CN114205578A - Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value - Google Patents

Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value Download PDF

Info

Publication number
CN114205578A
CN114205578A CN202111392348.6A CN202111392348A CN114205578A CN 114205578 A CN114205578 A CN 114205578A CN 202111392348 A CN202111392348 A CN 202111392348A CN 114205578 A CN114205578 A CN 114205578A
Authority
CN
China
Prior art keywords
frequency domain
video
matrix
skewness
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111392348.6A
Other languages
Chinese (zh)
Inventor
詹仁辉
徐哲鑫
许智杰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fujian Supwit Group Co ltd
Original Assignee
Fujian Supwit Group Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Fujian Supwit Group Co ltd filed Critical Fujian Supwit Group Co ltd
Priority to CN202111392348.6A priority Critical patent/CN114205578A/en
Publication of CN114205578A publication Critical patent/CN114205578A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N17/00—Diagnosis, testing or measuring for television systems or their details
    • H04N17/004—Diagnosis, testing or measuring for television systems or their details for digital television systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G06T7/41—Analysis of texture based on statistical description of texture
    • G06T7/42—Analysis of texture based on statistical description of texture using transform domain methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10016—Video; Image sequence
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20048—Transform domain processing
    • G06T2207/20056—Discrete and fast Fourier transform, [DFT, FFT]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/30—Subject of image; Context of image processing
    • G06T2207/30168—Image quality inspection

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Probability & Statistics with Applications (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a video fuzzy anomaly detection method and a system based on frequency domain skewness and frequency domain peak values, wherein the method comprises the following steps: step S2: extracting a reference video frame and a video frame to be detected from a video: continuously extracting a plurality of video frames at the starting moment of each period as reference video frames for calculating a definition measurement characteristic reference value; extracting a plurality of video frames in the video stream according to a set extraction cycle at each time interval to serve as video frames to be detected; step S3: performing image blocking on the extracted video frame and calculating a frequency domain skewness matrix of each image block and a normalized frequency domain peak value matrix; the method integrates the influence of local blurring of the image on the evaluation of the definition of the whole image, detects the image to be evaluated in different regions, performs element-level comparison on evaluation values obtained from all the regions to obtain a definition anomaly measurement matrix, and finally judges the blurring condition of the video frame according to the definition anomaly measurement matrix to realize a better image blurring detection effect.

