Fire accident video analysis method, device, equipment, storage medium and product

Abstract

This application relates to the field of fire accident analysis technology, and in particular to a method, apparatus, equipment, storage medium, and product for fire accident video analysis. The method extracts fire feature data and personnel / vehicle data from fire scene video data; then, it classifies and identifies the fire feature data to obtain flame information, smoke information, and light and shadow information; it performs super-resolution reconstruction and recognition of personnel behavior images from the personnel / vehicle data to obtain personnel feature information; it performs de-motion blurring processing on vehicle images from the personnel / vehicle data to obtain license plate feature information; finally, based on the above feature information, it generates a fire accident video analysis report. This method integrates multiple video analysis technologies to automatically extract fire feature data from fire scene video data, effectively improving the efficiency and accuracy of data extraction and recognition.

Description

Technical Field

[0001] This application relates to the field of fire accident analysis technology, and in particular to a fire accident video analysis method, apparatus, equipment, storage medium and product. Background Technology

[0002] With the acceleration of urbanization and the increase in building density, the frequency and complexity of fire accidents are constantly rising. Rapid response and accurate analysis of fire accidents have become urgent problems for fire departments and related units. Traditional fire accident analysis mainly relies on manual review of surveillance videos to replay and assess the situation on-site. This method is cumbersome and inefficient, consuming significant human resources and easily affected by human factors, leading to incomplete information extraction or inaccurate judgments. Furthermore, the complex environment at fire scenes, such as changes in lighting, smoke, and the rapid movement of people and vehicles, further increases the difficulty and error rate of manual analysis. In recent years, with the rapid development of artificial intelligence and computer vision technologies, automated analysis methods based on video data have gradually become a research hotspot. Deep learning models have significantly improved performance in image and video recognition, automatically extracting various feature information from fire scene videos, such as flames, smoke, changes in light and shadow, faces, and license plates, and performing efficient classification and recognition. However, most existing fire video analysis technologies only detect single types of feature information, lacking comprehensive analysis of multiple feature data, and cannot generate comprehensive and accurate fire accident analysis reports, resulting in poor performance of fire scene video analysis. Therefore, how to improve the effectiveness of fire video analysis has become an urgent technical problem to be solved. Summary of the Invention

[0003] The main objective of this application is to provide a method, apparatus, equipment, storage medium, and product for fire accident video analysis, aiming to solve the technical problem of how to improve the effectiveness of fire video analysis.

[0004] To achieve the above objectives, this application provides a method for analyzing fire accident videos, the method comprising the following steps:

[0005] Extract fire feature data and personnel and vehicle data from the fire scene video data;

[0006] The fire feature data is classified and identified to obtain flame information, smoke information, and light and shadow information.

[0007] Super-resolution reconstruction and recognition are performed on the images of human behavior in the personnel and vehicle data to obtain human feature information;

[0008] The vehicle images in the personnel and vehicle data are subjected to motion blur removal processing to obtain vehicle feature information;

[0009] Based on the flame information, smoke information, light and shadow information, personnel characteristic information, and vehicle characteristic information, a fire accident video analysis report is generated.

[0010] In one embodiment, before the step of extracting fire feature data and personnel and vehicle data from the fire scene video data, the method further includes:

[0011] Based on the topology of local storage for on-site monitoring, identify each video input channel;

[0012] The target video input channel is determined based on the device type, resolution, frame rate, and video encoding format corresponding to each video input channel.

[0013] Based on the target video input channel, acquire the target surveillance video;

[0014] The target surveillance video is optimized to obtain the fire scene video data.

[0015] In one embodiment, the step of optimizing the target surveillance video to obtain the fire scene video data includes:

[0016] Convert the target surveillance video to a preset standard format;

[0017] Based on a preset data recovery method, the lost frames of the target surveillance video after format conversion are recovered to obtain multiple restored surveillance videos.

[0018] Image enhancement processing is performed on the multiple restored surveillance videos to obtain multiple enhanced videos;

[0019] The timestamps of the multiple enhanced videos are synchronized.

[0020] The enhanced video after timestamp synchronization is subjected to integrity verification, and the video that passes the integrity verification is used as the fire scene video data.

[0021] In one embodiment, the step of extracting fire feature data and personnel and vehicle data from the fire scene video data includes:

[0022] The fire scene video is preprocessed to obtain a target frame set, wherein the preprocessing includes one or more of video frame decomposition, image enhancement, and region of interest detection;

[0023] Based on a preset fire feature detection model, flame detection, smoke detection, and light and shadow detection are performed on each frame in the target frame set to obtain the fire feature data.

[0024] Based on a preset personnel and vehicle detection model, personnel and vehicle detection are performed on each frame in the target frame set to obtain the personnel and vehicle data.

[0025] In one embodiment, the step of performing super-resolution reconstruction and recognition of personnel behavior images in the personnel and vehicle data to obtain personnel feature information includes:

[0026] Extract human behavior images from the personnel and vehicle data, and evaluate the quality of the extracted human behavior images;

[0027] Remove face images whose quality evaluation value is lower than the preset face image quality standard evaluation value to obtain quality face images;

[0028] The size, color space, brightness, and contrast of the quality face image are standardized to obtain a standard quality face image;

[0029] Based on a preset super-resolution reconstruction model, the standard quality face image is reconstructed and enhanced to obtain a reconstructed face image;

[0030] Based on a preset face recognition algorithm, feature extraction is performed on the reconstructed face image to obtain the person's feature information.

[0031] In one embodiment, the step of performing motion blur removal processing on the vehicle images in the personnel and vehicle data to obtain vehicle feature information includes:

[0032] The vehicle image is extracted from the personnel and vehicle data, and the quality evaluation is performed on the extracted vehicle image;

[0033] Vehicle images whose quality evaluation value is lower than the preset vehicle image quality standard evaluation value are removed to obtain quality vehicle images;

[0034] The standardization process is applied to the size, color space, brightness, and contrast of the quality vehicle image to obtain a standard quality vehicle image.

[0035] The standard quality vehicle image is divided into multiple license plate images, and the fuzzy information of each license plate image is determined.

[0036] Based on the fuzzy information, the images of each license plate are input into a preset deblurring model to obtain the vehicle feature information.

[0037] Furthermore, to achieve the above objectives, this application also proposes a fire accident video analysis device, which includes:

[0038] The data extraction module is used to extract fire feature data and personnel and vehicle data from the fire scene video data;

[0039] The information recognition module is used to classify and recognize the fire feature data to obtain flame information, smoke information and light and shadow information;

[0040] The personnel identification module is used to perform super-resolution reconstruction and identification of personnel behavior images in the personnel and vehicle data to obtain personnel feature information;

[0041] The vehicle recognition module is used to perform motion blur removal processing on the vehicle images in the personnel and vehicle data to obtain vehicle feature information;

[0042] The target module is used to generate a fire accident video analysis report based on the flame information, the smoke information, the light and shadow information, the personnel feature information, and the vehicle feature information.

