Video-based full-link monitoring method and system

By extracting feature points and calculating the impact degree of the full-link monitoring video of the power grid, and optimizing the compression processing, the problem of waste of storage space and reduction of access rate under the influence of environmental changes in the existing technology is solved, and more efficient video data compression is achieved and monitoring analysis.

CN118509561BActive Publication Date: 2024-11-12INFORMATION & COMMNUNICATION BRANCH STATE GRID JIANGXI ELECTRIC POWER CO
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410955989.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-17
Publication Date
2024-11-12
Estimated Expiration
2044-07-17

AI Technical Summary

Technical Problem

When the prior art compresses the full-link monitoring video of the power grid, it is difficult to effectively deal with the impact of environmental changes, resulting in wasted storage space and a decrease in access rate.

Method used

By performing preset sampling intervals on the full-link monitoring video of the power grid, multi-frame images are extracted, grayscale processing and feature point extraction, feature point sequences are constructed, timing and range influence degree, and the need to compress the data sequence to be compressed, And smoothing is performed according to the necessity threshold to optimize the compression result.

Benefits of technology

It improves the compression efficiency of video frame images, improves the video data access efficiency and abnormal analysis capabilities of the full-link monitoring system, and enhances the equipment fault monitoring efficiency.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118509561B_ABST
    Figure CN118509561B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of image data processing, and more specifically, the present invention relates to a video-based full-link monitoring method and system. The method includes obtaining a grayscale image to be tested; extracting feature points from any grayscale image to be tested, obtaining a feature point set, and calculating the timing influence of the target feature point based on the constructed target feature sequence; calculating the range influence of the target feature point based on the timing influence of all feature points within the neighborhood of the preset target feature point; obtaining a data sequence to be compressed based on the grayscale image to be tested, calculating the compression necessity of sample data points in the data sequence to be compressed, so as to construct a compression necessity sequence of the data sequence to be compressed; obtaining an optimized compression result, and completing intelligent monitoring of the full-link status. Through the technical solution of the present invention, the efficiency of in-video device fault monitoring of the full-link monitoring system can be effectively improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention generally relates to the field of image data processing and more specifically to a video-based full-link monitoring method and system. Background Art

[0002] The full-link monitoring of multi-level power grid systems can realize remote monitoring and intelligent operation and maintenance of power grids, and improve the reliability, safety and economy of power grid operation. Among them, video monitoring of substations, transmission towers, switchgear, important transmission lines, etc. can detect fires, damage or other safety issues in a timely manner. However, the amount of data generated in the full-link monitoring video of the power grid is very large, and its data storage requires more storage resources. Therefore, it is necessary to realize the compressed storage of the acquired full-link monitoring video frame images to improve the video data access efficiency of the full-link monitoring system and realize the rapid response and early warning processing of abnormal conditions in the full-link monitoring.

[0003] The existing technology uses The compression algorithm compresses the full-link monitoring video of the power grid, which can effectively save storage resources, improve the access rate of video data, and facilitate the rapid analysis of monitoring video anomalies of the full-link monitoring system; however, In the process of compression processing of the full-link monitoring video of the power grid by the compression algorithm, its compression efficiency depends on the repetitive characteristics of the buffer data in the sliding window. Only the repetitive compression processing within the buffer of the video frame image data sequence is adopted. It is difficult to cope with the impact of environmental changes in the video frame image, which often leads to a waste of storage space and a reduction in access rate. Summary of the invention

[0004] To solve one or more of the above technical problems, the present invention proposes a video-based full-link monitoring method and system. To this end, the present invention provides solutions in the following aspects.

[0005] In a first aspect, the present invention discloses a video-based full-link monitoring method, comprising: extracting multiple frames of images from a power grid full-link monitoring video at a preset sampling interval, graying the multiple frames of images to obtain a gray image to be tested, and constructing a set of gray images to be tested within a preset period; extracting feature points from any of the gray images to be tested to obtain a feature point set, and taking any feature point in the feature point set as a target feature point; constructing a target feature sequence of the target feature point at the same position of each of the gray images to be tested in the gray image set to be tested; and calculating the time of the target feature point according to the target feature sequence. The method comprises the following steps: calculating the degree of range influence of the target feature point according to the degree of temporal influence of all feature points in the neighborhood of the target feature point; obtaining a data sequence to be compressed based on the grayscale image to be tested; calculating the necessity of compression of sample data points in the data sequence to be compressed based on the degree of range influence, so as to construct a compression necessity sequence of the data sequence to be compressed; comparing the compression necessity with a preset necessity threshold, traversing the compression necessity sequence, and smoothing the compression necessity greater than the necessity threshold to obtain an optimized compression result, and complete intelligent monitoring of the entire link status.