Description

Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value
Technical Field
The invention relates to the technical field of video image detection, in particular to a video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak values.
Background
With the rapid development of modern science and technology and the remarkable improvement of safety requirements of people, the requirement for the definition detection of real-time pictures of a video monitoring system is increased. The video monitoring system is an important part forming a safety technology prevention system, but at present, imaging equipment has a large number of fuzzy images, so that the availability of videos is greatly reduced, and wrong information can be brought to a user. Therefore, it is necessary to detect the image sharpness of the video surveillance system in real time. At present, there are two methods for evaluating image quality, namely a subjective evaluation method and an objective evaluation method. The subjective evaluation method directly judges the image blurring degree by means of human vision, and the method is time-consuming, labor-consuming, poor in economy and incapable of being popularized comprehensively. The objective evaluation method is based on theory, a calculation model is designed, and images are judged through the established mathematical model, so that the method is fast and stable, and quantitative data can be given. The method is easy to implement, and efficient and quick. The currently common image objective evaluation methods are classified into three categories, which are respectively as follows: full reference quality evaluation, weak reference quality evaluation, no reference quality evaluation. For full-reference quality evaluation, an evaluation result needs to be obtained by comparing an evaluation image with a reference image, and for weak-reference quality evaluation, the quality evaluation of the evaluation image needs to be completed by utilizing partial characteristic information in the reference image.
In the quality evaluation of the non-reference image, the definition of the image is an important index for measuring the image quality, the definition of the image can correspond to the visual perception of people, and the evaluation of the definition of the image is not high, so that the image is blurred. Blur is a common form of image degradation that directly determines the sharpness of an image. At present, scholars at home and abroad have proposed a plurality of non-reference fuzzy image quality evaluation models and methods. Zhang Fan et al propose a perception-based hybrid video quality assessment model. The algorithm adaptively combines significant distortion and blurring artifacts to simulate the HVS perception process by using an enhanced nonlinear model. Wu Yujie et al propose an end-to-end fuzzy assistant feature aggregation network (BFAN) for video object detection. The algorithm mainly aims at the aggregation process under the influence of motion blur, defocus and other blur. Huang Rixing et al propose a no-reference image quality analysis method based on edge blur measure. The method disperses all fuzzy features into three values by regression analysis. Then input into a classifier to perform fuzzy classification on the image. TangChang et al propose a fuzzy metric method based on log-average spectral residuals to obtain a coarse fuzzy graph. The iterative updating mechanism proposed by the method can partially solve the problem of distinguishing the smooth area and the fuzzy smooth area in the focus. Although these algorithms have good evaluation effects, they all have a common problem. The algorithms are based on global image calculation or comparison, if the image to be judged is locally blurred, such as local blurring, local brightness abnormality, local pure background image, local occlusion and the like, the original clear image is easily detected as a blurred image by the algorithms, so that detection errors are caused, and meanwhile, the algorithms cannot judge the position of the local blurring. For example, a part of an image is blocked to be black, and the definition of the blocked part is normal, if the definition of the image is judged by calculating the gradient of the whole image through the above algorithm, the gradient of the black blocked part affects the gradient of the whole image, so that the calculated gradient value cannot accurately reflect the gradient value of the whole image, and the result is misjudged. Meanwhile, although there are some algorithms for recognizing local blur at present, these algorithms do not show excellent applicability.
Disclosure of Invention
Therefore, a video blurring anomaly detection method and a video blurring anomaly detection system based on frequency domain skewness and frequency domain peak values need to be provided, and the problem that the existing image blurring detection effect is not ideal is solved.
In order to achieve the above object, the present invention provides a video fuzzy anomaly detection method based on frequency domain skewness and frequency domain peak value, comprising the following steps:
step S2: extracting a reference video frame and a video frame to be detected from a video: continuously extracting a plurality of video frames at the starting moment of each period as reference video frames for calculating a definition measurement characteristic reference value; extracting a plurality of video frames in the video stream according to a set extraction cycle at each time interval to serve as video frames to be detected;
step S3: performing image blocking on the extracted video frame and calculating a frequency domain skewness matrix of each image block and a normalized frequency domain peak value matrix;
step S3-1, setting the image size as h x w pixels, and cutting the image into blocks according to proportion, wherein each image block is marked as Gk,iAnd Gd,iK is marked as a reference video frame, d is marked as a video frame to be detected, wherein i is 1,2, …, M and M is the number of image blocks after the video frame is cut; image block size after cutting is
Figure BDA0003364404580000031
Wherein: "[]"represents a ceiling operation; step S3-2, a frequency domain skewness matrix and a frequency domain peak value matrix of each image block are constructed, wherein alpha is the cutting number of h pixel directions, beta is the cutting number of w pixel directions;
step S4: judging whether a reference value exists in the current time period or not; if the reference value exists, performing step S5; otherwise, taking the result of averaging the frequency domain skewness matrix and the frequency domain peak value matrix of the reference video frame as the current reference value and executing the step S6;