[0043] In addition, to achieve the above objectives, this application also proposes a fire accident video analysis device, the device comprising: a memory, a processor, and a fire accident video analysis program stored in the memory and executable on the processor, the fire accident video analysis program being configured to implement the steps of the fire accident video analysis method described above.

[0044] In addition, to achieve the above objectives, this application also proposes a storage medium storing a fire accident video analysis program, which, when executed by a processor, implements the steps of the fire accident video analysis method described above.

[0045] In addition, to achieve the above objectives, this application also proposes a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the fire accident video analysis method described above.

[0046] This application extracts fire feature data and personnel / vehicle data from fire scene video data; then classifies and identifies the fire feature data to obtain flame, smoke, and lighting information; performs super-resolution reconstruction and recognition on personnel behavior images from the personnel / vehicle data to obtain personnel feature information; and performs de-motion blurring on vehicle images from the personnel / vehicle data to obtain vehicle feature information. Finally, based on the flame, smoke, lighting, personnel, and vehicle feature information, a fire accident video analysis report is generated. This application integrates multiple video analysis technologies to automatically extract fire feature data, personnel feature information, and vehicle feature information from fire scene video data, and performs classification, recognition, super-resolution reconstruction, and de-motion blurring on this information, effectively improving the efficiency and accuracy of data extraction and recognition. The classification and recognition of fire feature data can accurately assess the starting location and spread direction of a fire; super-resolution reconstruction improves the accuracy of face recognition in complex environments; and de-motion blurring of license plates enhances the ability to recognize fast-moving or blurred license plates. Ultimately, the fire accident video analysis report generated based on the above information can reflect the actual situation at the fire scene, providing strong decision support for fire emergency response, cause investigation and responsibility determination, thereby improving the effectiveness of fire video analysis. Attached Figure Description

[0047] Figure 1 This is a flowchart illustrating the first embodiment of the fire accident video analysis method of this application;

[0048] Figure 2 This is a schematic diagram of a sub-process in the second embodiment of the fire accident video analysis method of this application;

[0049] Figure 3 This is a schematic diagram of a sub-process in the third embodiment of the fire accident video analysis method of this application;

[0050] Figure 4 This is a schematic diagram of the module structure of the fire accident video analysis device according to an embodiment of this application;

[0051] Figure 5 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the fire accident video analysis method in this application embodiment.

[0052] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0053] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of this application.

[0054] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0055] It should be noted that with the acceleration of urbanization and the increase in building density, the frequency and complexity of fire accidents are constantly rising. Rapid response and accurate analysis of fire accidents have become urgent problems for fire departments and related units. Traditional fire accident analysis mainly relies on manual review of surveillance videos to replay and assess the situation on-site. This method is cumbersome and inefficient, consuming significant human resources and easily susceptible to human error, leading to incomplete information extraction or inaccurate judgments. Furthermore, the complex environment at fire scenes, such as changes in lighting, smoke, and the rapid movement of people and vehicles, further increases the difficulty and error rate of manual analysis. In recent years, with the rapid development of artificial intelligence and computer vision technologies, automated analysis methods based on video data have gradually become a research hotspot. Deep learning models have significantly improved performance in image and video recognition, automatically extracting various feature information from fire scene videos, such as flames, smoke, changes in light and shadow, faces, and license plates, and performing efficient classification and recognition. However, most existing fire video analysis technologies only detect single types of feature information, lacking comprehensive analysis of multiple feature data. This results in an inability to generate comprehensive and accurate fire accident analysis reports, leading to poor performance in fire scene video analysis. Therefore, improving the effectiveness of fire video analysis has become an urgent technical problem to be solved.

[0056] The main solution of this application is as follows: Fire feature data and personnel / vehicle data are extracted from fire scene video data; then, the fire feature data is classified and identified to obtain flame information, smoke information, and light and shadow information; super-resolution reconstruction and recognition of personnel behavior images in the personnel and vehicle data are performed to obtain personnel feature information; motion blur removal processing is applied to vehicle images in the personnel and vehicle data to obtain vehicle feature information; finally, a fire accident video analysis report is generated based on the flame information, smoke information, light and shadow information, personnel feature information, and vehicle feature information.

[0057] This application integrates multiple video analytics technologies to automatically extract fire feature data, personnel feature information, and vehicle feature information from fire scene video data. It then performs classification, identification, super-resolution reconstruction, and motion blur removal on this information, effectively improving the efficiency and accuracy of data extraction and identification. The classification and identification of fire feature data can accurately assess the starting location and spread direction of a fire; super-resolution reconstruction improves facial recognition accuracy in complex environments; and motion blur removal of license plates enhances the ability to identify fast-moving or blurred license plates. Ultimately, the fire accident video analysis report generated based on the above information reflects the actual situation at the fire scene, providing strong decision support for fire emergency response, cause investigation, and liability determination, thereby improving the effectiveness of fire video analysis.

[0058] It should be noted that the executing entity of the method in this embodiment can be a computing service device with data processing, network communication, and program execution functions, or it can be the aforementioned fire accident video analysis device with the same or similar functions. This embodiment and the following embodiments will be described using a fire accident video analysis device as an example.

[0059] Based on this, a first embodiment of the fire accident video analysis method of this application is proposed. Please refer to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the fire accident video analysis method of this application.

[0060] In this embodiment, the fire accident video analysis method includes the following steps:

[0061] S1: Extract fire feature data and personnel and vehicle data from the fire scene video data;

[0062] It should be noted that fire scene video data refers to raw video data collected at the fire scene using various monitoring devices (such as surveillance cameras, mobile phone cameras, dashcams, etc.). This data includes images, sounds, and other relevant information during the fire's occurrence and forms a crucial foundation for fire accident analysis. Fire characteristic data refers to specific fire-related information extracted from fire scene videos, including but not limited to key visual features indicating the fire's state and development, such as flames, smoke, and changes in light and shadow. This data is used to analyze the fire's origin, spread path, and intensity. Personnel and vehicle data refers to facial and vehicle images appearing in the fire scene videos. This data is used to identify relevant personnel and vehicle information at the fire scene to help determine the activity of people and vehicles at the accident site.

[0063] Specifically, fire feature data and personnel and vehicle data are extracted from the raw video data collected at the fire scene. When extracting fire feature data, the equipment automatically detects key features such as flames, smoke, and changes in light and shadow in video frames using a pre-trained deep learning model, and records information such as the time, location, shape, and trend of these features. Simultaneously, feature filtering algorithms are used to remove non-fire-related visual information to ensure that the extracted data accurately reflects the actual fire situation.