[0006] Preferably, constructing a target feature sequence of the target feature point at the same position of each of the grayscale images to be tested in the grayscale image set to be tested comprises the steps of: using The algorithm extracts feature points from any of the grayscale images to be tested to obtain a set of feature points; uses the edges of the grayscale images to be tested as the horizontal and vertical axes to construct a first quadrant coordinate system; marks the element value positions where feature points exist in the grayscale images to be tested as 1, and marks the element value positions where feature points do not exist as 0; traverses each of the grayscale images to be tested in the grayscale image set to obtain a target feature sequence of target feature points at the same position of each of the grayscale images to be tested in the grayscale image set.

[0007] Preferably, the range influence degree of the target feature point satisfies the relationship: in, Representing feature points The timing impact of represents the number of elements in the feature sequence, Indicates the first The value of the element, Representation and feature sequence The number of adjacent elements that have the same value.

[0008] Preferably, the range influence degree of the target feature point satisfies the relationship: ,in, Indicates The first feature point set Feature points The extent of the impact, Indicates The first feature point set Feature points The number of feature points in the neighborhood of Representing feature points The timing impact of Representing feature points In the neighborhood of Feature points The degree of timing impact.

[0009] Preferably, the step of obtaining the data sequence to be compressed comprises the following steps: treating any one of the grayscale images to be tested as a two-dimensional matrix of the same size as the grayscale image to be tested, wherein the element values ​​of the two-dimensional matrix are the grayscale values ​​of the corresponding positions of the image; connecting the grayscale values ​​on the column vectors of the two-dimensional matrix from left to right in sequence; and using the connected data as The data sequence to be compressed by the compression algorithm.

[0010] Preferably, the distance value of two feature points closest to the sample data point in the data sequence to be compressed is obtained, wherein the two feature points are respectively on both sides of the sample data point; multiple data points at the same position on multiple data sequences to be compressed within a preset time series neighborhood of the sample data point are obtained; and the compression necessity of the sample data point is calculated, and the compression necessity satisfies the polynomial:

[0011] ,

[0012] ,

[0013] in, Represents data points The influence degree of the neighborhood feature points, Represents data points The range of the feature point closest to the left side affects the degree of influence. Represents the data points in the data sequence to be compressed The distance value from the nearest feature point on the left, Represents data points The range of the feature point closest to the right side affects the degree of influence. Represents the data points in the data sequence to be compressed The distance value to the nearest feature point on the right side, Represents data points The necessity of compression, Represents data points The mean of the data points at the same position on multiple data sequences to be compressed within the preset time series neighborhood, Represents data points The average influence degree of neighborhood feature points of data points at the same position on multiple data sequences to be compressed within a preset time series neighborhood.

[0014] In a second aspect, the present invention discloses a video-based full-link monitoring system, comprising: a processor; and a memory, wherein the memory stores computer instructions, and when the computer instructions are executed by the processor, the system executes the above-mentioned video-based full-link monitoring method.

[0015] The present invention has the following technical effects:

[0016] It is possible to obtain the range of influence of feature points according to the changing characteristics of feature points in the video frame image of the time period, and thereby realize the acquisition of the necessity of compression of data points in the data sequence to be compressed extracted from the video frame image, thereby achieving further improvement in compression efficiency, and is more conducive to the rapid analysis of monitoring video anomalies of the full-link monitoring system, and improving the efficiency of device fault monitoring in the video of the full-link monitoring system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0018] Figure 1 It is a flow chart of a video-based full-link monitoring method according to an embodiment of the present invention.

[0019] Figure 2 is a schematic diagram of a video-based full-link monitoring system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0021] It should be understood that when the terms "first", "second", etc. are used in the claims, descriptions, and drawings of the present invention, they are only used to distinguish different objects, rather than to describe a specific order. The terms "include" and "comprise" used in the description and claims of the present invention indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their collections.

