Low-quality video evaluation method, system and device and storage medium

Through the combination of Gaussian contrast method and integral graph algorithm, the continuous video image quality collected by the aircraft is evaluated, which solves the problem that traditional methods cannot evaluate low-quality videos, and realizes efficient quality evaluation and redundant deletion of continuous videos.

CN118941520APending Publication Date: 2024-11-12WUHAN XINGHUAN HENGYU INFORMATION TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410976323.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Traditional image quality evaluation methods cannot be applied to continuous video images and cannot effectively evaluate low-quality video in the video data collected by the aircraft, resulting in storage and transmission redundancy issues.

Method used

The Gaussian contrast method is used to conduct preliminary quality evaluation on each frame of the video data, low-quality images are selected, and feature values ​​are calculated using the integral graph algorithm to evaluate the video quality based on the proportion of frames of low-quality images, so as to achieve quality evaluation of continuous videos .

Benefits of technology

Effectively identify and delete low-quality videos, reduce redundancy in storage and transmission, and are suitable for the quality evaluation of continuous video images without the need for reference images.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118941520A_ABST
    Figure CN118941520A_ABST
Patent Text Reader

Abstract

The invention provides a low-quality video evaluation method. The low-quality video evaluation method comprises the following steps: S1, acquiring collected video data; s2, evaluating the quality of each frame of image in the video data by using a Gaussian contrast method; s3, calculating a characteristic value of the current frame image by using an integrogram algorithm, further evaluating the quality of the current frame image according to the characteristic value, and if the current frame image is judged to be a low-quality image, updating the frame number of the low-quality image; s4, judging whether the current frame image is the last frame image of the video or not, and if yes, entering the step S5; otherwise, importing the next frame of image, and returning to the step S2; and S5, evaluating whether the video is a low-quality video or not according to the proportion of the frame number of the low-quality image in the total frame number of the video. According to the quality evaluation method, reference images are not needed, the method can be suitable for quality evaluation of continuous video images, low-quality videos can be effectively deleted, and the redundancy problem brought by the low-quality videos to storage and transmission is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and in particular to a low-quality video evaluation method, system, device and storage medium. Background Art

[0002] The aircraft (manned / unmanned aircraft) conducts reconnaissance of the ground environment during flight. The reconnaissance process requires video recording. For the convenience of storage and transmission, the video will be stored and transmitted in units of 2 minutes.

[0003] Since the amount of collected video data is large, the required storage space is also large. During the flight of the aircraft, affected by various factors such as the environment, the collected video image data contains many unclear images, which are invalid images. When a video contains a large proportion of invalid images, the video is usually considered to be low-quality video. Low-quality videos will occupy a large storage space and transmission channel.

[0004] Therefore, after acquiring the video data, its quality needs to be evaluated so as to eliminate the low-quality video data and retain the useful video data. Video quality assessment is ultimately to evaluate the quality of each frame of the image. When evaluating the quality of an image, traditional image quality assessment methods generally use a reference image to compare and analyze the image to be evaluated with the reference image to evaluate the image quality. However, the video images collected by the aircraft are continuous and there is no reference image for comparison. Therefore, traditional image quality assessment methods are not applicable. Summary of the invention

[0005] The present invention provides a low-quality video evaluation method, system, device and storage medium, which solve the problem that traditional image quality evaluation methods cannot perform quality evaluation on continuous video images.

[0006] The technical solution of the present invention is achieved in this way:

[0007] A first aspect of the present invention provides a low-quality video evaluation method, comprising the following steps:

[0008] S1, obtaining collected video data;

[0009] S2, using the Gaussian contrast method to evaluate the quality of each frame of the video data, if the current frame is determined to be a low-quality image, then update the frame number of the low-quality image and proceed to step S4; otherwise, proceed to step S3;

[0010] S3, using the integral graph algorithm to calculate the characteristic value of the current frame image, and further evaluating the quality of the current frame image according to the characteristic value. If it is judged to be a low-quality image, the frame number of the low-quality image is updated, and the process proceeds to step S4; otherwise, the process proceeds directly to step S4;

[0011] S4, determine whether the current frame image is the last frame image of the video, if so, proceed to step S5; otherwise, import the next frame image and return to step S2;

[0012] S5, evaluating whether the video is a low-quality video according to the proportion of the number of low-quality image frames in the total number of video frames.

[0013] The present invention uses the Gaussian contrast method to preliminarily evaluate the quality of video images, preliminarily screens out low-quality images, and then uses the integral graph algorithm to further evaluate the quality of the remaining images. Finally, the quality of the video is evaluated according to the proportion of the number of frames of low-quality images in the total number of video frames, thereby effectively deleting low-quality videos and solving the redundancy problem of storage and transmission caused by low-quality videos. The quality evaluation method of the present invention does not require the use of reference images and can be applicable to the quality evaluation of continuous video images.