step S5: comparing the average frequency domain skewness matrix and each element value of the frequency domain peak value of the reference video frame with the reference value matrix of the current corresponding time period respectively to judge whether the average frequency domain skewness matrix and each element value of the frequency domain peak value are within a preset allowable variation range; if the current reference value is within the allowable variation range, the current reference value is weighted and combined with the element value in the reference value matrix to serve as the current reference value; otherwise, keeping the element value in the reference value matrix as the current reference value;
and step S6, element-level comparison is carried out on the frequency domain kurtosis matrix and the normalized frequency domain peak matrix of the video frame to be detected and a current reference value to obtain a definition anomaly measurement matrix, if the frequency domain skewness and the frequency domain peak are both in a normal range, the image block is judged to be normal and is set as a first flag bit, if any condition is not met, the image block is judged to be abnormal and is set as a second flag bit, after the image block is completely detected, the flag bit matrix is output and is marked as the definition anomaly measurement matrix, according to the definition anomaly measurement matrix, if a second flag bit element exists in the matrix, the video frame to be detected is judged to have a fuzzy condition, and if not, the video frame to be detected is judged to be a normal image.
Further, the method also includes step S1: for the video stream that is continuously input, after the video frame definition abnormality detection is triggered at the start time of each set period cycle, the process proceeds to step S2.
Further, the continuously input video stream is a video stream input by a monitoring camera.
Further, the step S3-2 includes the following steps:
(1) for Gk,iAnd Gd,iPerforming the same treatment, uniformly adopting GiPerforming a calculation of GiFourier transform d ofiComprises the following steps:
di＝F(u,ν)＝∫∫f(x,y)e-j2π(ux+vy)dxdy
calculate the image block GiD ofiAmplitude | d ofiAnd calculating its skewness xiiThereby obtaining Gk,iAnd Gd,iFrequency domain skewness xik,i,ξd,i；
Figure BDA0003364404580000041
Wherein n is the total amplitude number;
(2) based on the determined Gk,iAnd Gd,iFrequency domain skewness xik,i,ξd,iConstructing frequency domain skewness matrixes of the video frames fk and fd, which are respectively marked as xik,Ξd；
Figure BDA0003364404580000042
Figure BDA0003364404580000043
(3) Image block GiD ofiAmplitude | d ofiAnd calculating the frequency domain peak value Dmax after the | is normalized, and constructing a frequency domain peak value matrix as above.
Further, the method also comprises the following steps when the image is cut: if the image is not completely cut, the remaining image block length is
Figure BDA0003364404580000044
Has a width of
Figure BDA0003364404580000045
Further, the first flag bit is 0, and the second flag bit is 1.
The invention provides a video fuzzy anomaly detection system based on frequency domain skewness and frequency domain peaks, which comprises a memory and a processor, wherein the memory is stored with a computer program, and the computer program realizes the steps of the method according to any one of the embodiments of the invention when being executed by the processor.
Different from the prior art, the technical scheme includes that a reference video frame and a video frame needing to be detected are extracted firstly, then image cutting is carried out on the extracted video frame, a frequency domain skewness matrix and a frequency domain peak value matrix are constructed, element-level comparison is carried out on the frequency domain skewness matrix of the video frame to be detected and a reference value to obtain an image definition abnormity measurement matrix, and finally the fuzzy condition of the video frame is output according to the definition abnormity measurement matrix. The influence of local blurring of the image on the evaluation of the definition of the whole image is integrated, the image to be evaluated is detected in different regions, the evaluation values obtained from the regions are subjected to element level comparison to obtain a definition anomaly measurement matrix, and finally the blurring condition of the video frame is judged according to the definition anomaly measurement matrix, so that a better image blurring detection effect is realized.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
To explain technical contents, structural features, and objects and effects of the technical solutions in detail, the following detailed description is given with reference to the accompanying drawings in conjunction with the embodiments.
Referring to fig. 1, in order to overcome the shortcomings of the conventional research, the present invention provides a method for detecting video frame sharpness based on frequency domain skewness and kurtosis of image blocks. Firstly, the inventor finds out through research that the more fuzzy the image is, the larger the frequency domain skewness of the image is; the clearer the image is, the smaller the deviation of the image frequency domain is. Meanwhile, as the image is more and more blurred, the frequency domain skewness of the image is not infinite, but continuously approaches a fixed value, namely an upper bound exists, the frequency domain peak value of the image is higher and higher, but the increasing rate of the frequency domain peak value is continuously reduced, the position of the frequency domain peak value is slowly shifted to the left, but the moving rate is continuously reduced. Based on the theory, the algorithm is provided. The method considers the influence of the local blurring of the image on the evaluation of the definition of the whole image, detects the subareas of the image to be evaluated, performs element-level comparison on evaluation values obtained from all the subareas to obtain a definition anomaly measurement matrix, and finally judges the blurring condition of the video frame according to the definition anomaly measurement matrix. The method comprises the steps of firstly extracting a reference video frame and a video frame to be detected, then carrying out image cutting on the extracted video frame, constructing a frequency domain skewness matrix and a frequency domain peak value matrix, carrying out element-level comparison on the frequency domain skewness matrix of the video frame to be detected and a reference value to obtain an image definition anomaly measurement matrix, and finally outputting the fuzzy condition of the video frame according to the definition anomaly measurement matrix, thereby achieving better effect.
The invention relates to a video frame definition detection method based on frequency domain skewness and kurtosis of image blocks, which comprises the following steps:
