduce the recognition accuracy. At the same time, performing image recognition on the entire photo will also increase the computational complexity and time cost. Summary of the Invention

[0004] The purpose of the present invention is to provide an active alarm information generation method and system based on AI video analysis to solve the following technical problems:

[0005] The number of features contained in the entire photo is large and complex, which leads to a large computing power requirement for image recognition. Moreover, the entire photo may contain a large amount of irrelevant information, which will interfere with the recognition result and thus reduce the recognition accuracy.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] An active alarm information generation method based on AI video analysis includes the following steps:

[0008] Select one from the camera list as the monitoring object, create a fireworks detection task, set the execution time of the task, select the corresponding video analysis algorithm and draw the ROI recognition area in the algorithm list; set the corresponding video frame extraction frequency;

[0009] Adjust the depth of field of the camera lens to a shallow depth of field to blur the ROI recognition area, mark the blurred light spots of the known light sources in the ROI recognition area, and perform real-time light spot detection on the ROI recognition area according to the video frame extraction frequency;

[0010] When an unmarked blurred light spot appears in the ROI recognition area, identify the color temperature value and the maximum diameter of the light spot of the blurred light spot, mark the blurred light spot with the color temperature value within the set range and the maximum diameter of the light spot less than the set threshold as a pending light spot, and mark the minimum circumscribed rectangle area of the pending light spot as the target area;

[0011] Adjust the depth of field of the camera lens to a deep depth of field, collect the image frame of the ROI recognition area again, crop the image block corresponding to the target area, input the image block into a preset image recognition model, output the recognition result of the image block, and when the recognition result is fireworks, trigger an alarm push, otherwise continue to perform light spot detection on the ROI recognition area.

[0012] As a further solution of the present invention: The process of adjusting the depth of field of the camera lens to a shallow depth of field is as follows:

[0013] Adjust the distance between the focusing plane of the camera and the camera lens so that the distance is within the interval [A, A + a], where A is the nearest focusing distance of the camera and a is the set redundancy value.

[0014] As a further solution of the present invention: The process of adjusting the depth of field of the camera lens to a deep depth of field is as follows:

[0015] Adjust the aperture value of the camera lens so that the entire ROI recognition area is within the imaging depth of field of the camera lens.

[0016] As a further solution of the present invention: The process of performing real-time light spot detection on the ROI recognition area is as follows:

[0017] Set a selection box according to the shape of the lens aperture blades, set the deformation range of the selection box, perform edge detection on the image frame of the ROI recognition area, and perform binarization processing to obtain a binary image. Exclude the marked blurred light spots from the binary image, select the contours of the connected regions with a gray value of 255 in the binary region based on the selection box, filter out the contours within the deformation range of the selection box, and mark them as light spot contours. The light spot contours and the internal area are the blurred light spots.

[0018] As a further solution of the present invention: The process of identifying the color temperature value of the blurred light spot is as follows:

[0019] For any blurred light spot, when the color of the blurred light spot is white, the color temperature of the blurred light spot is the white balance parameter of the current camera;

[0020] When the color of the blurred light spot is light blue or blue, gradually increase the white balance parameter of the camera until the color of the blurred light spot becomes white. Then, the color temperature of the blurred light spot is the current white balance parameter of the camera;

[0021] When the color of the blurred light spot is yellow or orange, gradually decrease the white balance parameter of the camera until the color of the blurred light spot becomes white. Then, the color temperature of the blurred light spot is the current white balance parameter of the camera.

[0022] As a further solution of the present invention: The process of identifying the maximum diameter of the blurred light spot is as follows:

[0023] For any blurred light spot, select a number of points on the contour line of the blurred light spot at equal intervals. Taking each point as the starting point, draw a connecting line from the starting point through the centroid of the contour to the other side of the contour. Select the connecting line with the longest length and mark it as the maximum diameter of the blurred light spot.

[0024] As a further solution of the present invention: The specific process of marking the blurred light spot with a color temperature value within a set range and a maximum light spot diameter less than a set threshold as a pending light spot is as follows:

[0025] Obtain the flame color temperature K of the fireworks to be detected, and set an error threshold k. Then, the set range is [K - k, K + k];

[0026] Place the combustion material of the fireworks to be detected at the farthest distance in the recognition area. Keep the camera lens in a shallow depth of field, and record the size of the blurred light spot of the fireworks combustion material in the camera screen at this time. Use this size as the set threshold.