[0014] Specifically, in step S1, after acquiring the collected video data, the video data is segmented into 2-minute segments to obtain a plurality of segmented videos for subsequent quality evaluation.

[0015] Specifically, in step S2, the method of using the Gaussian contrast method to evaluate the image quality includes the following:

[0016] The Gaussian contrast method is used to determine the texture information in the image. The formula is as follows:

[0017] diffGuass=abs(guass(image)-image)

[0018] Where, image represents the original image, guass(image) represents the Gaussian blurred image of the original image, and diffGuass represents the absolute value of the grayscale difference between the Gaussian blurred image and the original image;

[0019] The image after Gaussian contrast is divided into 9 regions of 3×3, and the variance of the 9 regions is calculated respectively;

[0020] Sort the variances of the 9 regions from small to large, and take the average of the 3 largest variance values ​​as the variance of the image;

[0021] The variance of the image is compared with a reference threshold T0. If the variance of the image is less than the reference threshold T0, the image is considered to be a low-quality image.

[0022] Specifically, in step S3, the method for calculating the feature value of the current frame image is as follows:

[0023] Calculate the integral image of the current frame image;

[0024] With width / 3 and height / 3 as the step size, traverse the entire integral graph and get the rectangular area as follows:

[0025]

[0026] Wherein, rect(i) = {x, y, w, h} represents the i-th rectangular area, (x, y) represents the coordinates of the starting point of the i-th rectangular area; w, h represent the width and height of the i-th rectangular area respectively; width, height represent the width and height of the image respectively;

[0027] Calculate the mean and variance of the grayscale values ​​in each rectangular area respectively, normalize the mean and variance of the grayscale values ​​in each rectangular area, and synthesize the feature parameters:

[0028]

[0029] Among them, Mean rect(i) 、Var rect(i) Respectively represent the mean and variance of the i-th rectangular area, Feat rect(i) Represents the characteristic parameters of the i-th rectangular area;

[0030] The feature parameters of the nine rectangular areas are sorted in ascending order, and the average value of the three largest feature parameters is taken as the feature value of the current frame image.

[0031] Specifically, if the characteristic value is less than the reference threshold T1, the current frame image is determined to be a low-quality image.

[0032] Specifically, the method for evaluating whether a video is a low-quality video is:

[0033] Calculate the ratio of low-quality image frames to the total number of video frames:

[0034] ratio = count / N

[0035] Where count is the number of frames of the low-quality image, and N is the total number of frames of the video;

[0036] If the ratio is greater than the reference threshold T2, the current video is determined to be a low-quality video.

[0037] A second aspect of the present invention provides a low-quality video assessment system, comprising:

[0038] Video acquisition module, used for acquiring video data;

[0039] The first image quality assessment module uses a Gaussian contrast method to assess the quality of each frame of the video data;

[0040] The second image quality assessment module calculates the characteristic value of the current frame image by using the integral graph algorithm, and further assesses the quality of the current frame image according to the characteristic value;

[0041] A counting module, used to update the frame number of low-quality images;

[0042] The judgment module evaluates whether the video is of low quality based on the proportion of the number of low quality image frames in the total number of video frames.

[0043] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the above-mentioned low-quality video assessment method when executing the computer program.

[0044] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned low-quality video assessment method are implemented. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 The figure is a flow chart of a low-quality video evaluation method of the present invention. DETAILED DESCRIPTION

[0047] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only 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 ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] Reference Figure 1 A first aspect of an embodiment of the present invention provides a low-quality video evaluation method, comprising the following steps:

[0049] S1, obtaining collected video data;

[0050] S2, using the Gaussian contrast method to evaluate the quality of each frame of the video data, if the current frame is determined to be a low-quality image, then update the frame number of the low-quality image and proceed to step S4; otherwise, proceed to step S3;

[0051] S3, using the integral graph algorithm to calculate the characteristic value of the current frame image, and further evaluating the quality of the current frame image according to the characteristic value. If it is judged to be a low-quality image, the frame number of the low-quality image is updated, and the process proceeds to step S4; otherwise, the process proceeds directly to step S4;

[0052] S4, determine whether the current frame image is the last frame image of the video, if so, proceed to step S5; otherwise, import the next frame image and return to step S2;

[0053] S5, evaluating whether the video is a low-quality video according to the proportion of the number of low-quality image frames in the total number of video frames.