[0022] The present invention provides a full-link monitoring method based on video. Figure 1 As shown, the video-based full-link monitoring method includes steps S1 to S5, which are described in detail below.

[0023] S1, obtaining a grayscale image to be tested.

[0024] In one embodiment, the full-link monitoring video of the power grid obtained in a preset period extracts video frame images according to a preset sampling interval. For example, the preset period is one day, and the preset sampling frequency is once every 5 minutes. The extracted image is an RGB (redgreenblue) image. The RGB image is grayscaled to obtain a grayscale image to be tested, and then a sequence of grayscale images to be tested within the preset period is obtained.

[0025] Specifically, the video surveillance devices set up in the key monitoring areas of multiple stages such as the power generation stage and the transmission stage of the full-link monitoring system of the power grid are used to obtain the full-link monitoring video of the power grid, and the video frame image extraction is performed on the historical video surveillance video of a single location. Among them, the grayscale image to be tested in the monitoring video frame often has a small change feature of the background, which is greatly affected by environmental changes. Specifically, the image features of the equipment area such as the transmission line in the image tend to remain unchanged when it is operating normally, and its changes are mainly caused by equipment abnormalities and environmental influences. For example, the influence of icing on the transmission line, fire in the distribution box, and changes in the brightness of the environment, etc., leads to a large change in the grayscale features of the image, which reduces the repeatability of the acquired image data, and further affects the compression efficiency of the compression algorithm.

[0026] S2, extracting feature points from any grayscale image to be tested, obtaining a feature point set, and calculating the temporal influence of the target feature points according to the constructed target feature sequence.

[0027] In one embodiment, using The algorithm (Scale-Invariant Feature Transform) extracts feature points from any grayscale image to be tested, obtains a set of feature points, and constructs a first quadrant coordinate system using the edges of the grayscale image to be tested as the horizontal and vertical axes.

[0028] Mark the element value position where the feature point exists in the grayscale image to be tested as 1, and mark the element value position where the feature point does not exist as 0, traverse each grayscale image to be tested in the grayscale image set to obtain the target feature sequence of the target feature point at the same position of each grayscale image to be tested in the grayscale image set.

[0029] Calculate the timing influence of the target feature points, and the timing influence satisfies the relationship: in, Representing feature points The timing impact of represents the number of elements in the feature sequence, Indicates the first The value of the element, Indicates the first The number of adjacent elements that have the same value.

[0030] It should be noted that The acquisition method can use a preset sliding window to traverse the element values ​​of the feature sequence. When there are elements with the same element value, the number is increased by one until there are different elements, and then the counting is restarted. For example, the feature sequence is 1100111001111, and the size of the preset sliding window is 3. Then the value of the adjacent element with the same value as the 5th element in the feature sequence is is 3.

[0031] S3, calculating the range influence degree of the target feature point according to the temporal influence degrees of all feature points within the neighborhood range of the preset target feature point.

[0032] In one embodiment, for the local area where the target feature point is located, the neighborhood range of the target feature point is preset. Exemplarily, the neighborhood range of the present invention is 5×5. The range influence degree of the target feature point is calculated according to the temporal influence degree of all feature points within the neighborhood range of the preset target feature point. The range influence degree of the target feature point satisfies the relationship:

[0033] ,in, Indicates The first feature point set Feature points The extent of the impact, Indicates The first feature point set Feature points The number of feature points in the neighborhood of Representing feature points The timing impact of Representing feature points In the neighborhood of Feature points The degree of timing impact.

[0034] in, Refers to the feature point The degree of influence of the range is measured by the feature points The stability in time series, that is, the reflection of the characteristic points The probability that the corresponding area in the image belongs to the background area is large enough to facilitate the subsequent acquisition of feature points. Necessity of compression of local neighborhoods on the data sequence to be compressed.

[0035] S4, obtaining a data sequence to be compressed based on the grayscale image to be tested, and calculating the compression necessity of sample data points in the data sequence to be compressed to construct a compression necessity sequence of the data sequence to be compressed.