[0027] The present invention further includes an active warning information generation system based on AI video analysis for implementing the above-mentioned active warning information generation method based on AI video analysis, including:

[0028] A task configuration module, used to create a fireworks detection task, set the execution time of the task, select the corresponding video analysis algorithm and draw the ROI recognition area in the algorithm list; set the corresponding video frame extraction frequency;

[0029] A depth of field adjustment module, used to adjust the depth of field of the camera lens to a shallow depth of field or a deep depth of field;

[0030] A light spot monitoring module, used to mark the blurred light spots of known light sources in the ROI recognition area, and perform real-time light spot detection on the ROI recognition area according to the video frame extraction frequency; when an unmarked blurred light spot appears in the ROI recognition area, identify the color temperature value and the maximum diameter of the blurred light spot, mark the blurred light spot with a color temperature value within the set range and a maximum light spot diameter less than the set threshold as a pending light spot, and mark the minimum circumscribed rectangle area of the pending light spot as the target area;

[0031] An image recognition module, which is used to collect image frames in the ROI recognition area, crop the image block corresponding to the target area, input the image block into a preset image recognition model, and output the recognition result of the image block.

[0032] An alarm prompt module, which is used to trigger alarm push when the recognition result is fireworks, otherwise continue to perform spot detection on the ROI recognition area.

[0033] Advantages of the present invention:

[0034] The present invention first identifies the blurred spot of fireworks and crops the target area, and then performs image recognition, which significantly reduces the amount of data to be processed, thereby reducing the computing power requirement of the system. This method avoids complex image recognition processing of the entire photo, enabling more effective utilization of computing resources. By accurately positioning the target area and identifying the blurred spot, the interference of irrelevant information can be effectively excluded, thereby improving the accuracy of subsequent image recognition. Especially when detecting specific targets such as fireworks, this targeted processing method can significantly enhance the reliability of the recognition result. After cropping the target area, only image recognition is performed on this smaller area, greatly reducing the computational complexity and time cost. This not only improves the processing efficiency of the system but also enables the system to complete the recognition task in a shorter time, meeting the requirements of real-time monitoring. Due to the reduction of computational complexity and the improvement of processing speed, the method of the present invention can generate alarm information faster, enhancing the real-time response ability of the system. This is crucial for timely discovering potential safety hazards and taking preventive measures in advance. Description of the Drawings

[0035] The present invention will be further described below with reference to the accompanying drawings.

[0036] Figure 1 is a flow diagram of the present invention. Detailed Embodiments

[0037] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0038] Please refer to Figure 1 As shown, the present invention is an active alarm information generation method based on AI video analysis, including the following steps:

[0039] 1. Select one or more cameras from the camera list as the objects for monitoring and analysis. Create a task and fill in the task name: Smoke detection task, which is for smoking detection. After clicking the "Add Task" button, enter the task creation page. Next, select a suitable video analysis service node according to the requirements. This node will be responsible for processing and analyzing the video data from this camera. Set the execution time of the task, frequency plan parameter configuration, and select the task plan according to the actual needs. In the algorithm list, select the corresponding video analysis algorithm and draw the ROI recognition area. Set a reasonable video frame extraction frequency according to the analysis requirements. This helps to reduce resource consumption while ensuring the accuracy of analysis. After completing all the above settings, click the "Save" button to successfully create the task.

[0040] Task start / stop: Through the query and filtering functions of the task list, you can quickly find the task you want to manage. In the task list, select one or more tasks you want to start or stop. Click the "Start" or "Stop" button to perform the corresponding operation on the selected tasks.

[0041] Batch configuration modification: For convenient management, the task scheduling center also provides the function of batch configuration modification. Similarly, through the query and filtering functions of the task list, find the tasks for which you want to modify the configuration. Select one or more tasks that need to modify the configuration. Click the "Batch Configuration Modification" button to enter the configuration modification page. On this page, you can batch modify parameters such as the default frame extraction rate and default recognition delay of the selected tasks.

[0042] Batch configuration of plans: In addition, the task scheduling center also supports the function of batch configuration of plans. Through the query and filtering functions of the task list, find the tasks for which you want to set the analysis time plan. Select one or more tasks that need to set the analysis time plan. Click the "Batch Configuration of Plans" button to enter the plan setting page. On this page, you can batch set the analysis time plan for the selected tasks to ensure that they can perform video algorithm analysis work according to the scheduled time.