[0054] The present invention uses the Gaussian contrast method to preliminarily evaluate the quality of video images, preliminarily screens out low-quality images, and then uses the integral graph algorithm to further evaluate the quality of the remaining images. Finally, the quality of the video is evaluated according to the proportion of the number of frames of low-quality images in the total number of video frames, thereby effectively deleting low-quality videos and solving the redundancy problem of storage and transmission caused by low-quality videos. The quality evaluation method of the present invention does not require the use of reference images and can be applicable to the quality evaluation of continuous video images.

[0055] Specifically, in step S1, after acquiring the collected video data, the video data is segmented into 2-minute segments to obtain a plurality of segmented videos for subsequent quality evaluation.

[0056] During the specific implementation process, the aircraft (manned / unmanned aircraft) will install a photoelectric ball at the pod position to collect infrared or color video data. After one set of video image data is encoded and compressed, a video recording is generated at a 2-minute time interval, and another set of data is analyzed by an algorithm. From the beginning of video data collection, the present invention begins to count whether the video segment is a low-quality video to determine whether the video recording is retained.

[0057] Specifically, in step S2, the method of using the Gaussian contrast method to evaluate the image quality includes the following:

[0058] The Gaussian contrast method is used to determine the texture information in the image. The formula is as follows:

[0059] diffGuass=abs(guass(image)-image)

[0060] Where image represents the original image, guass(image) represents the Gaussian blurred image of the original image (5×5 Gaussian kernel), and diffGuass represents the absolute value of the grayscale difference between the Gaussian blurred image and the original image.

[0061] The image after Gaussian contrast is divided into 9 regions of 3×3 to obtain 9-grid image blocks, and the variance of the grayscale values ​​of the 9 regions is calculated respectively;

[0062] Sort the variances of the 9 regions from small to large, and take the average of the 3 largest variance values ​​as the variance of the image;

[0063] The variance of the image is compared with the reference threshold T0. If the variance of the image is less than the reference threshold T0, the image is considered to be a low-quality image; if the variance of the image is greater than or equal to T0, the analysis continues.

[0064] Specifically, in step S3, the method for calculating the feature value of the current frame image is as follows:

[0065] Calculate the integral image of the current frame image;

[0066] Integral image is a method to quickly calculate the sum of rectangular areas in an image. The main advantage of this algorithm is that once the integral image is calculated first, we can calculate the sum of rectangular areas of any size in the image in constant time.

[0067] By using the above characteristics of the integral image, the mean and variance of each rectangular area can be quickly obtained. For the convenience of subsequent steps, the integral image of the video image is first calculated. This calculation method is a conventional method.

[0068] With width / 3 and height / 3 as the step size, traverse the entire integral graph and get the rectangular area as follows:

[0069]

[0070]

[0071] Wherein, rect(i) = {x, y, w, h} represents the i-th rectangular area, (x, y) represents the coordinates of the starting point of the i-th rectangular area (the lower left vertex of the rectangular area is defined as the starting point); w and h represent the width and height of the i-th rectangular area respectively; width and height represent the width and height of the image respectively;

[0072] After knowing the starting point and width and height data of the above 9 rectangular areas, the coordinates of the 4 vertices of the rectangular area can be obtained. Using the calculation principle of the integral graph, the mean and variance of each area can be obtained;

[0073] Calculate the mean and variance of the grayscale values ​​in each rectangular area respectively, normalize the mean and variance of the grayscale values ​​in each rectangular area, and synthesize the feature parameters:

[0074]

[0075] Among them, Mean rect(i) 、Var rect(i) Respectively represent the mean and variance of the i-th rectangular area, Feat rect(i) Represents the characteristic parameters of the i-th rectangular area;

[0076] The feature parameters of the nine rectangular areas are sorted in ascending order, and the average value of the three largest feature parameters is taken as the feature value of the current frame image.

[0077] If the characteristic value is less than the reference threshold T1, the current frame image is judged to be a low-quality image.

[0078] Specifically, the method for evaluating whether a video is a low-quality video is:

[0079] Calculate the ratio of low-quality image frames to the total number of video frames:

[0080] ratio = count / N

[0081] Where count is the number of frames of the low-quality image, and N is the total number of frames of the video;

[0082] If the ratio is greater than the reference threshold T2, the current video is judged to be a low-quality video and the current video recording needs to be deleted; if the ratio is less than the reference threshold T2, the current video is retained and the quality assessment of the next video continues.

[0083] A second aspect of an embodiment of the present invention provides a low-quality video assessment system, including:

[0084] Video acquisition module, used for acquiring video data;

[0085] The first image quality assessment module uses a Gaussian contrast method to assess the quality of each frame of the video data;

[0086] The second image quality assessment module calculates the characteristic value of the current frame image by using the integral graph algorithm, and further assesses the quality of the current frame image according to the characteristic value;

[0087] A counting module, used to update the frame number of low-quality images;

[0088] The judgment module evaluates whether the video is of low quality based on the proportion of the number of low quality image frames in the total number of video frames.

