A method, system, medium, device and terminal for monitoring video key frame extraction

Abstract

The application belongs to the technical field of multimedia information processing, and discloses a kind of monitoring video key frame extraction method, system, medium, equipment and terminal, collect original video stream data, original video stream data is decomposed into image frame set;Image frame set obtained by decomposition is sampled, and the image frame result set obtained by sampling is filtered;The image frame set after filtering is adaptively clustered, and the result after clustering is collected to form a video abstract.In order to better utilize the memory space of storage medium and let the user better quickly browse the general content of original video stream, the application provides a key frame screening algorithm for original video stream by sampling, filtering and clustering.The key frame extraction method of the application screens out similar frames, redundant frames and fuzzy frames in original video stream through key frame screening algorithm for original video data, forms a video abstract storage of a section of original video, thereby greatly reduces the occupied storage space.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of multimedia information processing, and particularly relates to a monitoring video key frame extraction method, system, medium, device and terminal. BACKGROUND

[0002] Common video monitoring adopts the way of storing original video for saving, which results in that the captured video occupies a large amount of disk space and is not used for a long time, directly leading to the waste of storage resources and low utilization rate. Meanwhile, there are a large number of redundant frames, similar frames and blurred frames in the captured original video stream, and these image frames cannot be used by operators again, and there are also some redundant similar frames. It is just because of the lack of a video summary of the overall key information of the original video that this problem is derived, causing great waste of storage resources. The goal of video key frame extraction is to extract some representative static image frames from a video sequence for video summary, video retrieval or fast browsing and other applications. In recent years, the key frame extraction methods mainly include key frame extraction based on traditional image processing methods, key frame extraction based on machine learning and key frame extraction based on deep learning. However, although the methods based on machine learning and deep learning have obvious effects, the processing speed is slow and they cannot be completely suitable for current high-frame-rate monitoring video equipment; although the traditional image processing method is simple, the accuracy of extracting key frames is low.

[0003] Through the above analysis, the problems and defects of the prior art are that:

[0004] (1) The existing method is to decompress all original video frame data and then screen, and there is no need to decompress some blurred image frames without effective information;

[0005] (2) The key frame extraction method based on deep learning has a complex process and low processing efficiency;

[0006] (3) The existing method cannot be completely suitable in foggy weather environment; SUMMARY

[0007] In view of the problems existing in the prior art, the present application provides a monitoring video key frame extraction method, system, medium, device and terminal, and particularly relates to a monitoring video key frame real-time extraction method, system, medium, device and terminal based on improved adaptive clustering analysis.

[0008] The application is implemented by a monitoring video key frame extraction method, which comprises the following steps: collecting original video stream data and decomposing the original video stream data into an image frame set; sampling the decomposed image frame set and filtering the sampled image frame result set; adaptively clustering the filtered image frame set and collecting the clustered result to form a video summary.

[0009] Further, the monitoring video key frame extraction method further comprises the following steps: collecting original video stream data and performing decomposition processing to obtain an image frame set; performing image frame sampling based on a Gaussian mixture model background difference algorithm and performing image key frame sampling based on an IBP frame; calculating a decision segmentation weight and performing mean value sampling according to the decision segmentation weight; performing similarity filtering on the sampling result by using a hash fingerprint algorithm and performing adaptive clustering by using an adaptive clustering algorithm; and collecting the adaptive clustering result to form a final key frame set.

[0010] Further, the monitoring video key frame extraction method comprises the following steps:

[0011] Step one, first, the key area of the video is divided and extracted from the compressed domain. Since the I frame is the main carrier of image information, the DCT coefficient of the video image frame can be obtained by analyzing the I frame. The DCT coefficient is composed of a direct current coefficient (DC) and an alternating current coefficient (AC). The direct current coefficient is the main part and contains a large amount of information of the image frame. The alternating current coefficient contains some detailed information. Assuming that there are N I frames in a video segment, the video segment can be represented as The difference formula between two I frames is as shown in the following formula (2):

[0012]

[0013] Wherein, H i represents the histogram of the DC image of the i-th I frame. If HD(I i ,I i+1 ) exceeds the threshold value, it indicates that the two I frames come from different shots, which means that the video content has a shot change or the video picture change content is large. Otherwise, it indicates that the two I frames come from the same shot, and the video picture change is not large.

[0014] Step two, the video is divided into an image frame set. First, a first frame image is selected, and a frame image is extracted after an interval of k frames. The difference value of the two extracted frame images is calculated by background difference change, and the formula is as follows.