[0036] In one embodiment, any grayscale image to be tested is regarded as a two-dimensional matrix of the same size as the grayscale image to be tested, and the element value of the two-dimensional matrix is ​​the grayscale value of the corresponding position of the image; the grayscale values ​​on the column vectors of the two-dimensional matrix are connected from left to right in sequence; the connected data is used as The data sequence to be compressed of the compression algorithm. For the data sequence to be compressed corresponding to the grayscale images to be tested that are adjacent in time sequence, the more similar the distribution structure of the feature points on the data sequence to be compressed is, the more likely the grayscale changes between the feature points on the data sequence to be compressed are to be stable, the more likely the actual grayscale differences are to be affected by environmental factors, the more likely the grayscale changes of the data points on the data sequence to be compressed are to be considered as acceptable error losses, and the more necessary the data points are to be compressed.

[0037] Obtain the distance value of the two feature points closest to the sample data point in the data sequence to be compressed, wherein the two feature points are respectively on both sides of the sample data point. , the distance of the feature point closest to the sample data point on the left side of the sample data point in the data sequence to be compressed is recorded as , the distance of the feature point closest to the sample data point on the right side of the sample data point in the data sequence to be compressed is recorded as .

[0038] Obtain multiple data points at the same position on multiple data sequences to be compressed within a preset time series neighborhood of the sample data point. The greater the degree of change in the distance between these data points and the nearest feature points in the left and right neighborhoods, the smaller the range of the feature points, and the smaller the necessity of compression corresponding to the data points. Exemplarily, the preset time series neighborhood of the present invention is set to .

[0039] Calculate the compression necessity of the sample data points, which satisfies the polynomial:

[0040] ,

[0041] ,

[0042] in, Represents data points The influence degree of the neighborhood feature points, Represents data points The range of the feature point closest to the left side affects the degree of influence. Represents the data points in the data sequence to be compressed The distance value from the nearest feature point on the left, Represents data points The range of the feature point closest to the right side affects the degree of influence. Represents the data points in the data sequence to be compressed The distance value to the nearest feature point on the right side, Represents data points The necessity of compression, Represents data points The mean of the data points at the same position on multiple data sequences to be compressed within the preset time series neighborhood, Represents data points The average influence degree of neighborhood feature points of data points at the same position on multiple data sequences to be compressed within a preset time series neighborhood.

[0043] Traverse all data points to obtain the compression necessity sequence of the data sequence to be compressed.

[0044] S5, obtains optimized compression results and completes intelligent monitoring of the entire link status.

[0045] In one embodiment, the compression necessity is compared with a preset necessity threshold. For example, the present invention sets the necessity threshold to 0.86. The compression necessity sequence is traversed, and the compression necessity greater than the necessity threshold is smoothed, thereby completing the environmental influencing factors for The compression algorithm handles the computational redundancy of the data sequence. Specifically, the revolving door algorithm can be used to perform straight line fitting on all data points on the filtered data sequence to be compressed, and the data sequence after the straight line fitting can be Compression algorithm processing obtains optimized compression results and completes intelligent monitoring of the entire link status.

[0046] The embodiment of the present invention also discloses a full-link monitoring system based on video, referring to Figure 2 , including a processor and a memory, the memory storing computer program instructions, and when the computer program instructions are executed by the processor, the video-based full-link monitoring method according to the present invention is implemented.

[0047] The above system also includes other components well known to those skilled in the art, such as a communication bus and a communication interface, and their configuration and functions are known in the art, so they will not be described in detail here.

[0048] In the present invention, the aforementioned memory may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, apparatus or device. For example, a computer-readable storage medium may be any appropriate magnetic storage medium or magneto-optical storage medium, such as a resistive random access memory RRAM (Resistive Random Access Memory), a dynamic random access memory DRAM (Dynamic Random Access Memory), a static random access memory SRAM (Static Random-Access Memory), an enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), a high-bandwidth memory HBM (High-Bandwidth Memory), a hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application, a module or both. Any such computer storage medium may be part of a device or accessible or connectable to a device.