step S1: and triggering video frame definition abnormity detection at the starting moment of each set period aiming at the video stream continuously input by the monitoring camera. The video stream that is continuously input may be generated by a monitoring camera, or in some embodiments, may be a network video stream, and detection of the network video may be implemented. And step S1 is an unnecessary step, and if the video is a video preset for a certain segment, step S2 may be directly performed.
Step S2: extracting a reference video frame and a video frame to be detected: continuously extracting a plurality of video frames at the starting moment of each period as reference video frames for calculating a definition measurement characteristic reference value; and extracting a plurality of video frames in the video stream as video frames to be detected according to a set extraction period (namely, a certain period is arranged between every two adjacent video frames) in each period.
Step S3: and carrying out image blocking on the extracted video frame according to the proportion, and calculating a frequency domain skewness matrix of each image block and a normalized frequency domain peak value matrix. Step S3 includes step S3-1 and step S3-2.
Step S3-1: setting the image size as h x w pixels, partitioning and cutting the image according to proportion (preferably integer proportion), and marking each image block as Gk,iAnd Gd,iK is marked as a reference video frame, d is marked as a video frame to be detected, wherein i is 1,2, …, M; image block size after cutting is
Figure BDA0003364404580000061
If an image can be completely cut, there will be no remaining image blocks. If the image is not completely cut, the length of the remaining image block is
Figure BDA0003364404580000062
Has a width of
Figure BDA0003364404580000063
Wherein: "[]"means a ceiling operation. α is the number of h pixel direction cuts, β is the number of w pixel direction cuts, and preferably, the values can be as follows:
α∈[3,4,5]
β∈[3,4]
and then step S3-2 is carried out to construct a frequency domain skewness matrix and a frequency domain peak value matrix. The method specifically comprises the following steps:
(1) for Gk,iAnd Gd,iThe same treatment is carried out as described above,unified adoption of GiPerforming a calculation of GiFourier transform d ofiComprises the following steps:
di＝F(u,ν)＝∫∫f(x,y)e-j2π(ux+vy)dxdy
calculate the image block GiD ofiAmplitude | d ofiAnd calculating its skewness xiiThereby obtaining Gk,iAnd Gd,iFrequency domain skewness xik,i,ξd,i；
Figure BDA0003364404580000071
Wherein n is the total amplitude number;
(2) based on the determined Gk,iAnd Gd,iFrequency domain skewness xik,i,ξd,iConstructing frequency domain skewness matrixes of the video frames fk and fd, which are respectively marked as xik,Ξd；
Figure BDA0003364404580000072
Figure BDA0003364404580000073
(3) Image block GiD ofiAmplitude | d ofiAnd calculating the frequency domain peak value Dmax after the | is normalized, and constructing a frequency domain peak value matrix as above.
Step S4: judging whether a reference value exists in the current time period or not, wherein the reference value is a preset value; if yes, go to step S5; otherwise, the result of averaging the frequency domain skewness matrix and the frequency domain peak value matrix of the reference video frame is used as the current reference value and step S6 is executed.
Step S5: comparing each element of the average frequency domain skewness matrix and the frequency domain peak value of the reference video frame with the reference value matrix of the current corresponding time period respectively to judge whether the average frequency domain skewness matrix and the frequency domain peak value are in an allowable variation range; and if so, weighting and combining the current value and the reference value of the reference value matrix and the element value in the reference value matrix to serve as the current reference value, and weighting the current value and the reference value of the reference video average frequency domain skewness matrix and the current value and the reference value of the frequency domain peak value during weighting. The specific weighting may be as follows: w is the current value + (1-w) the reference value, wherein w is a weighted value of 0-1. Otherwise, keeping the element value in the reference value matrix as the current reference value.
And step S6, performing element-level comparison on the frequency domain kurtosis matrix of the video frame to be detected and the normalized frequency domain peak matrix with a reference value to obtain a definition anomaly measurement matrix, judging that the image is normal if the frequency domain skewness and the frequency domain peak are both in a normal range, setting a flag bit '0' (the specific value of the flag bit can be modified according to actual needs), judging that the image is abnormal if any condition is not met, setting a flag bit '1' (the specific value of the flag bit can be modified according to actual needs), outputting the flag bit matrix which is hereinafter referred to as the definition anomaly measurement matrix after the image block is completely detected, judging that the video frame has a fuzzy condition if the matrix has '1' elements according to the definition anomaly measurement matrix, and otherwise, judging that the image is normal.
The invention provides a video fuzzy anomaly detection system based on frequency domain skewness and frequency domain peaks, which comprises a memory and a processor, wherein the memory is stored with a computer program, and the computer program realizes the steps of the method according to any one of the embodiments of the invention when being executed by the processor. When the system carries out video frame fuzzy detection, the influence of local fuzzy of the image on the definition evaluation of the whole image is integrated, the image to be evaluated is detected in different regions, element-level comparison is carried out on evaluation values obtained from the regions, a definition anomaly measurement matrix is obtained, finally the fuzzy condition of the video frame is judged according to the definition anomaly measurement matrix, and a better image fuzzy detection effect is realized.
It should be noted that, although the above embodiments have been described in the present invention, the scope of the present invention is not limited thereby. Therefore, based on the innovative concepts of the present invention, the technical solutions of the present invention can be directly or indirectly applied to other related technical fields by making changes and modifications to the embodiments of the present invention or by using equivalent structures or equivalent processes performed in the content of the present specification and the attached drawings, which are included in the scope of the present invention.