[0015]

[0016] Assuming the area of the changed region accounts for a proportion r of the entire frame image, if r = 0, it means that the two frames of images and the images in the interval have not changed, the frame image is discarded and a frame is reselected for background difference change calculation after an interval of k frames from the frame image; if 0 < r < t, it means that the frame image has changed compared to the selected first frame image, but the area of the changed region accounts for a small proportion, so the target may be far away from the gun-shaped camera or just enter the detection range of the gun-shaped camera, which means that the interval images between the selected two frames of images do not have important reference information, and only the current frame needs to be selected into the key frame set; if r ≥ t, it means that the frame image has obvious difference with the selected first frame image, and the area of the changed region reaches the set threshold t, which means that there is obvious difference between the frame image and the previous key frame image, and there is also a key frame in the interval image frame set between the two frames of images, after the current image frame is added to the key frame set, the dichotomy is used to find the key frame in the interval image frame set and add it to the key frame set; continue to execute according to the process until the last frame of image is calculated.

[0017] Step three, the preliminary image frame sampling result can be obtained through the sampling algorithm of the previous step, however, these results may have the problem of too high similarity between different segments of image frames and information redundancy, so it is necessary to further filter the similarity of the obtained initial image frame set, so as to filter out the redundant image frames that are too similar in the initial image frame set;

[0018] Step four, the sampling algorithm and the filtering algorithm are pre-processing of data, the image frame set obtained after the image frame similarity filtering of the previous step is used as the input data of the clustering algorithm, and the corresponding video summary is output after the clustering algorithm processing. Before clustering the image key frame set, the image features need to be extracted first, and the image is characterized by mathematical method. The commonly used image feature extraction methods at present include histogram-based image feature extraction and image entropy-based feature extraction technology. Since the scene in this paper is in a foggy environment, the image pictures collected are seriously blurred, and the detail texture and color information are severely damaged. The histogram-based image feature extraction is mostly based on color distribution information to extract image features, which brings difficulty to feature extraction. Therefore, this paper adopts the image entropy density-based feature extraction technology. The entropy density data of the image can reflect the information amount and complexity of the image. The larger the entropy value is, the more information it contains. The calculation method of image entropy is shown in the formula.

[0019]

[0020] In the formula, p xThe probability corresponding to the gray level x is represented by D, and the total number of gray levels is represented by D. The greater the entropy H(x) of the image, the richer the pixel gray scale contained in the image, the more uniform the gray scale distribution, and the greater the information quantity of the image. The feature extraction of the image first needs to block the target image. In this paper, the method of sliding window is used to start from the top left corner of the image, and then slide sequentially until the whole image is traversed. The target image is divided into 36 image blocks. Then calculate the entropy density value of each image block, and take it as the feature vector of the block. For each pixel in the block, calculate the entropy density value of the interval it belongs to, and take it as a dimension of the feature vector. In this paper, the number of clusters in the final clustering is specified as 15% of the total number of frames of the video segment. First, arrange all frames in descending order according to the size of image entropy density, then select the frame with the largest image entropy density in the current set as the initial clustering center in the initial stage, then measure the difference between each frame in the set and the current clustering center, if the maximum difference between the two frames is less than the given threshold t, they are classified into one class, otherwise the frame is too far away from the current clustering center and forms a class by itself, repeat the process until all key frame sets in the set are taken.

[0021] Step five, recombine the final key frame set into a video summary for the original video stream data.

[0022] Further, in step one, the duration of the collected original video stream data of the target person is not less than 1 second;

[0023] Save the image frame data obtained by decomposing the original video stream in a separate folder.

[0024] Further, in step two, the background difference based on Gaussian mixture model and the adaptive mean sampling algorithm of IBP frame are divided into two branches to make decisions. The first branch divides the image frame set of the entire video stream into multiple GOPs according to the key degree of interframe information. Each GOP starts with a key frame--I frame and serves as the interval between GOPs. Each GOP is composed of three types of frames: key frame--I frame, front and rear reference frame--B frame, and forward reference frame--P frame. The second branch divides the video into an image frame set, selects the first image, and extracts another image after an interval of k frames. For the two extracted images, calculate the background difference to get the ratio of the changed area to the entire image frame, which is r. Fuse the information of the two branches. The higher the corresponding weight value of the key frame segment selected by the two branches, the higher the interest value of the corresponding image frame set. After converting and counting all image frames in the video, set different w values for different segments according to the interest value--weight. The higher the w value, the more interesting image frames are obtained by sampling.

[0025] Further, in step three, after sampling is completed, the sampling result set of the original video is obtained, and the next step of filtering operation for similar frames and redundant frames is performed to complete the filtering of the candidate video frames from the similarity of the fingerprint hash of the adjacent video frame images. After step, a preliminary key frame set is obtained, and the obtained set is clustered frame by frame; wherein the frame-by-frame clustering method of the set is: taking the first frame in the set as the initial clustering center, and each frame in the set is similar to the current clustering center; if the maximum similarity between the two frames is less than a given threshold t, they are classified into a class, otherwise the frame is too far away from the current clustering center and forms a class by itself, and the clustering process is repeated until all key frames in the set are taken.

[0026] Further, in step four, after clustering is completed, the frame closest to the clustering center is taken into the final key frame result set to obtain the final key frame result set.

[0027] Another object of the present application is to provide a monitoring video key frame extraction system applying the monitoring video key frame extraction method, which comprises:

[0028] The data acquisition and decomposition module is used for acquiring the original video stream data of the target person and performing frame decomposition on the original video stream data to obtain a complete image frame set of the original video stream.

[0029] The image frame set sampling module is used for sampling the original video stream data by using the background difference based on the Gaussian mixture model and the adaptive mean sampling algorithm of the IBP frame.

[0030] The image frame filtering and clustering module is used for performing similarity filtering on the obtained sampling result set by using the image hash fingerprint filtering algorithm, and performing adaptive clustering on the filtered image frame result set.

[0031] The key frame extraction module is used for collecting the adaptive clustered result, taking the frame closest to the clustering center into the final key frame result set to obtain the final key frame result set.

[0032] Another object of the present application is to provide a computer device, which comprises a memory and a processor, the memory stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the monitoring video key frame extraction method.

[0033] Another object of the present application is to provide a computer readable storage medium, which stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the monitoring video key frame extraction method.

[0034] Another object of the present application is to provide an information data processing terminal for implementing the monitoring video key frame extraction system.

[0035] In combination with the above technical solutions and the solved technical problems, the technical solution to be protected by the present application has the following advantages and positive effects:

[0036] In order to better utilize the memory space of the storage medium and enable the user to better and quickly browse the general content of the original video stream, the present application screens out similar frames, redundant frames and fuzzy frames in the original video stream through a key frame screening algorithm to form a video summary storage of a section of original video, thereby greatly reducing the occupied storage space. The number of video frames contained in the original video stream obtained from the imaging device is very large. For example, the frame rate of a Hikvision camera is generally about 30 to 70, and if a video of 10 seconds is shot, the video image frames extracted from the video are about 300 to 700 frames. If all these frame images are processed, the calculation amount is very large, and a large amount of time is required for processing, which seriously affects the overall efficiency of the video defogging system. Therefore, the video image frames in the original video stream need to be screened to reduce the overall workload of the video defogging system and thereby improve the efficiency. The key frame extraction method of the present application first separates the original video into image frames, then samples the original image frame set, uses an IBP key frame-based adaptive mean sampling method to complete the sampling work, then filters the image frame set obtained by sampling, uses an image histogram to complete the filtering, and obtains a preliminary image key frame set after the filtering. Then, adaptive clustering is performed, and after the clustering is completed, the frame closest to the cluster center in each cluster is calculated as a key frame and is added to the final key frame set to form a video summary of the original video.

[0037] The present application provides a monitoring video key frame real-time extraction method based on inter-frame difference clustering analysis in the case of foggy monitoring video blurring, and realizes the key frame screening algorithm of the original video stream through sampling, filtering and clustering. The present application obtains the final key frame set through data acquisition, framing, sampling, similarity filtering and adaptive clustering. Framing, sampling and similarity filtering are the preprocessing of data samples, which can greatly reduce the input data of clustering and improve the processing efficiency of the entire algorithm, so as to cover all key frame set data sets and make the results more true and reliable.

[0038] At the same time, the algorithm based on the combination of compression domain and non-compression domain for sampling is first proposed, which realizes more sampling in the segment with high interest value weight and less sampling in the segment with low weight.

[0039] The present application fills the gap in the case of foggy weather video monitoring picture blurring, and provides a fast and effective method for extracting video key frames.

[0040] Some of the methods currently exist are based on multi-modal to complete the extraction of video key frame, and in the outdoor only single gun type camera, can't very good extraction of video key frame. The paper innovatively proposes based on compression domain and non-compression domain to sample the original video data, greatly reduces the number of processing image frame. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly introduced as follows. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can also be obtained according to these drawings without creative labor for those skilled in the art.