[0064] Furthermore, for the extraction of face and license plate data, face detection algorithms (such as MTCNN or YOLO models based on deep learning) are used to scan the video frame by frame to identify the appearance of faces and extract their image information; simultaneously, license plate recognition algorithms (such as YOLO or SSD) are used to detect and extract the image of the license plate area. To improve the accuracy of recognition, the device can combine multi-frame analysis methods to verify the reliability of the recognition results by comparing the changes in the positions of faces and license plates in consecutive frames, ensuring that the identity and information of each detected target are clear and unambiguous.

[0065] By automatically extracting fire feature data and personnel and vehicle data from fire scene videos, this method significantly improves the efficiency and accuracy of fire video analysis. Utilizing deep learning algorithms and feature detection technology, it can quickly and accurately identify key fire features (such as flames, smoke, and lighting changes) and specific information about personnel and vehicles in the video, reducing the workload of manual video viewing and analysis and subjective judgment errors. Furthermore, multi-frame analysis technology enhances the robustness of the identification, effectively avoiding single-frame misjudgments and missed detections, providing a high-quality data foundation for subsequent fire development assessment, personnel identification, and vehicle tracking. The automation and intelligence of this method make it more suitable for dealing with complex and ever-changing fire scene situations, supporting rapid and accurate emergency response and accident investigation.

[0066] S2: Classify and identify the fire feature data to obtain flame information, smoke information, and light and shadow information;

[0067] It should be noted that flame information refers to characteristic data related to flames in the video, including the color, shape, brightness, area, location, and trend of flame changes. It is primarily used to determine the location of the fire source and the spread of the fire. Smoke information refers to characteristic data related to smoke in the video, including the color, density, diffusion speed, direction, and area coverage of smoke. It is used to assess the impact of smoke generated by the fire and its diffusion path. Light and shadow information refers to abnormal changes in light and images caused by factors such as flames and smoke in the video, including flickering light sources and changes in light and shadow reflections. This helps determine the onset time of the fire and the fire's intensity under specific conditions.

[0068] Specifically, in the process of classifying and identifying fire feature data, firstly, raw data related to flames, smoke, and light and shadow are obtained from the extracted fire scene video data, and this data is input into a trained deep learning model. For flame information identification, the model uses features such as the color, shape, and brightness of the flames for classification and identification, identifying the precise location, area, and trend of change of the flames; for smoke information identification, the model locates and identifies the specific location, range, and concentration changes of smoke by analyzing grayscale changes, texture features, and regional diffusion; and for light and shadow information identification, it detects light intensity changes and reflection features in video frames to discover abnormal light and shadow phenomena caused by flames and smoke.

[0069] Furthermore, to improve the accuracy of classification and recognition, the computer equipment utilizes temporal analysis technology to comprehensively analyze data from multiple video frames, eliminating misjudgments and noise interference in single-frame detection. By tracking the temporal evolution of flames, smoke, and light, the system can establish a timeline of dynamic changes at the fire scene, accurately assess the fire's starting location, spread rate, and impact range, thereby forming a complete dataset of flame, smoke, and light information.

[0070] By classifying and identifying fire characteristic data, this step significantly improves the accuracy and efficiency of fire analysis. The precise extraction and identification of flame, smoke, and light information enables the system to monitor the dynamic changes at the fire scene in real time, identify the fire's origin and spread path, and provide a more accurate basis for fire emergency decision-making. Furthermore, the application of deep learning and time-series analysis techniques avoids the subjective misjudgments and delays inherent in traditional manual analysis, effectively improving the response speed and reliability of fire monitoring. This comprehensive data classification and identification method provides crucial decision support and an information foundation for rapid emergency response, cause investigation, and follow-up handling at fire scenes.

[0071] S3: Perform super-resolution reconstruction and recognition on the personnel behavior images in the personnel and vehicle data to obtain personnel feature information;

[0072] It should be noted that images of people's behavior refer to facial regions identified and extracted from videos. Due to limitations in video quality or on-site conditions, these images may often be blurry or have low resolution. Super-resolution reconstruction is a technique that uses deep learning or image processing technology to convert low-resolution images into high-resolution images. By filling in details, it makes the images clearer and more accurate, facilitating further identification and analysis. Personnel feature information refers to the set of feature information used for personnel identification and verification, obtained through high-resolution reconstruction of facial images.

[0073] Specifically, the extracted low-resolution images of people's behavior are input into a pre-trained super-resolution reconstruction model. This model processes the input face image through a series of convolutional neural network layers, extracting its features and utilizing a high-resolution image feature library learned within the model to gradually recover and reconstruct a higher-resolution image. The model iteratively optimizes by comparing the differences between the original low-resolution image and the generated high-resolution image, producing clearer and more detailed face images.

[0074] Furthermore, feature extraction algorithms are used to analyze the reconstructed high-resolution images of human behavior, extracting key facial feature points (such as the position and contour of the eyes, nose, and mouth) and generating facial feature vectors. These feature vectors can be used for subsequent face recognition and comparison to help identify and confirm the identities of relevant personnel at the fire scene.

[0075] Super-resolution reconstruction of facial images significantly improves the clarity and recognizability of faces in low-quality videos, solving the problem of facial recognition difficulties caused by low video resolution, insufficient lighting, or blurry images. This technology enables accurate identification and confirmation of individuals even in complex environments such as fires, providing more reliable data support for personnel screening and subsequent investigations at fire scenes. Furthermore, the super-resolution reconstruction method can quickly process large numbers of low-resolution images, reducing recognition time and improving the efficiency and accuracy of fire emergency response.

[0076] S4: Perform motion blur removal processing on the vehicle images in the personnel and vehicle data to obtain vehicle feature information;

[0077] It should be noted that personnel and vehicle data refers to visual information data, including facial and vehicle images, extracted from fire scene videos. This data is used to identify and confirm relevant personnel and vehicles at the scene. Vehicle images refer to license plate area images of vehicles identified and extracted from the video. These images may be blurry or difficult to identify due to rapid vehicle movement or camera shake. Motion blur removal is an image restoration technique used to eliminate image blur caused by camera shake or object movement, restoring details and making blurry images clear and distinguishable. Vehicle feature information refers to the clear license plate data obtained after motion blur removal of vehicle images, including license plate number, characters, and related vehicle information, used for vehicle identification and tracking.

[0078] Specifically, the extracted blurred vehicle image is input into a motion blur removal model. This model uses deep learning algorithms or traditional image restoration-based algorithms to detect and estimate motion blur features in the vehicle image. By analyzing the image's blur kernel (such as the direction and speed of motion), the model can generate a deconvolution or inverse filter to eliminate the blur effect. This process uses multiple convolutional layers to progressively restore details in the vehicle image, optimizing the sharpness of each pixel, and ultimately generating a higher-resolution vehicle image.

[0079] Furthermore, Optical Character Recognition (OCR) technology is applied to perform character recognition on the deblurred vehicle image. Based on the clear shapes of the license plate characters, the OCR algorithm accurately converts the letters and numbers in the license plate into text information. Simultaneously, by combining the license plate's format rules and regional features, the correctness of the recognition results is further verified, ensuring the accurate extraction of vehicle feature information.

[0080] By applying motion blur removal processing to vehicle images, image blurring caused by rapid vehicle movement or camera shake is effectively eliminated, making license plate characters clearly legible. This processing method significantly improves the accuracy and reliability of license plate recognition, especially in low-light, complex background, or fast-moving scenarios, ensuring accurate acquisition of vehicle feature information. The application of motion blur removal technology solves the recognition errors and omissions caused by image blurring in traditional license plate recognition, providing more reliable evidence for vehicle tracking, liability determination, and accident investigation at fire scenes. Simultaneously, this technology can process large numbers of images in a short time, improving recognition efficiency and facilitating rapid response and decision-making.

[0081] S5: Based on the flame information, smoke information, light and shadow information, personnel characteristic information, and vehicle characteristic information, generate a fire accident video analysis report.

[0082] Specifically, by combining flame, smoke, and light and shadow information, a dynamic change model of the fire scene is established. By analyzing the color, area, location, and trend of the flames, the starting point and spread path of the fire are determined; by combining the diffusion speed, concentration, and coverage of the smoke, the extent of the fire's spread and its impact on the surrounding environment are assessed; and by utilizing light and shadow change information (such as light source flickering and shadow movement), the time of the fire's occurrence and its development process are further precisely located, forming a comprehensive analysis of the fire scene.

[0083] Furthermore, personnel and vehicle characteristic information is used to identify and track the movements of people and vehicles at the fire scene. The extracted high-resolution personnel characteristic information is matched with personnel information in a known database to determine the identities and behavioral patterns of relevant personnel at the fire scene. Simultaneously, vehicle characteristic information is identified and analyzed to track vehicles appearing at the scene and their movement paths. All information is integrated and correlated to generate a fire accident video analysis report that includes the cause of the fire, its spread, and the activities of personnel and vehicles.

[0084] By comprehensively analyzing flame, smoke, light and shadow, personnel, and vehicle characteristics, the generated fire accident video analysis report can comprehensively and accurately reflect the occurrence, development, and impact of a fire accident. This report not only provides the fire's origin, spread path, and possible causes, but also identifies relevant personnel and vehicles at the scene, providing strong evidence for accident liability determination and subsequent investigations. Furthermore, the automated generation of the analysis report improves the efficiency and accuracy of fire accident analysis, reduces the workload and subjective errors of manual intervention, and significantly enhances emergency response speed and the scientific basis of decision-making.

[0085] This embodiment extracts fire feature data and personnel / vehicle data from fire scene video data; then, it classifies and identifies the fire feature data to obtain flame, smoke, and lighting information; it performs super-resolution reconstruction and identification of personnel behavior images from the personnel / vehicle data to obtain personnel feature information; it performs de-motion blurring on vehicle images from the personnel / vehicle data to obtain vehicle feature information; finally, based on the flame, smoke, lighting, personnel, and vehicle feature information, it generates a fire accident video analysis report. This embodiment integrates multiple video analysis technologies to automatically extract fire feature data, personnel feature information, and vehicle feature information from fire scene video data, and performs classification, identification, super-resolution reconstruction, and de-motion blurring on this information, effectively improving the efficiency and accuracy of data extraction and identification. The classification and identification of fire feature data can accurately assess the starting location and spread direction of a fire; super-resolution reconstruction improves the accuracy of face recognition in complex environments; and de-motion blurring of license plates enhances the ability to identify fast-moving or blurred license plates. Ultimately, the fire accident video analysis report generated based on the above information can reflect the actual situation at the fire scene, providing strong decision support for fire emergency response, cause investigation and responsibility determination, thereby improving the effectiveness of fire video analysis.

[0086] Based on the first embodiment described above, a second embodiment of the fire accident video analysis method of this application is proposed. Please refer to... Figure 2 , Figure 2 This is a schematic diagram of a sub-process in the second embodiment of the fire accident video analysis method of this application.

[0087] like Figure 2 As shown, in this embodiment, before step S1, the following steps are also included:

[0088] S1a: Identify each video input channel based on the topology of local storage for on-site monitoring;

[0089] S1b: Determine the target video input channel based on the device type, resolution, frame rate, and video encoding format corresponding to each video input channel;

[0090] S1c: Obtain the target surveillance video based on the target video input channel;

[0091] S1d: Optimize the target monitoring video to obtain the fire scene video data.

[0092] It should be noted that the topology of the monitoring network refers to the layout and connection relationships of various video surveillance devices at the fire scene, including the physical location of the devices, network connection methods, and data transmission paths, used for effective management and scheduling of video resources. Video input channels refer to the communication paths used to transmit monitoring video data. Each channel corresponds to a specific monitoring device or camera, capable of independently acquiring, transmitting, and storing monitoring video. Device type, resolution, frame rate, and video encoding format refer to the specific type of monitoring device (e.g., fixed camera, mobile camera), the clarity of the video image (e.g., 1080p, 4K), the number of frames displayed per second (e.g., 30fps, 60fps), and the compression encoding format of the video data (e.g., H.264, H.265), respectively. These parameters determine the quality and transmission performance of the video data. Target video input channels refer to the monitoring video channels that, after screening and optimization, meet predetermined conditions (e.g., video quality and device stability). These channels are selected to obtain the most useful video data. Target monitoring video refers to the fire scene video data acquired through the target video input channels, which, after processing and optimization, is used for further analysis. Optimization processing refers to a series of enhancement and adjustment operations performed on the acquired video data, such as noise reduction, contrast adjustment, and deblurring, in order to improve video quality and analysis results.

[0093] Specifically, based on the topology of the fire scene monitoring network, each video input channel is identified, and the monitoring equipment connected to each channel, its physical location, and network connection status are determined. Combined with the on-site monitoring layout, key information such as the device type (e.g., fixed or mobile camera), resolution, frame rate, and video encoding format for each video input channel is extracted to comprehensively understand the performance characteristics and video data quality of each channel. Based on the identified information, channels are filtered and prioritized to determine the target video input channels. The filtering process mainly considers video clarity (high resolution), smoothness (high frame rate), compression efficiency (suitable encoding format), and device reliability (e.g., low failure rate and good network connectivity). After selecting the target video input channels, the target monitoring video data is obtained from them and stored in the designated storage system.

[0094] Furthermore, after acquiring the target surveillance video, these videos are optimized to improve their effectiveness in fire analysis. Optimization includes noise reduction to remove random noise from the video; adjusting contrast and brightness to improve image clarity and legibility; and using deblurring algorithms to remove blur caused by equipment shake or insufficient lighting to restore image details. The resulting optimized fire scene video data possesses higher quality and reliability, making it suitable for subsequent fire feature extraction and analysis.

[0095] By identifying video input channels based on the topology of the on-site monitoring network and selecting the optimal target video input channel according to parameters such as device type, resolution, frame rate, and encoding format, it helps to select the highest quality and most reliable video source, ensuring high-quality video data acquisition. Optimization of the target monitoring video further improves video clarity and usability, enhancing the accuracy and effectiveness of fire scene video data, and providing more accurate and efficient basic data support for subsequent fire feature analysis, personnel identification, and accident investigation. This optimization method significantly improves the overall effectiveness of fire video analysis and the speed of emergency response.

[0096] In this embodiment, step S1d includes:

[0097] S1d1: Convert the format of the target surveillance video to a preset standard format;

[0098] S1d2: Based on a preset data recovery method, the lost frames of the target surveillance video after format conversion are recovered to obtain multiple restored surveillance videos;

[0099] S1d3: Perform image enhancement processing on the multiple restored surveillance videos to obtain multiple enhanced videos;

[0100] S1d4: Synchronize the timestamps of the multiple enhanced videos;

[0101] S1d5: Perform integrity verification on the enhanced video after timestamp synchronization, and use the video that passes the integrity verification as the fire scene video data.

[0102] It should be noted that "target surveillance video" refers to surveillance video data that has been filtered and optimized, and then further processed for fire scene analysis. "Preset standard format" refers to a unified video format (such as MP4, AVI, etc.) pre-defined in the fire video analysis system, ensuring that all video data can be processed and analyzed under the same standard. "Data recovery methods" refer to methods used to recover video data lost due to transmission interruptions, equipment failures, etc., including interpolation algorithms and machine learning models, used to reconstruct and recover lost image frames. "Restored surveillance video" refers to video data that has been supplemented with lost frames after data recovery processing, restoring the integrity of the video. "Image enhancement processing" refers to optimizing video images, such as noise reduction, contrast enhancement, and edge sharpening, to improve video clarity and recognizability. "Timestamp synchronization" refers to calibrating the time stamps of multiple video data sets to ensure that videos from different channels are accurately aligned on the timeline, facilitating subsequent analysis and comparison. "Integrity verification" refers to verifying the processed video data to ensure data integrity and consistency, eliminating the possibility of data loss, corruption, or errors.

[0103] Specifically, the target surveillance video format is converted to a preset standard format for unified processing and analysis within the fire video analysis system. During the format conversion, the video data is re-encoded and compressed according to the requirements of the preset standard format (such as resolution, encoding format, frame rate, etc.) to ensure compatibility with the system's analysis software and algorithms. For the converted video data, a preset data recovery method is used to recover lost frames. During data recovery, interpolation algorithms or machine learning models are employed to estimate and supplement frames lost due to transmission interruptions, equipment failures, etc., generating multiple restored surveillance videos. Subsequently, image enhancement processing is performed on the restored video data. Through techniques such as noise reduction, contrast adjustment, and edge sharpening, the video quality is improved, enhancing clarity and detail reproduction, resulting in multiple enhanced videos.

[0104] Furthermore, the enhanced videos undergo timestamp synchronization to ensure accurate alignment of multiple video data from different channels on the timeline. The timestamp synchronization process involves adjusting and matching the start time and inter-frame time of the videos to guarantee that all videos are analyzed synchronously on the same time base. After synchronization, the timestamp-calibrated enhanced videos undergo integrity verification. By comparing the hash value or checksum of the verification data, it is ensured that the video data has not been lost, damaged, or tampered with. Only videos that pass the integrity verification are considered qualified and used as the final fire scene video data for subsequent analysis.

[0105] By converting the target surveillance video format to a preset standard format, the compatibility and uniformity of the video data are ensured, making subsequent processing and analysis smoother and more efficient. Data recovery methods are used to recover lost frames, effectively solving the problem of video data loss caused by transmission interruptions or equipment failures, ensuring video integrity. Image enhancement processing improves the clarity and recognizability of the video, providing high-quality basic data for the accurate extraction of fire features. Timestamp synchronization and integrity verification further ensure the precise alignment of multiple video data streams on the timeline and the integrity and consistency of the data, providing reliable data support for the accurate analysis and judgment of fire accidents. Overall, this series of steps significantly improves the accuracy, reliability, and efficiency of fire video analysis.

[0106] This embodiment extracts fire feature data and personnel / vehicle data from fire scene video data; then, it classifies and identifies the fire feature data to obtain flame, smoke, and lighting information; it performs super-resolution reconstruction and identification of personnel behavior images from the personnel / vehicle data to obtain personnel feature information; it performs de-motion blurring on vehicle images from the personnel / vehicle data to obtain vehicle feature information; finally, based on the flame, smoke, lighting, personnel, and vehicle feature information, it generates a fire accident video analysis report. This embodiment integrates multiple video analysis technologies to automatically extract fire feature data, personnel feature information, and vehicle feature information from fire scene video data, and performs classification, identification, super-resolution reconstruction, and de-motion blurring on this information, effectively improving the efficiency and accuracy of data extraction and identification. The classification and identification of fire feature data can accurately assess the starting location and spread direction of a fire; super-resolution reconstruction improves the accuracy of face recognition in complex environments; and de-motion blurring of license plates enhances the ability to identify fast-moving or blurred license plates. Ultimately, the fire accident video analysis report generated based on the above information can reflect the actual situation at the fire scene, providing strong decision support for fire emergency response, cause investigation and responsibility determination, thereby improving the effectiveness of fire video analysis.

[0107] Based on the second embodiment described above, a third embodiment of the fire accident video analysis method of this application is proposed. Please refer to... Figure 3 , Figure 3 This is a schematic diagram of a sub-process in the third embodiment of the fire accident video analysis method of this application.

[0108] In this embodiment, step S1 includes:

[0109] S11: Preprocess the fire scene video to obtain a target frame set, wherein the preprocessing includes one or more of video frame decomposition, image enhancement, and region of interest detection;

[0110] S12: Based on the preset fire feature detection model, flame detection, smoke detection and light and shadow detection are performed on each frame in the target frame set to obtain the fire feature data;

[0111] S13: Based on the preset personnel and vehicle detection model, perform personnel detection and vehicle detection on each frame in the target frame set to obtain the personnel and vehicle data.