[0089] A third aspect of an embodiment of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and the processor implements the steps of the above-mentioned low-quality video assessment method when executing the computer program.

[0090] A fourth aspect of the embodiments of the present invention provides a computer-readable storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the above-mentioned low-quality video assessment method are implemented.

[0091] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A low-quality video evaluation method, It is characterized in that The following steps are involved: S1, obtaining collected video data; S2, using the Gaussian contrast method to evaluate the quality of each frame of the video data, if the current frame is determined to be a low-quality image, then updating the frame number of the low-quality image, and entering step S4; Otherwise, proceed to step S3; S3, using the integral graph algorithm to calculate the characteristic value of the current frame image, and further evaluating the quality of the current frame image according to the characteristic value. If it is judged to be a low-quality image, the frame number of the low-quality image is updated, and the process proceeds to step S4; otherwise, the process proceeds directly to step S4; S4, determine whether the current frame image is the last frame image of the video, if so, proceed to step S5; otherwise, import the next frame image and return to step S2; S5, evaluating whether the video is a low-quality video according to the proportion of the number of low-quality image frames in the total number of video frames.

2. A low-quality video assessment method as claimed in claim 1, It is characterized in that In step S1, after acquiring the collected video data, the video data is segmented into 2-minute segments to obtain a plurality of segmented videos for subsequent quality evaluation.

3. A low-quality video assessment method as claimed in claim 1, It is characterized in that In step S2, the method of using the Gaussian contrast method to evaluate the image quality includes the following: The Gaussian contrast method is used to determine the texture information in the image. The formula is as follows: diffGuass=abs(guass(image)-image) Where, image represents the original image, guass(image) represents the Gaussian blurred image of the original image, and diffGuass represents the absolute value of the grayscale difference between the Gaussian blurred image and the original image; The image after Gaussian contrast is divided into 9 regions of 3×3, and the variance of the 9 regions is calculated respectively; Sort the variances of the 9 regions from small to large, and take the average of the 3 largest variance values ​​as the variance of the image; The variance of the image is compared with a reference threshold T0. If the variance of the image is less than the reference threshold T0, the image is considered to be a low-quality image.

4. A low-quality video assessment method as claimed in claim 1, It is characterized in that In step S3, the method for calculating the feature value of the current frame image is as follows: Calculate the integral image of the current frame image; With width / 3 and height / 3 as the step size, traverse the entire integral graph and get the rectangular area as follows: Wherein, rect(i) = {x, y, w, h} represents the i-th rectangular area, (x, y) represents the coordinates of the starting point of the i-th rectangular area; w, h represent the width and height of the i-th rectangular area respectively; width, height represent the width and height of the image respectively; Calculate the mean and variance of the grayscale values ​​in each rectangular area respectively, normalize the mean and variance of the grayscale values ​​in each rectangular area, and synthesize the feature parameters: Among them, Mean rect(i) 、Var rect(i) Respectively represent the mean and variance of the i-th rectangular area, Feat rect(i) Represents the characteristic parameters of the i-th rectangular area; The feature parameters of the nine rectangular areas are sorted in ascending order, and the average value of the three largest feature parameters is taken as the feature value of the current frame image.

5. A low-quality video assessment method as claimed in claim 1, It is characterized in that If the characteristic value is less than the reference threshold T1, the current frame image is judged to be a low-quality image.

6. A low-quality video assessment method as claimed in claim 1, It is characterized in that The method to evaluate whether a video is low quality is: Calculate the ratio of low-quality image frames to the total number of video frames: ratio = count / N Where count is the number of frames of the low-quality image, and N is the total number of frames of the video; If the ratio is greater than the reference threshold T2, the current video is determined to be a low-quality video.

7. A low-quality video assessment system, based on the low-quality video assessment method according to any one of claims 1 to 6, It is characterized in that include: Video acquisition module, used for acquiring video data; The first image quality assessment module uses a Gaussian contrast method to assess the quality of each frame of the video data; The second image quality assessment module calculates the characteristic value of the current frame image by using the integral graph algorithm, and further assesses the quality of the current frame image according to the characteristic value; A counting module, used to update the frame number of low-quality images; The judgment module evaluates whether the video is of low quality based on the proportion of the number of low quality image frames in the total number of video frames.

8. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, It is characterized in that When the processor executes the computer program, the steps of the low-quality video assessment method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program. It is characterized in that When the computer program is executed by a processor, the steps of the low-quality video assessment method according to any one of claims 1 to 6 are implemented.