Claims (7)

1. The video fuzzy anomaly detection method based on the frequency domain skewness and the frequency domain peak value is characterized by comprising the following steps of:
step S2: extracting a reference video frame and a video frame to be detected from a video: continuously extracting a plurality of video frames at the starting moment of each period as reference video frames for calculating a definition measurement characteristic reference value; extracting a plurality of video frames in the video stream according to a set extraction cycle at each time interval to serve as video frames to be detected;
step S3: performing image blocking on the extracted video frame and calculating a frequency domain skewness matrix of each image block and a normalized frequency domain peak value matrix;
step S3-1, setting the image size as h x w pixels, and cutting the image into blocks according to proportion, wherein each image block is marked as Gk，iAnd Gd，iK is marked as a reference video frame, d is marked as a video frame to be detected, wherein i is 1,2, …, M and M is the number of image blocks after the video frame is cut; image block size after cutting is
Figure FDA0003364404570000011
Wherein: "[]"represents a ceiling operation; step S3-2, a frequency domain skewness matrix and a frequency domain peak value matrix of each image block are constructed, wherein alpha is the cutting number of h pixel directions, beta is the cutting number of w pixel directions;
step S4: judging whether a reference value exists in the current time period or not; if the reference value exists, performing step S5; otherwise, taking the result of averaging the frequency domain skewness matrix and the frequency domain peak value matrix of the reference video frame as the current reference value and executing the step S6;
step S5: comparing the average frequency domain skewness matrix and each element value of the frequency domain peak value of the reference video frame with the reference value matrix of the current corresponding time period respectively to judge whether the average frequency domain skewness matrix and each element value of the frequency domain peak value are within a preset allowable variation range; if the current reference value is within the allowable variation range, the current reference value is weighted and combined with the element value in the reference value matrix to serve as the current reference value; otherwise, keeping the element value in the reference value matrix as the current reference value;
and step S6, element-level comparison is carried out on the frequency domain kurtosis matrix and the normalized frequency domain peak matrix of the video frame to be detected and a current reference value to obtain a definition anomaly measurement matrix, if the frequency domain skewness and the frequency domain peak are both in a normal range, the image block is judged to be normal and is set as a first flag bit, if any condition is not met, the image block is judged to be abnormal and is set as a second flag bit, after the image block is completely detected, the flag bit matrix is output and is marked as the definition anomaly measurement matrix, according to the definition anomaly measurement matrix, if a second flag bit element exists in the matrix, the video frame to be detected is judged to have a fuzzy condition, and if not, the video frame to be detected is judged to be a normal image.
2. The method for detecting video blurring abnormality based on frequency domain skewness and frequency domain peaks of claim 1, further comprising step S1: for the video stream that is continuously input, after the video frame definition abnormality detection is triggered at the start time of each set period cycle, the process proceeds to step S2.
3. The method according to claim 2, wherein the video blur anomaly detection method based on the frequency domain skewness and the frequency domain peak value is characterized in that: the video stream continuously input is the video stream input by the monitoring camera.
4. The method for detecting video blurring abnormality according to claim 1, wherein said step S3-2 includes the following steps:
(1) for Gk，iAnd Gd，iPerforming the same treatment, uniformly adopting GiPerforming a calculation of GiFourier transform d ofiComprises the following steps:
di＝F(u，ν)＝∫∫f(x，y)e-j2π(ux+vy)dxdy
calculate the image block GiD ofiAmplitude | d ofiAnd calculating its skewness xiiThereby obtaining Gk，iAnd Gd，iIn the frequency domain ofSkewness xik，i，ξd，i；
Figure FDA0003364404570000021
Wherein n is the total amplitude number;
(2) based on the determined Gk，iAnd Gd，iFrequency domain skewness xik，i，ξd，iConstructing frequency domain skewness matrixes of the video frames fk and fd, which are respectively marked as xik，Ξd；
Figure FDA0003364404570000022
(3) Image block GiD ofiAmplitude | d ofiAnd calculating the frequency domain peak value Dmax after the | is normalized, and constructing a frequency domain peak value matrix as above.
5. The method for detecting video blurring-anomaly according to claim 1, further comprising the steps of: if the image is not completely cut, the remaining image block length is
Figure FDA0003364404570000031
Has a width of
Figure FDA0003364404570000032
6. The method according to claim 1, wherein the video blur anomaly detection method based on the frequency domain skewness and the frequency domain peak value is characterized by: the first flag bit is 0, and the second flag bit is 1.
7. Video fuzzy anomaly detection system based on frequency domain skewness and frequency domain peak value is characterized in that: comprising a memory, a processor, said memory having stored thereon a computer program which, when being executed by the processor, carries out the steps of the method according to any one of claims 1 to 6.
CN202111392348.6A 2021-11-19 2021-11-19 Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value Pending CN114205578A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111392348.6A CN114205578A (en) 2021-11-19 2021-11-19 Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111392348.6A CN114205578A (en) 2021-11-19 2021-11-19 Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value