[0042] Figure 1 It is the monitoring video key frame extraction method flow chart provided by the embodiment of the present application;

[0043] Figure 2 It is the whole flow chart of the monitoring video key frame extraction method provided by the embodiment of the present application;

[0044] Figure 3 It is the flow chart of the I frame obtained by FFmpeg provided by the embodiment of the present application;

[0045] Figure 4 It is the key frame sampling algorithm flow chart based on GMM provided by the embodiment of the present application;

[0046] Figure 5 It is the algorithm flow chart of adaptive clustering provided by the embodiment of the present application;

[0047] Figure 6 It is the principle diagram of the monitoring video key frame extraction method provided by the embodiment of the present application.

[0048] Figure 7 It is the sampling process schematic diagram based on GMM provided by the embodiment of the present application;

[0049] Figure 8 It is the sampling process schematic diagram based on GMM provided by the embodiment of the present application;

[0050] Figure 9 It is the key frame extraction result diagram provided by the embodiment of the present application;

[0051] Figure 10 It is the comparison result diagram with other methods provided by the embodiment of the present application;

[0052] Figure 11 It is the process schematic diagram of the I frame detected by FFmpeg provided by the embodiment of the present application;

[0053] Figure 12 is a schematic diagram of the extracted key frame set provided by the embodiment of the present application;

[0054] Figure 13 is a schematic diagram of the correspondence between the original frame index and the result frame provided by the embodiment of the present application. DETAILED DESCRIPTION

[0055] In order to make the objectives, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and should not be used to limit the present application.

[0056] In view of the problems in the prior art, the present application provides a monitoring video key frame extraction method, system, medium, device and terminal, which are described in detail below with reference to the accompanying drawings.

[0057] As shown in Figure 1 , the monitoring video key frame extraction method provided by the embodiment of the present application includes the following steps:

[0058] S101, collecting original video stream data;

[0059] S102, decomposing the original video stream data into an image frame set;

[0060] S103, sampling the decomposed image frame set;

[0061] S104, filtering the sampled image frame result set;

[0062] S105, adaptively clustering the filtered image frame set;

[0063] S106, collecting the adaptively clustered result to form a video summary.

[0064] As a preferred embodiment, as shown in Figures 2-6 , the monitoring video key frame extraction method provided by the embodiment of the present application specifically includes the following steps:

[0065] First step: collecting original video stream data of a target person.

[0066] Second step: frame splitting the original video stream data to obtain a complete image frame set of the original video stream.

[0067] Third step: the sampling algorithm is based on the background difference change and IBP frame adaptive sampling algorithm based on Gaussian mixture model, the algorithm is divided into two branches to make decisions, the first branch divides the image frame set of the entire video stream into multiple GOP (Group of picture) according to the key degree of interframe information, each GOP starts with a key frame (I frame) and takes it as the interval between GOPs, and each GOP is composed of three structure frames of key frame (I frame), front and rear reference frame (B frame) and forward reference frame (P frame) (if the I frame is broken, the entire GOP structure is also broken); the second branch divides the video into an image frame set, first selects the first frame image, extracts a frame image after an interval of k frames, then calculates the background difference change of the two extracted frames to obtain the proportion r of the area of the changed region to the entire frame image; then the information of the two branches is fused, the higher the corresponding weight value of the key frame segment selected by the two branches, the higher the interest value of the corresponding image frame set, after the conversion and statistics of the entire video frame, different w values are set according to the interest value, that is, the weight, the higher the interest value, the higher the corresponding w value, so that the number of the image frames of interest obtained by sampling is more.

[0068] Fourth step: after the sampling is completed, a sampling result set of the original video is obtained, and then the filtering operation of similar frames and redundant frames is performed in the next step.

[0069] Fifth step: after the above steps, a preliminary key frame set is obtained, and then the obtained set needs to be clustered frame by frame, so that the entire key frame set data set can be covered to make the result more real and reliable. First, the first frame in the set is taken as the initial clustering center, then each frame in the set is measured for similarity with the current clustering center, if the maximum similarity between the two frames is less than a given threshold t, they are classified into a class, otherwise the frame is too far away from the current clustering center and forms a class by itself, and the process is repeated until all key frames in the set are taken.

[0070] Sixth step: after clustering, the frame closest to the clustering center is taken into the final key frame result set to obtain the final key frame result set.