[0043] 2. Adjust the depth of field of the camera lens to a shallow depth of field. Adjusting the depth of field of the camera lens to a shallow depth of field is a crucial step. Specifically, this involves adjusting the distance between the focusing plane of the camera and the camera lens so that this distance is within a specific interval [A, A + a]. Here, A represents the nearest focusing distance of the camera, and a is a preset redundancy value used to ensure the accuracy and flexibility of the adjustment. Through such adjustment, the images in the ROI (region of interest) recognition area will be slightly blurred, which helps the subsequent spot detection process.

[0044] In the ROI recognition area, the system will first mark the blurred light spots generated by known light sources. These light spots may be caused by lights, reflections, or other light sources in the monitoring environment. The purpose of marking these blurred light spots is to exclude them during the subsequent light spot detection process to avoid interference with the detection results.

[0045] Next, the system will perform real-time light spot detection on the ROI recognition area according to the video frame extraction frequency. The video frame extraction frequency refers to the frequency of extracting frames from the video stream for detection, which determines the real-time performance and accuracy of the system. By setting an appropriate frame extraction frequency, the system can ensure that important information is not missed while not causing performance degradation due to processing excessive data.

[0046] During light spot detection, the system will set a selection box according to the shape of the lens aperture blades and set the deformation range of this selection box. This selection box is used for edge detection on the image frame of the ROI recognition area and obtains a binary image through binarization processing. During this process, the system will exclude the already marked blurred light spots from the binary image to avoid their influence on subsequent processing.

[0047] Then, the system will select the contours of the connected regions with a gray value of 255 in the binary region. These connected regions usually correspond to the bright parts in the image, such as light spots or reflective points. The system will filter out the contours located within the deformation range of the selection box and mark these contours as light spot contours. Finally, the light spot contours and the regions inside them are determined as blurred light spots.

[0048] 3. When unmarked blurred light spots appear in the ROI recognition area, a series of steps are taken to identify the color temperature value and the maximum diameter of these blurred light spots.

[0049] For the process of identifying the color temperature value of the blurred light spot, the system will make a judgment based on the color of the blurred light spot. If the color of the blurred light spot is white, then the color temperature of this blurred light spot is the current white balance parameter of the camera. This is because white usually corresponds to a neutral color temperature, that is, there is no obvious color cast.

[0050] If the color of the blurred light spot is light blue or blue, the system will gradually increase the white balance parameter of the camera until the color of the blurred light spot becomes white. During this process, the system will record the white balance parameter that makes the color of the blurred light spot become white and use it as the color temperature value of this blurred light spot.

[0051] On the contrary, if the color of the blurred light spot is yellow or orange, the system will gradually decrease the white balance parameter of the camera until the color of the blurred light spot becomes white. Similarly, the system will record the white balance parameter that makes the color of the blurred light spot become white and use it as the color temperature value of this blurred light spot.

[0052] Next is the process of identifying the maximum diameter of the blurred light spot. For any blurred light spot, the system will evenly select several points on its contour line. Then, starting from each point, a connecting line from the starting point to the other side of the contour passing through the centroid of the contour is drawn. Among these connecting lines, the system will select the longest one as the maximum diameter of the blurred light spot.

[0053] For the identified blurred light spots, the system will further process them to determine whether they are potential firework targets. Specifically, the system will mark the blurred light spots with color temperature values within a set range and the maximum diameter of the light spot less than a set threshold as pending light spots.

[0054] To set this range and threshold, the system first needs to obtain the flame color temperature K of the firework to be detected. This color temperature value is usually pre-determined according to the type and combustion conditions of the firework. Then, the system will set an error threshold k to define the acceptable range of the color temperature value. In this way, the set range is determined as [K - k, K + k]. Only when the color temperature value of the blurred light spot falls within this range will it be considered a possible firework target.

[0055] Next, the system will set up the firework combustibles to be detected. It will place the firework combustibles at the farthest distance in the recognition area and keep the camera lens in a shallow depth of field state. In this way, when the firework burns, the size of the blurred light spot in the camera image can be accurately recorded. This size will be used as the set threshold for subsequent judgment of whether the blurred light spot is a pending light spot.

[0056] Once the set range and set threshold are determined, the system will process the pending light spots. It will calculate the minimum bounding rectangle area of each pending light spot and mark this area as the target area. This target area will be used in the subsequent image recognition and alarm information generation processes. In this way, the system can more accurately locate potential firework targets, thereby improving the accuracy and reliability of recognition.

[0057] 4. When the system identifies potential firework targets, it will further adjust the depth of field of the camera lens to a deep depth of field to ensure that the entire ROI recognition area is within the imaging depth of field of the camera lens. This step is crucial for improving the accuracy of image recognition because only when the entire ROI area is clearly visible can the image recognition model accurately analyze and judge.