[0049] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

[0050] The above are all preferred embodiments of the present invention, and are not intended to limit the protection scope of the present invention. Therefore, any equivalent changes made based on the structure, shape, and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. Video-based full-link monitoring method, It is characterized in that include: Extract multiple frames of images from the full-link monitoring video of the power grid according to a preset sampling interval, perform grayscale processing on the multiple frames of images to obtain grayscale images to be tested, and construct a set of grayscale images to be tested within a preset period; Extracting feature points from any grayscale image to be tested to obtain a feature point set, and taking any feature point in the feature point set as a target feature point; Constructing a target feature sequence of the target feature point at the same position of each of the grayscale images to be tested in the grayscale image set to be tested; The temporal influence degree of the target feature point is calculated according to the target feature sequence, and the temporal influence degree satisfies the relationship: Among them, f(P x,y ) represents the feature point P x,y The degree of temporal influence, I represents the number of elements in the feature sequence, a i represents the value of the i-th element in the feature sequence, l i Represents the number of adjacent elements with the same value as the i-th element in the feature sequence; The range influence degree of the target feature point is calculated according to the temporal influence degree of all feature points within the preset neighborhood of the target feature point, and the range influence degree satisfies the relationship: Among them, AE j,r Represents the rth feature point P on the jth feature point set j,r The influence degree of the range, K represents the rth feature point P on the jth feature point set j,r The number of feature points in the neighborhood, f(P j,r ) represents the feature point P j,r The degree of timing influence, f(P j,r,k Represents the feature point P j,r The kth feature point P in the neighborhood of j,r,k The degree of timing impact; Obtaining a data sequence to be compressed based on the grayscale image to be tested; The compression necessity of the sample data points in the to-be-compressed data sequence is calculated based on the range influence degree to construct a compression necessity sequence of the to-be-compressed data sequence. Obtaining the compression necessity includes the steps of: Obtaining distance values ​​of two feature points closest to the sample data point in the to-be-compressed data sequence, wherein the two feature points are respectively located on both sides of the sample data point; Acquire multiple data points at the same position on multiple data sequences to be compressed within a preset time series neighborhood of the sample data point; Calculate the compression necessity of the sample data points, which satisfies the polynomial: Among them, s(Q m,n ) represents the data point Q m,n The influence degree of the neighborhood feature points, AE m,n, - represents data point Q m,n The influence degree of the range of the feature point closest to the left side, d m,n. - represents the data point Q in the data sequence to be compressed m,n The distance value of the nearest feature point on the left, AE m,n,+ Represents data point Q m,n The range influence degree of the feature point closest to the right side, d m,n,+ Represents the data point Q in the data sequence to be compressed m,n The distance value from the nearest feature point on the right, EN m,n Represents data point Q m,n The necessity of compression, Represents data point Q m,n The mean of the data points at the same position on multiple data sequences to be compressed within the preset time series neighborhood, Represents data point Q m,n The average influence degree of neighborhood feature points of data points at the same position on multiple data sequences to be compressed within a preset time series neighborhood; Compare the compression necessity with a preset necessity threshold, traverse the compression necessity sequence, and smooth the compression necessity greater than the necessity threshold to obtain an optimized compression result and complete intelligent monitoring of the entire link status.

2. The video-based full-link monitoring method according to claim 1, It is characterized in that Constructing a target feature sequence of the target feature point at the same position of each of the grayscale images to be tested in the grayscale image set to be tested comprises the steps of: Using the SIFT algorithm to extract feature points from any grayscale image to be tested, to obtain a feature point set; Using the edge of the grayscale image to be measured as the horizontal axis and the vertical axis to construct a first quadrant coordinate system; Marking the element value position where the feature point exists in the grayscale image to be measured as 1, and marking the element value position where the feature point does not exist as 0; Each of the grayscale images to be tested in the grayscale image set to be tested is traversed to obtain a target feature sequence of target feature points at the same position of each of the grayscale images to be tested in the grayscale image set to be tested.

3. The video-based full-link monitoring method according to claim 1, It is characterized in that The step of obtaining the data sequence to be compressed comprises: Treat any one of the grayscale images to be tested as a two-dimensional matrix of the same size as the grayscale image to be tested, wherein the element values ​​of the two-dimensional matrix are the grayscale values ​​of the corresponding positions of the image; Connect the grayscale values ​​on the column vectors of the two-dimensional matrix from left to right in sequence; The concatenated data is used as the data sequence to be compressed by the LZ77 compression algorithm.

4. Video-based full-link monitoring system, It is characterized in that include: processor; And a memory storing computer instructions for a video-based full-link monitoring method, wherein when the computer instructions are executed by the processor, the system executes the video-based full-link monitoring method according to any one of claims 1 to 3.