[0071] The monitoring video key frame extraction system provided by the embodiment of the application comprises:

[0072] A data acquisition and decomposition module is configured to acquire original video stream data of a target person, frame the original video stream data, and obtain a complete image frame set of the original video stream;

[0073] An image frame set sampling module is configured to sample the original video stream data by using a background difference based on a Gaussian mixture model and an adaptive mean sampling algorithm of IBP frames;

[0074] An image frame filtering and clustering module is configured to perform similarity filtering on the obtained sampling result set by using an image hash fingerprint filtering algorithm, and to perform adaptive clustering on the filtered image frame result set;

[0075] A key frame extraction module is configured to collect the adaptive clustering result, to take each frame closest to the clustering center to join a final key frame result set, and to obtain a final key frame result set.

[0076] In order to prove the creativity and technical value of the technical scheme of the present application, this part is an application embodiment of the technical scheme of the claim on a specific product or related technology.

[0077] The method proposed in this paper has been tested on the West Circuit and some streets of the West Point Qingdao Research Institute, thereby verifying the feasibility and effectiveness of the method. Some results in the experimental process are as follows:

[0078] Figure 7 is a sampling process schematic diagram based on GMM of the present application;

[0079] Figure 8 is a sampling process schematic diagram based on GMM of the present application;

[0080] Figure 9 is a key frame extraction result diagram of the present application;

[0081] Figure 10 is a comparison result diagram of the present application with other methods;

[0082] Figure 11 is a process schematic diagram of FFmpeg detecting I frames of the present application;

[0083] Figure 12 is a schematic diagram of the extracted key frame set of the present application;

[0084] Figure 13 is a schematic diagram of the corresponding relationship between the original frame index and the result frame provided by the embodiment of the present application.

[0085] It should be noted that embodiments of the present application can be realized by hardware, software, or a combination of software and hardware. The hardware portion can be realized by a special logic; the software portion can be stored in a memory and executed by a proper instruction execution system, such as a microprocessor or a specially designed hardware. A person of ordinary skill in the art can understand that the above-mentioned apparatus and method can be realized by computer executable instructions and / or included in processor control codes, such as a carrier medium, such as a magnetic disk, CD or DVD-ROM, a programmable memory, such as a read-only memory (firmware), or a data carrier, such as an optical or electronic signal carrier. The apparatus of the present application and its modules can be realized by a hardware circuit, such as a very large scale integrated circuit or a gate array, a semiconductor, such as a logic chip, a transistor, or a programmable hardware device, such as a field programmable gate array, a programmable logic device, or the like, by software executed by various types of processors, or by a combination of the above-mentioned hardware circuit and software, such as firmware.

[0086] The above description is merely a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any modification, equivalent replacement, and improvement within the technical range disclosed by the present application, and within the spirit and principle of the present application, should be covered within the protection scope of the present application.

Claims

1. A method for monitoring video key frame extraction, characterized in that, The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps:

2. The method of claim 1, wherein, The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps:

3. The method of claim 1, wherein, The monitoring video key frame extraction method comprises the following steps:

4. The method of claim 1, wherein, The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises the following steps: The monitoring video key frame extraction method comprises The set frame-by-frame clustering method is as follows: taking the first frame in the set as an initial clustering center, performing similarity measurement between each frame in the set and the current clustering center; if the maximum similarity between two frames is less than a given threshold t, the two frames are classified into one class, otherwise the frame is too far away from the current clustering center and forms a class by itself, and the clustering process is repeated until all key frame sets in the set are taken. In step four, after clustering is completed, the frame closest to the clustering center is taken to the final key frame result set to obtain the final key frame result set.

5. A surveillance video key frame extraction system applying the surveillance video key frame extraction method according to any one of claims 1 to 4, characterized by, The monitoring video key frame extraction system comprises: A data acquisition decomposition module is configured to acquire original video stream data of a target person and frame the original video stream data to obtain a complete image frame set of the original video stream data. An image frame set sampling module is configured to sample the original video stream data by using a background difference based on a Gaussian mixture model and an IBP frame adaptive mean sampling algorithm. An image frame filtering and clustering module is configured to perform similarity filtering on the obtained sampling result set by using an image hash fingerprint filtering algorithm and to perform adaptive clustering on the filtered image frame result set. A key frame extraction module is configured to collect the adaptive clustering result, take the frame closest to the clustering center to the final key frame result set, and obtain the final key frame result set.

6. A computer device, comprising: The computer device comprises a memory and a processor, the memory stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the monitoring video key frame extraction method according to any one of claims 1 to 4.

7. A computer readable storage medium storing a computer program, the computer program being executed by a processor to make the processor execute the steps of the monitoring video key frame extraction method according to any one of claims 1 to 4.

8. An information data processing terminal, characterized by The information data processing terminal is configured to implement the monitoring video key frame extraction system according to claim 5.

Images