[0058] To achieve the deep depth of field effect, the system will adjust the aperture value of the camera lens. The size of the aperture value directly affects the depth of field range of the camera. A smaller aperture value can increase the depth of field, making more scene elements stay within the clear focus range. By precisely controlling the aperture value, the system can ensure that all details within the ROI area are clearly captured.

[0059] After adjusting the depth of field of the camera lens, the system will collect the image frames of the ROI recognition area again. This time, due to the use of a deep depth of field setting, the obtained images will be clearer and more detailed. Next, the system will process the collected images, especially for the previously marked target areas. The system will crop out the image blocks corresponding to the target areas, and this image block contains information about the possible firework targets and their surrounding environments.

[0060] Subsequently, the system will input this image block into a preset image recognition model. This model is specially trained to recognize various types of firework features. Deep learning models, such as convolutional neural networks, long short-term memory networks, and Transformer models, are introduced for feature extraction and classification recognition of video frames. By training these models, they can extract key features from video frames and achieve accurate detection and recognition of targets. The model will analyze the input image block and output a recognition result. If the recognition result is a firework, then the system will immediately trigger the alarm push mechanism and send alarm messages to relevant personnel, including:

[0061] Alarm monitoring: By analyzing various metrics, states, and data of the structured data generated by algorithms, once abnormal or unexpected behaviors are found, the alarm push will be immediately triggered.

[0062] Alarm notification: After the alarm is triggered, according to the preset alarm strategy, notifications will be sent to specified users through various means, such as SMS, email, instant messaging, etc.

[0063] Alarm handling: Handle the alarms generated by algorithm analysis, including confirming alarms, ignoring alarms, resolving alarms, and setting corresponding dispositions.

[0064] Alarm configuration strategy: Allow users to customize alarm strategies and configurations, including alarm templates, alarm levels, and alarm strategy settings. Custom alarm templates can be set for sending alarm notifications. Custom alarm templates can be set for sending alarm notifications. Custom alarm templates can be set for sending alarm notifications.

[0065] If the recognition result is not a firework, then the system will continue to perform spot detection on the ROI recognition area. This means that the system does not abandon the monitoring of potential firework targets but continuously conducts detection and analysis to ensure that no possible safety hazards are missed. This continuous monitoring method helps improve the reliability and accuracy of the system and provides more secure protection for users.

[0066] The present invention further includes an active warning information generation system based on AI video analysis for implementing the above-mentioned active warning information generation method based on AI video analysis, including:

[0067] A task configuration module, which is used to create a fire detection task, set the execution time of the task, select the corresponding video analysis algorithm and draw the ROI recognition area in the algorithm list; set the corresponding video frame extraction frequency;

[0068] A depth-of-field adjustment module, which is used to adjust the depth of field of the camera lens to a shallow depth of field or a deep depth of field;

[0069] A light spot monitoring module, which is used to mark the blurred light spots of known light sources in the ROI recognition area and perform real-time light spot detection on the ROI recognition area according to the video frame extraction frequency; when an unmarked blurred light spot appears in the ROI recognition area, identify the color temperature value and the maximum diameter of the light spot of the blurred light spot, mark the blurred light spot with the color temperature value within the set range and the maximum diameter of the light spot less than the set threshold as a pending light spot, and mark the minimum circumscribed rectangle area of the pending light spot as the target area;

[0070] An image recognition module, which is used to collect the image frames of the ROI recognition area, crop the image block corresponding to the target area, input the image block into a preset image recognition model, and output the recognition result of the image block.

[0071] An alarm prompt module, which is used to trigger alarm push when the recognition result is a fire, otherwise continue to perform light spot detection on the ROI recognition area.

[0072] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that includes one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, or magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0073] Those of ordinary skill in the art will realize that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0074] As described above, the above are only specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0075] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. An active alarm information generation method based on AI video analysis, characterized in that, The steps include: Select one camera from the camera list as the monitoring object, create a fire detection task, set the execution time of the task, select the corresponding video analysis algorithm in the algorithm list and draw the ROI recognition area; set the corresponding video frame extraction frequency; Adjust the depth of field of the camera lens to a shallow depth of field to blur the ROI recognition area, mark the blurred light spots of the known light sources in the ROI recognition area, and perform real-time light spot detection on the ROI recognition area according to the video frame extraction frequency; When an unmarked blurred light spot appears in the ROI recognition area, identify the color temperature value and the maximum diameter of the light spot of the blurred light spot, mark the blurred light spot with the color temperature value within the set range and the maximum diameter of the light spot less than the set threshold as a pending light spot, and mark the minimum circumscribed rectangle area of the pending light spot as the target area; Adjust the depth of field of the camera lens to a deep depth of field, collect the image frame of the ROI recognition area again, crop the image block corresponding to the target area, input the image block into a preset image recognition model, output the recognition result of the image block, and when the recognition result is fire, trigger an alarm push, otherwise continue to perform light spot detection on the ROI recognition area; The process of performing real-time light spot detection on the ROI recognition area is: Set a selection box according to the shape of the lens aperture blades, set the deformation range of the selection box, perform edge detection on the image frame of the ROI recognition area and perform binaryzation processing to obtain a binary image, exclude the marked blurred light spots from the binary image, select the contours of the connected areas with a gray value of 255 in the binary area based on the selection box, filter out the contours located within the deformation range of the selection box, and mark them as light spot contours. The light spot contours and the internal area are the blurred light spots.

2. The active warning information generation method based on AI video analysis according to claim 1, characterized in that, The process of adjusting the depth of field of the camera lens to a shallow depth of field is: Adjust the distance between the focusing plane of the camera and the camera lens so that the distance is within the interval [A, A + a], where A is the nearest focusing distance of the camera and a is the set redundancy value.

3. The active alarm information generation method based on AI video analysis according to claim 1, wherein The process of adjusting the depth of field of the camera lens to a deep depth of field is: Adjust the aperture value of the camera lens so that the entire ROI recognition area is within the imaging depth of field of the camera lens.

4. The active warning information generation method based on AI video analysis according to claim 1, wherein The process of identifying the color temperature value of the blurred light spot is: For any blurred light spot, when the color of the blurred light spot is white, the color temperature of the blurred light spot is the current white balance parameter of the camera; When the color of the blurred light spot is light blue or blue, gradually increase the white balance parameter of the camera until the color of the blurred light spot becomes white, then the color temperature of the blurred light spot is the current white balance parameter of the camera; When the color of the blurred light spot is yellow or orange, gradually decrease the white balance parameter of the camera until the color of the blurred light spot becomes white, then the color temperature of the blurred light spot is the current white balance parameter of the camera.

5. The active warning information generation method based on AI video analysis according to claim 1, wherein The process of identifying the maximum diameter of the blurred light spot is: For any blurred light spot, select several points on the contour line of the blurred light spot at equal intervals. Starting from each point, draw a connection line from the starting point to the other side of the contour through the centroid of the contour, and select the connection line with the longest length and mark it as the maximum diameter of the blurred light spot.

6. The active warning information generation method based on AI video analysis according to claim 1, characterized in that The specific process of marking a blurred light spot with a color temperature value within a set range and a maximum light spot diameter smaller than a set threshold as a pending light spot is as follows: Obtain the flame color temperature K of the fireworks to be detected, and set an error threshold k. Then the set range is [K - k, K + k]; Set the combustion material of the fireworks to be detected at the farthest distance in the recognition area. Keep the camera lens in a shallow depth of field, record the size of the blurred light spot of the fireworks combustion material in the camera screen at this time, and use this size as the set threshold.

7. An active warning information generation system based on AI video analysis, which is used to implement an active warning information generation method based on AI video analysis according to any one of claims 1-6, and is characterized in that, It includes: A task configuration module, used to create a fireworks detection task, set the execution time of the task, select the corresponding video analysis algorithm in the algorithm list, and draw the ROI recognition area; Set the corresponding video frame extraction frequency; A depth of field adjustment module, used to adjust the depth of field of the camera lens to a shallow depth of field or a deep depth of field; A light spot monitoring module, used to mark the blurred light spots of known light sources in the ROI recognition area, and perform real-time light spot detection on the ROI recognition area according to the video frame extraction frequency; when an unmarked blurred light spot appears in the ROI recognition area, identify the color temperature value and the maximum light spot diameter of the blurred light spot, mark the blurred light spot with a color temperature value within the set range and a maximum light spot diameter smaller than the set threshold as a pending light spot, and mark the minimum circumscribed rectangle area of the pending light spot as the target area; An image recognition module, used to collect the image frames of the ROI recognition area, crop the image block corresponding to the target area, input the image block into a preset image recognition model, and output the recognition result of the image block; An alarm prompt module, used to trigger an alarm push when the recognition result is fireworks, otherwise continue to perform light spot detection on the ROI recognition area.

Images