[0112] It should be noted that fire scene video refers to raw video data captured by on-site monitoring equipment during a fire, containing information about the fire's development and related personnel and vehicles. Preprocessing refers to a series of preliminary processing operations performed on the fire scene video before further analysis, including video frame decomposition, image enhancement, and region of interest (ROI) detection, to improve the accuracy and efficiency of subsequent detection. The target frame set refers to the preprocessed video frame set, containing all filtered and enhanced image frames, used for fire feature detection and face / license plate detection. Video frame decomposition refers to breaking down the video stream into a series of independent static image frames for frame-by-frame analysis and processing. Image enhancement refers to improving the quality and clarity of video frames through image processing techniques, such as noise reduction, contrast adjustment, and brightness optimization. ROI detection refers to identifying and marking regions in the image that may contain fire features (such as flames and smoke) or important information such as faces and license plates, to narrow the analysis scope and improve detection efficiency. The fire feature detection model refers to a deep learning model specifically trained to identify fire scene features (such as flames, smoke, and changes in light and shadow). The face and license plate detection model refers to a deep learning model specifically trained to recognize faces and license plates, which can quickly and accurately locate and identify facial and vehicle feature information in video frames.

[0113] Specifically, the fire scene video undergoes preprocessing to generate a target frame set. The preprocessing steps include video frame decomposition, breaking down the continuous video stream into a series of static image frames for frame-by-frame analysis; image enhancement processing is performed on each decomposed frame, improving image clarity and recognizability through methods such as noise reduction, contrast adjustment, and brightness optimization; and region of interest detection (ROI) technology is used to identify and label regions in the image that may contain important information such as flames, smoke, faces, and license plates, thereby narrowing the analysis scope and reducing computational load and false positive rate. These preprocessing steps result in a high-quality target frame set, facilitating subsequent feature detection and analysis.

[0114] Furthermore, based on a pre-defined fire feature detection model, flame detection, smoke detection, and light and shadow detection are performed on each frame in the target frame set. The fire feature detection model utilizes deep learning algorithms to automatically analyze information such as color, brightness, shape, texture, and dynamic changes in each frame to identify and classify the flame, smoke, and light and shadow changes at the fire scene, generating detailed fire feature data. Simultaneously, based on a pre-defined face and license plate detection model, face and license plate detection are performed on each frame in the target frame set. This model uses advanced algorithms such as convolutional neural networks to quickly locate and identify face and license plate regions in the image and extract relevant data to obtain personnel and vehicle data containing information about people and vehicles.

[0115] By preprocessing fire scene videos (such as video frame decomposition, image enhancement, and region of interest detection), the clarity and analyzability of the video data are effectively improved, and interference from noise and irrelevant information is reduced, making subsequent fire feature and face / license plate detection more accurate and efficient. Using a pre-defined fire feature detection model to detect flames, smoke, and light, the main characteristics of a fire can be quickly identified, providing crucial information for fire cause analysis and development prediction. Simultaneously, the face / license plate detection model accurately identifies personnel and vehicle information at the fire scene, ensuring comprehensive monitoring and rapid response to the situation. Overall, this multi-layered, multi-step detection method significantly improves the accuracy and efficiency of fire accident analysis, providing reliable data support for emergency response and accident investigation.

[0116] Based on the second embodiment described above, in this embodiment, step S3 includes:

[0117] S31: Extract personnel behavior images from the personnel and vehicle data, and evaluate the quality of the extracted personnel behavior images;

[0118] S32: Remove face images whose quality evaluation value is lower than the preset face image quality standard evaluation value to obtain quality face images;

[0119] S33: Standardize the size, color space, brightness, and contrast of the quality face image to obtain a standard quality face image;

[0120] S34: Based on a preset super-resolution reconstruction model, the standard quality face image is reconstructed and enhanced to obtain a reconstructed face image;

[0121] S35: Based on a preset face recognition algorithm, feature extraction is performed on the reconstructed face image to obtain the personnel feature information.

[0122] It should be noted that personnel and vehicle data refers to visual information data containing facial and vehicle images extracted from fire scene videos, used to identify and confirm relevant personnel and vehicles at the scene. Facial images refer to images of facial regions identified and extracted from the video, used for further analysis and recognition. Quality evaluation refers to the quality assessment of the extracted facial images, judging their quality by measuring parameters such as image sharpness, contrast, and noise level to ensure that the facial images used for subsequent processing and recognition meet quality standards. Standardization processing refers to the uniformization of the size, color space, brightness, and contrast of facial images to make them conform to preset analysis standards. Super-resolution reconstruction models refer to models that use deep learning or image processing techniques to convert low-resolution images into high-resolution images, enhancing details and sharpness to make the images easier to recognize and analyze. Facial recognition algorithms refer to algorithms used to extract facial feature vectors from images; by comparing these feature vectors, the identity of people in the image can be identified and confirmed.

[0123] Specifically, all facial images are extracted from the personnel and vehicle data, and these images are then evaluated for quality. The quality evaluation process involves measuring parameters such as image sharpness, contrast, noise level, and illumination uniformity to score or rate each facial image, and removing low-quality images whose quality evaluation values ​​are lower than the preset facial image quality standard evaluation value, thus obtaining an image set that meets the quality standard, i.e., high-quality facial images.

[0124] Furthermore, these high-quality face images are standardized by adjusting their size, color space, brightness, and contrast to a uniform standard to ensure all images are input into the subsequent analysis model under the same conditions. Next, a pre-defined super-resolution reconstruction model is used to reconstruct and enhance the standard-quality face images. A deep learning model is used to progressively increase the image resolution and details, making the face images clearer and easier to recognize. Finally, based on a pre-defined face recognition algorithm, features are extracted from the reconstructed and enhanced high-resolution face images to generate personnel feature information for identification.

[0125] By evaluating and filtering the extracted facial images, high-quality facial data for subsequent analysis was ensured, thereby reducing recognition errors caused by poor image quality. Standardization ensured that all images were analyzed under uniform conditions, avoiding model instability caused by excessive differences in image features. Using a super-resolution reconstruction model to enhance facial images transformed low-resolution or blurry facial images into high-resolution images, improving the accuracy and robustness of facial recognition. Ultimately, the high-quality facial feature information extracted by the facial recognition algorithm enables more reliable identification and tracking of individuals, effectively improving the accuracy and efficiency of personnel identification at fire scenes, and providing strong data support for fire accident investigation and liability determination.

[0126] Based on the second embodiment described above, in this embodiment, step S4 includes:

[0127] S41: Extract the vehicle image from the personnel and vehicle data, and perform the quality evaluation on the extracted vehicle image;

[0128] S42: Remove vehicle images whose quality evaluation value is lower than the preset vehicle image quality standard evaluation value to obtain quality vehicle images;

[0129] S43: Perform the normalization process on the size, color space, brightness, and contrast of the quality vehicle image to obtain a standard quality vehicle image;

[0130] S44: Divide the standard quality vehicle image into multiple license plate images and determine the fuzzy information of each license plate image;

[0131] S45: Based on the fuzzy information, input the images of each license plate into a preset deblurring model to obtain the vehicle feature information.

[0132] It should be noted that vehicle images refer to the license plate area images identified and extracted from videos, used for further license plate recognition and analysis. Quality evaluation refers to the quality assessment of the extracted vehicle images, judging image quality by measuring parameters such as image sharpness, contrast, and noise level to ensure that the vehicle images used for subsequent processing and recognition meet standards. Standardization processing refers to unifying the size, color space, brightness, and contrast of the vehicle images to make them conform to preset analysis standards. License plate sub-images refer to further dividing the standardized vehicle images into multiple smaller image blocks to facilitate the analysis and processing of the blurriness of each local area. Blur information refers to the blurry feature information in the license plate image, including the degree, direction, and range of blur, reflecting the state of image sharpness. Deblurring model refers to the deep learning model or image processing algorithm used to eliminate image blur, improving image sharpness and recognizability by restoring image details. Vehicle feature information refers to the clear license plate data containing the license plate number, characters, and related vehicle information after deblurring processing.

[0133] Specifically, vehicle images are extracted from personnel and vehicle data, and the quality of each extracted vehicle image is evaluated. The quality evaluation process analyzes parameters such as image sharpness, contrast, and noise level to determine the quality of the vehicle image; images with quality evaluation values ​​lower than the preset vehicle image quality standard are discarded to ensure that only high-quality images are used for subsequent processing, thereby obtaining high-quality vehicle images.

[0134] Furthermore, the high-quality vehicle images undergo standardization processing, including adjusting image size, color space, brightness, and contrast to conform to preset analysis standards, generating standard-quality vehicle images. Then, the standard-quality vehicle images are further segmented into multiple license plate images (i.e., smaller regions of the vehicle image), and the blurring information (such as blur degree and direction) of each license plate image is calculated and determined. Based on this blurring information, the license plate images are input into a preset deblurring model. The model eliminates blurring through algorithms such as deconvolution and iterative optimization, restoring the details of the vehicle image and ultimately obtaining clear and identifiable vehicle feature information.

[0135] By evaluating and screening the extracted vehicle images, high-quality images were ensured for subsequent deblurring, reducing interference from low-quality images in license plate recognition. Standardization further unified image features, avoiding recognition errors caused by excessive image differences. Segmenting vehicle images into multiple sub-images and deblurring based on blur information helps to more accurately recover image details, especially under complex backgrounds or lighting conditions, effectively improving license plate recognition accuracy. The resulting clear vehicle feature information provides reliable data support for vehicle tracking and liability determination at fire scenes, facilitating rapid and accurate analysis of the fire scene and identification of responsible parties.

[0136] This embodiment extracts fire feature data and personnel / vehicle data from fire scene video data; then, it classifies and identifies the fire feature data to obtain flame, smoke, and lighting information; it performs super-resolution reconstruction and identification of personnel behavior images from the personnel / vehicle data to obtain personnel feature information; it performs de-motion blurring on vehicle images from the personnel / vehicle data to obtain vehicle feature information; finally, based on the flame, smoke, lighting, personnel, and vehicle feature information, it generates a fire accident video analysis report. This embodiment integrates multiple video analysis technologies to automatically extract fire feature data, personnel feature information, and vehicle feature information from fire scene video data, and performs classification, identification, super-resolution reconstruction, and de-motion blurring on this information, effectively improving the efficiency and accuracy of data extraction and identification. The classification and identification of fire feature data can accurately assess the starting location and spread direction of a fire; super-resolution reconstruction improves the accuracy of face recognition in complex environments; and de-motion blurring of license plates enhances the ability to identify fast-moving or blurred license plates. Ultimately, the fire accident video analysis report generated based on the above information can reflect the actual situation at the fire scene, providing strong decision support for fire emergency response, cause investigation and responsibility determination, thereby improving the effectiveness of fire video analysis.

[0137] This application also provides a fire accident video analysis device. Please refer to... Figure 4 , Figure 4 This is a schematic diagram of the module structure of the fire accident video analysis device according to an embodiment of this application. The fire accident video analysis device includes:

[0138] Data extraction module 401 is used to extract fire feature data and personnel and vehicle data from the fire scene video data;

[0139] Information recognition module 402 is used to classify and recognize the fire feature data to obtain flame information, smoke information and light and shadow information;

[0140] Personnel identification module 403 is used to perform super-resolution reconstruction of personnel behavior images in the personnel and vehicle data to obtain personnel feature information;

[0141] The vehicle recognition module 404 is used to perform motion blur removal processing on the vehicle images in the personnel and vehicle data to obtain vehicle feature information;

[0142] The target module 405 is used to generate a fire accident video analysis report based on the flame information, the smoke information, the light and shadow information, the personnel feature information, and the vehicle feature information.

[0143] The fire accident video analysis device provided in this application, employing the fire accident video analysis method described in the above embodiments, can solve the technical problem of how to improve the effectiveness of fire video analysis. Compared with the prior art, the beneficial effects of the fire accident video analysis device provided in this application are the same as those of the fire accident video analysis method provided in the above embodiments, and other technical features in the fire accident video analysis device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0144] This application provides a fire accident video analysis device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the fire accident video analysis method in the above embodiments.

[0145] The following is for reference. Figure 5 The diagram illustrates a structural schematic of a fire accident video analysis device suitable for implementing embodiments of this application. The fire accident video analysis device in this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), and vehicle terminals (e.g., vehicle navigation terminals), as well as fixed terminals such as digital TVs and desktop computers. Figure 5 The fire accident video analysis device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0146] like Figure 5As shown, the fire accident video analysis device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the fire accident video analysis device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to the I / O interface 1006: input devices 1007 including, for example, a touch screen, touchpad, keyboard, mouse, image sensor, microphone, accelerometer, gyroscope, etc.; output devices 1008 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 1003 including, for example, magnetic tape, hard disk, etc.; and communication devices 1009. Communication device 1009 allows the fire incident video analysis equipment to communicate wirelessly or wiredly with other devices to exchange data. Although the figure shows fire incident video analysis equipment with various systems, it should be understood that implementation or possession of all the systems shown is not required. More or fewer systems may be implemented alternatively.

[0147] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0148] The fire accident video analysis device provided in this application, employing the fire accident video analysis method in the above embodiments, can solve the technical problem of how to improve the effect of fire video analysis. Compared with the prior art, the beneficial effects of the fire accident video analysis device provided in this application are the same as those of the fire accident video analysis method provided in the above embodiments, and other technical features in this fire accident video analysis device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0149] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0150] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0151] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the fire accident video analysis method in the above embodiments.

[0152] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0153] The aforementioned computer-readable storage medium may be included in the fire accident video analysis device; or it may exist independently and not assembled into the fire accident video analysis device.

[0154] The aforementioned computer-readable storage medium carries one or more programs. When these programs are executed by the fire accident video analysis device, the fire accident video analysis device performs the following actions: extracts fire feature data and personnel / vehicle data from the fire scene video data; classifies and identifies the fire feature data to obtain flame information, smoke information, and light and shadow information; performs super-resolution reconstruction and recognition of personnel behavior images in the personnel / vehicle data to obtain personnel feature information; performs de-motion blurring processing on vehicle images in the personnel / vehicle data to obtain vehicle feature information; and generates a fire accident video analysis report based on the flame information, smoke information, light and shadow information, personnel feature information, and vehicle feature information. Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof. These programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as C or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0155] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0156] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0157] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described fire accident video analysis method, thereby solving the technical problem of how to improve the effectiveness of fire video analysis. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the fire accident video analysis method provided in the above embodiments, and will not be repeated here.

[0158] This application provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the fire accident video analysis method described above.

[0159] The computer program product provided in this application can solve the technical problem of how to improve the effect of fire video analysis. Compared with the prior art, the beneficial effects of the computer program product provided in the embodiments of this application are the same as the beneficial effects of the fire accident video analysis method provided in the above embodiments, and will not be repeated here.

[0160] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent scope of this application.

Claims

1. A method for analyzing fire accident videos, characterized in that, The method includes: Based on the topology of local storage for on-site monitoring, identify each video input channel; The target video input channel is determined based on the device type, resolution, frame rate, and video encoding format corresponding to each video input channel. Based on the target video input channel, acquire the target surveillance video; Convert the target surveillance video to a preset standard format; Based on a preset data recovery method, the lost frames of the target surveillance video after format conversion are recovered to obtain multiple restored surveillance videos. Image enhancement processing is performed on the multiple restored surveillance videos to obtain multiple enhanced videos; The timestamps of the multiple enhanced videos are synchronized. The enhanced video after timestamp synchronization is subjected to integrity verification, and the video that passes the integrity verification is used as fire scene video data. Extract fire feature data and personnel and vehicle data from fire scene video data; The fire feature data is classified and identified to obtain flame information, smoke information, and light and shadow information. Super-resolution reconstruction and recognition are performed on the images of human behavior in the personnel and vehicle data to obtain human feature information; The vehicle images in the personnel and vehicle data are subjected to motion blur removal processing to obtain vehicle feature information; Based on the flame information, smoke information, light and shadow information, personnel characteristic information, and vehicle characteristic information, a fire accident video analysis report is generated.

2. The method as described in claim 1, characterized in that, The step of extracting fire feature data and personnel and vehicle data from the fire scene video data includes: The fire scene video is preprocessed to obtain a target frame set, wherein the preprocessing includes one or more of video frame decomposition, image enhancement, and region of interest detection; Based on a preset fire feature detection model, flame detection, smoke detection, and light and shadow detection are performed on each frame in the target frame set to obtain the fire feature data. Based on a preset personnel and vehicle detection model, personnel and vehicle detection are performed on each frame in the target frame set to obtain the personnel and vehicle data.

3. The method as described in claim 1, characterized in that, The step of performing super-resolution reconstruction and recognition of personnel behavior images in the personnel and vehicle data to obtain personnel feature information includes: Extract human behavior images from the personnel and vehicle data, and evaluate the quality of the extracted human behavior images; Remove face images whose quality evaluation value is lower than the preset face image quality standard evaluation value to obtain quality face images; The size, color space, brightness, and contrast of the quality face image are standardized to obtain a standard quality face image; Based on a preset super-resolution reconstruction model, the standard quality face image is reconstructed and enhanced to obtain a reconstructed face image; Based on a preset face recognition algorithm, feature extraction is performed on the reconstructed face image to obtain the person's feature information.

4. The method as described in claim 3, characterized in that, The step of performing motion blur removal processing on the vehicle images in the personnel and vehicle data to obtain vehicle feature information includes: The vehicle image is extracted from the personnel and vehicle data, and the quality evaluation is performed on the extracted vehicle image; Vehicle images whose quality evaluation value is lower than the preset vehicle image quality standard evaluation value are removed to obtain quality vehicle images; The standardization process is applied to the size, color space, brightness, and contrast of the quality vehicle image to obtain a standard quality vehicle image. The standard quality vehicle image is divided into multiple license plate images, and the fuzzy information of each license plate image is determined. Based on the fuzzy information, the images of each license plate are input into a preset deblurring model to obtain the vehicle feature information.

5. A fire accident video analysis device, characterized in that, The device includes: The data extraction module is used to identify each video input channel based on the topology of the local storage of on-site monitoring; determine the target video input channel according to the device type, resolution, frame rate, and video encoding format corresponding to each video input channel; acquire the target monitoring video based on the target video input channel; convert the format of the target monitoring video to a preset standard format; recover lost frames of the converted target monitoring video using a preset data recovery method to obtain multiple restored monitoring videos; perform image enhancement processing on the multiple restored monitoring videos to obtain multiple enhanced videos; synchronize the timestamps of the multiple enhanced videos; perform integrity verification on the timestamp-synchronized enhanced videos, and use the videos that pass the integrity verification as fire scene video data; and extract fire feature data and personnel and vehicle data from the fire scene video data. The information recognition module is used to classify and recognize the fire feature data to obtain flame information, smoke information and light and shadow information; The personnel identification module is used to perform super-resolution reconstruction and identification of personnel behavior images in the personnel and vehicle data to obtain personnel feature information; The vehicle recognition module is used to perform motion blur removal processing on the vehicle images in the personnel and vehicle data to obtain vehicle feature information; The target module is used to generate a fire accident video analysis report based on the flame information, the smoke information, the light and shadow information, the personnel feature information, and the vehicle feature information.

6. A computer device, characterized in that, The device includes: a memory, a processor, and a fire accident video analysis program stored in the memory and executable on the processor, the fire accident video analysis program being configured to implement the steps of the fire accident video analysis method as described in any one of claims 1 to 4.

7. A storage medium, characterized in that, The storage medium stores a fire accident video analysis program, which, when executed by a processor, implements the steps of the fire accident video analysis method as described in any one of claims 1 to 4.

8. A computer program product, characterized in that, The computer program product includes a computer program that, when executed by a processor, implements the steps of the fire accident video analysis method as described in any one of claims 1 to 4.

Images