Publications (1)

Publication Number Publication Date
CN114205578A true CN114205578A (en) 2022-03-18

Family

ID=80648416

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111392348.6A Pending CN114205578A (en) 2021-11-19 2021-11-19 Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value

Country Status (1)

Country Link
CN (1) CN114205578A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115439501A (en) * 2022-11-09 2022-12-06 慧视云创(北京)科技有限公司 Video stream dynamic background construction method and device and moving object detection method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115439501A (en) * 2022-11-09 2022-12-06 慧视云创(北京)科技有限公司 Video stream dynamic background construction method and device and moving object detection method
CN115439501B (en) * 2022-11-09 2023-04-07 慧视云创(北京)科技有限公司 Video stream dynamic background construction method and device and moving object detection method

Similar Documents

Publication Publication Date Title
Bahrami et al. A fast approach for no-reference image sharpness assessment based on maximum local variation
EP3104327B1 (en) Anomalous pixel detection
CN105678700B (en) Image interpolation method and system based on prediction gradient
CN109167997B (en) Video quality diagnosis system and method
US20120008832A1 (en) Region-of-interest video quality enhancement for object recognition
CN104486618B (en) The noise detecting method and device of video image
WO2003101120A1 (en) A method and system for estimating sharpness metrics based on local edge statistical distribution
CN110059666B (en) Attention detection method and device
CN113411571B (en) Video frame definition detection method based on sliding window gradient entropy
CN104182983B (en) Highway monitoring video definition detection method based on corner features
Li et al. GridSAR: Grid strength and regularity for robust evaluation of blocking artifacts in JPEG images
CN102231844A (en) Video image fusion performance evaluation method based on structure similarity and human vision
CN108174198B (en) Video image quality diagnosis analysis detection device and application system
CN114205578A (en) Video fuzzy anomaly detection method and system based on frequency domain skewness and frequency domain peak value
CN105469413B (en) It is a kind of based on normalization ring weighting without refer to smear restoration image synthesis method for evaluating quality
CN114708532A (en) Monitoring video quality evaluation method, system and storage medium
CN104657996B (en) The image quality evaluating method of Laplce's gaussian signal based on non-linear normalizing
Bahrami et al. Efficient image sharpness assessment based on content aware total variation
Bong et al. Objective blur assessment based on contraction errors of local contrast maps
CN111445435B (en) Multi-block wavelet transform-based reference-free image quality evaluation method
CN110349133B (en) Object surface defect detection method and device
Chen et al. A universal reference-free blurriness measure
Qureshi et al. An information based framework for performance evaluation of image enhancement methods
Maalouf et al. Offline quality monitoring for legal evidence images in video-surveillance applications
CN117011288B (en) Video quality diagnosis method and system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination