CN106385580B - Video jitter detection method based on gradation of image distribution characteristics - Google Patents
Video jitter detection method based on gradation of image distribution characteristics Download PDFInfo
- Publication number
- CN106385580B CN106385580B CN201610873192.6A CN201610873192A CN106385580B CN 106385580 B CN106385580 B CN 106385580B CN 201610873192 A CN201610873192 A CN 201610873192A CN 106385580 B CN106385580 B CN 106385580B
- Authority
- CN
- China
- Prior art keywords
- row
- row gray
- gray
- value
- variance
- 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.)
- Active
Links
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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/144—Movement detection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
Abstract
The present invention relates to a kind of video jitter detection method based on gradation of image distribution characteristics, including：The first step, Statistics Bar gray value and row gray value, calculate the expectation of row gray value and expectation and the variance of variance and row gray value respectively；Second step, a hypothesis testing is done in the row direction, judges shaken vertical direction whether there is；3rd step, a hypothesis testing is done in column direction, judgement whether there is in the horizontal direction is shaken.The present invention judge video whether there is shake when, fractional object can be moved and be made a distinction with video jitter, amount of calculation is small, and detection efficiency is fast, and Detection accuracy height, algorithm real-time is good.
Description
Technical field
It is more particularly to a kind of to be based on gradation of image distribution characteristics the present invention relates to a kind of detection method of monitor video shake
Video jitter detection method.
Background technology
Important component of the video monitoring system as vision Internet of Things, be widely used in urban safety, intelligent transportation,
In each field application such as wisdom environmental protection, border security, the contradiction of its regular maintenance inspection is growing more intense.According to statistics, it is domestic at present
In the monitoring system of operation, the ratio of the video camera of energy normal use is less than 60%, and the video monitoring system of substantial amounts
Operation and maintenance work majority manually detect and handle.How the efficiency of video monitoring system operation maintenance work is improved,
The running situation of head end video equipment is understood in time, is built intelligentized video monitoring quality diagnosis system and is led as video monitoring
Domain practical problem in the urgent need to address.
Video image dithering is the recurrent picture quality anomaly of video monitoring equipment.Under normal circumstances, move
Transition is smooth that picture correlation is more continuous between the continuous multiple frames of image sequence, but if correlation between them
There is great fluctuation process in property, and the situation of shake just occurs in video.In video monitoring, camera is typically all to be fixed on some position
The reason for putting, therefore causing video pictures jitter phenomenon mainly has：1) camera is disturbed (such as high wind) to occur by environment
It is regular to swing so as to cause the upper and lower or left and right of image to shake；2) camera is moved by people, causes float.Appoint
It a kind of what situation, can all cause picture periodicity chatter or irregular distortion occur, it is different to can mean that camera work occurs
Often, the work effectiveness of video monitoring system is seriously affected, therefore, it is necessary to intelligence is carried out to the video image of video monitoring system
Analysis detection, finds video jitter failure problems, realizes instant alarming and reparation in time.
In the research related to video jitter, Niu Yaoyao, Hong Danfeng are in paper《A kind of Traffic Surveillance Video based on HGPC
Shake method for detecting abnormality》Propose a kind of Gray Projection correlation coefficient process (Hierarchical Gray based on classification
Project Correlation, HGPC).The algorithm is mainly the identical of the row Gray scale projection curve of two field pictures before and after
Degree judges to shake vertical direction whether there is, then the goodness of fit for the row Gray scale projection curve for passing through front and rear two field pictures judges
Horizontal direction, which whether there is, is shaken, and is shaken so as to judge whole video whether there is.
We experimentally found that the video image of shake and the ranks grey value profile feature of normal video image are poor
It is different obvious, based on this, it is proposed that a kind of video jitter detection method based on gray distribution features.This method Detection accuracy
Height, algorithm real-time are good.
The content of the invention
It is an object of the invention to provide a kind of video jitter detection method based on gradation of image distribution characteristics.The invention according to
Meet normal distribution according to normal video image ranks gray value, and the ranks gray value for shaking video image does not meet normal state point
Cloth, using this distribution characteristics difference, by the ranks gray value of front and rear frame of video, construction APPROXIMATE DISTRIBUTION is the system of normal distribution
Metering, calculate ranks gray scale and examine the factor, to weigh the continuity of front and rear two field pictures and correlation, and then judge that the video is
It is no to be shaken.
The technical step of the present invention is as follows：
It is as follows based on the video jitter detection method of gradation of image distribution characteristics, step：
Step 1：Intercept front and rear adjacent two field pictures in video.
Step 2：The two field pictures of interception are transformed into gray space.
Step 3：The often row gray value and each column gray value of previous frame image are counted, calculates trip gray average and row gray scale
Variance yields and row gray average and row gray variance value.
Step 4：The often row gray value and each column gray value of latter two field picture are counted, calculates trip gray average and row gray scale
Variance yields and row gray average and row gray variance value.
Step 5：The row gray average and row gray variance for the front and rear two field pictures that step 3 and step 4 are calculated are done
One row hypothesis testing, obtain a performing check factor.
Step 6：The row gray average and row gray variance for step 3 and step 4 being calculated front and rear two frame do one
Row hypothesis testing, obtain a row and examine the factor.
Step 7：The performing check factor that step 5 and step 6 are calculated and row examine the factor respectively with given threshold value
Compare, if one of them exceeds threshold value, judge previous frame image to shake frame.
Step 8：Count in whole video, shake the ratio shared by frame, if it exceeds setting dithering threshold, then judge that this is regarded
Frequency is shake.
Further, in step 3, row gray value, row gray average and row gray variance computational methods are as follows：
Row gray value
Row gray average
Row gray variance
Wherein f (i, j) is the gray value of each pixel in gray level image, and A is picturedeep, and B is picturewide.
Row gray value, row gray average and row gray variance computational methods are as follows：
Row gray value
Row gray average
Row gray variance
Further in steps of 5, the computational methods of the performing check factor are as follows：
The performing check factor
Wherein,sx 2 1It is the row gray average and row gray variance of previous frame image respectively,sx 2 2It is latter respectively
The row gray average and row gray variance of two field picture.
Further, in step 6, row examine the computational methods of the factor as follows：
Row examine the factor
Wherein,sy 2 1It is the row gray average and row gray variance of previous frame image respectively,sy 2 2It is latter respectively
The row gray average and row gray variance of frame.
Further, in step 7, the performing check factor and row examine the threshold value of the factor to be set to α, if θx>α or θy>α has one
Set up, then judge previous frame image to shake frame.
Further, in step 8, the value of the shake frame number counted according to step 7, the ratio that shake frame number accounts for totalframes is calculated
Example k, if dithering threshold is β, if k>β is set up, then judges that the video is shaken.
Beneficial effects of the present invention：
The present invention detects respectively to the ranks of image, can detect the shake of different directions.Amount of calculation is small, detection effect
Rate is high.Whether monitor video or the video of other equipment shooting, can effectively be detected with the algorithm, be detected
The degree of accuracy reaches more than 90%.
Brief description of the drawings
Fig. 1 is the algorithm flow chart of the present invention.
Embodiment
Below according to accompanying drawing, contain the video detection process of 101 frame of video to this with reference to a size for 1280*720
Invention is described in detail, and specific implementation step is as follows：
Step 1：Front and rear adjacent two field pictures in video are intercepted successively, are intercepted 100 times altogether.
Step 2：The two field pictures of interception are transformed into gray space.
Step 3：The often row gray value and each column gray value of former frame are counted, calculates trip gray average and row gray variance
Value and row gray average and row gray variance.
Step 4：The often row gray value and each column gray value of a later frame are counted, calculates trip gray average and row gray variance
Value and row gray average and row gray variance.
Step 5：One is done using the row gray average and row gray variance of the counted front and rear two field pictures of step 3 and step 4
Individual row hypothesis testing, obtain a performing check factor.
Step 6：One is done using the row gray average and row gray variance of the counted front and rear two field pictures of step 3 and step 4
Individual row hypothesis testing, obtain a row and examine the factor.
Step 7：The factor is examined to be compared with given threshold value using step 5 and the counted performing check factor of step 6 and row,
If wherein there is one to exceed threshold value, judging former frame for shake frame.
Step 8：Count in whole video, shake frame shared by ratio, if it exceeds the threshold, then judge the video there occurs
Shake.
Further, in step 3, row gray average and row gray variance computational methods are as follows：
Row gray value
Row gray average
Row gray variance
Wherein f (i, j) is the gray value of each pixel in gray level image, and A is the line number of image, and B is the columns of image.
Further, in step 3, row gray average and row gray variance computational methods are as follows：
Row gray value
Row gray average
Row gray variance
Further in steps of 5, the computational methods of the performing check factor are as follows：
The performing check factor
Wherein,sx 2 1It is the row gray average and row gray variance of former frame respectively,sx 2 2It is a later frame respectively
Row gray average and row gray variance.
Further in step 6, row examine the computational methods of the factor as follows：
The performing check factor
Wherein,sy 2 1It is the row gray average and row gray variance of former frame respectively,sy 2 2It is a later frame respectively
Row gray average and row gray variance.
Further, in step 7, the performing check factor and row examine the threshold value of the factor to be set to 0.2, if θx>0.2 or θy>0.2
There is an establishment, then judge previous frame image to shake frame.
Further, in step 8, the value m of the shake frame number counted according to step 7, calculate shake frame number accounts for totalframes
Ratio k, whereinIf threshold value is 60%, if k>60% sets up, then judges that the video is shaken.
Claims (5)
1. the video jitter detection method based on gradation of image distribution characteristics, it is characterised in that comprise the following steps：
Step 1：Intercept front and rear adjacent two field pictures in video；
Step 2：The two field pictures of interception are transformed into gray space；
Step 3：The often row gray value and each column gray value of previous frame image are counted, calculates trip gray average and row gray variance
Value and row gray average and row gray variance value；
Step 4：The often row gray value and each column gray value of latter two field picture are counted, calculates trip gray average and row gray variance
Value and row gray average and row gray variance value；
Step 5：The row gray average and row gray variance for the front and rear two field pictures that step 3 and step 4 are calculated do one
Row hypothesis testing, obtain a performing check factor；
Step 6：The row gray average of front and rear two frame is calculated to step 3 and step 4 and row gray variance does a row vacation
If examining, obtain a row and examine the factor；
Step 7：The factor is examined to do ratio with given threshold value respectively the performing check factor and row that step 5 and step 6 are calculated
Compared with if one of them exceeds threshold value, judgement previous frame image is shake frame；
Step 8：Count in whole video, shake the ratio shared by frame, if it exceeds setting dithering threshold, then judge that the video is sent out
Raw shake；
The computational methods of the performing check factor described in step 5 are as follows：
The performing check factor
Wherein, It is the row gray average and row gray variance of previous frame image respectively, It is a later frame respectively
The row gray average and row gray variance of image；
Row described in step 6 examine the computational methods of the factor as follows：
Row examine the factor
Wherein, It is the row gray average and row gray variance of previous frame image respectively, It is a later frame respectively
Row gray average and row gray variance；Wherein, A is the line number of image, and B is picturewide.
2. the video jitter detection method according to claim 1 based on gradation of image distribution characteristics, it is characterised in that step
In rapid 3, row gray value, row gray average and row gray variance computational methods are as follows：
Row gray value
Row gray average
Row gray variance
Wherein f (i, j) is the gray value of each pixel of gray level image, and A is the line number of image, and B is picturewide.
3. the video jitter detection method according to claim 2 based on gradation of image distribution characteristics, it is characterised in that step
In rapid 3, row gray value, row gray average and row gray variance computational methods are as follows：
Row gray value
Row gray average
Row gray variance
4. the video jitter detection method according to claim 3 based on gradation of image distribution characteristics, it is characterised in that step
In rapid 7, the performing check factor and row examine the threshold value of the factor to be set to α, if θx>α or θy>α has an establishment, then judges former frame figure
As being shake frame.
5. the video jitter detection method according to claim 4 based on gradation of image distribution characteristics, it is characterised in that step
In rapid 8, the value of the shake frame number counted according to step 7, the ratio k that shake frame number accounts for totalframes is calculated, if dithering threshold is β,
If k>β is set up, then judges that the video is shaken.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610873192.6A CN106385580B (en) | 2016-09-30 | 2016-09-30 | Video jitter detection method based on gradation of image distribution characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610873192.6A CN106385580B (en) | 2016-09-30 | 2016-09-30 | Video jitter detection method based on gradation of image distribution characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106385580A CN106385580A (en) | 2017-02-08 |
| CN106385580B true CN106385580B (en) | 2018-02-06 |
Family
ID=57937234
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610873192.6A Active CN106385580B (en) | 2016-09-30 | 2016-09-30 | Video jitter detection method based on gradation of image distribution characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106385580B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101009769A (en) * | 2007-02-09 | 2007-08-01 | 于中 | Jitter-prevention processing method of TV image |
| CN103067741A (en) * | 2013-01-24 | 2013-04-24 | 浙江理工大学 | Shaking detection algorithm based on multi-feature fusion |
| CN103679750A (en) * | 2013-11-25 | 2014-03-26 | 武汉东智科技有限公司 | Camera shake detecting method based on videos |
| CN104301712A (en) * | 2014-08-25 | 2015-01-21 | 浙江工业大学 | Monitoring camera shaking detection method based on video analysis |
| CN105681663A (en) * | 2016-02-26 | 2016-06-15 | 北京理工大学 | Video jitter detection method based on inter-frame motion geometric smoothness |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7525572B2 (en) * | 2003-03-28 | 2009-04-28 | Victor Company Of Japan, Ltd. | Video camera with anti-shake system |
-
2016
- 2016-09-30 CN CN201610873192.6A patent/CN106385580B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101009769A (en) * | 2007-02-09 | 2007-08-01 | 于中 | Jitter-prevention processing method of TV image |
| CN103067741A (en) * | 2013-01-24 | 2013-04-24 | 浙江理工大学 | Shaking detection algorithm based on multi-feature fusion |
| CN103679750A (en) * | 2013-11-25 | 2014-03-26 | 武汉东智科技有限公司 | Camera shake detecting method based on videos |
| CN104301712A (en) * | 2014-08-25 | 2015-01-21 | 浙江工业大学 | Monitoring camera shaking detection method based on video analysis |
| CN105681663A (en) * | 2016-02-26 | 2016-06-15 | 北京理工大学 | Video jitter detection method based on inter-frame motion geometric smoothness |
Non-Patent Citations (1)
| Title |
|---|
| 《一种基于HGPC的交通监控视频抖动异常检测方法》;牛瑶瑶等;《青岛大学学报（自然科学版）》;20140815;第27卷(第3期);第38-43页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106385580A (en) | 2017-02-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111160125B (en) | Railway foreign matter intrusion detection method based on railway monitoring | |
| US11102417B2 (en) | Target object capturing method and device, and video monitoring device | |
| CN103903008B (en) | A kind of method and system of the mist grade based on image recognition transmission line of electricity | |
| CN104410855A (en) | Jitter detection method of monitoring video | |
| US7982774B2 (en) | Image processing apparatus and image processing method | |
| CN103761826B (en) | The recognition methods of a kind of thermal imaging video two mirror forest fires recognition system | |
| CN103096121A (en) | Camera moving detecting method and device | |
| EP3104327A2 (en) | Anomalous pixel detection | |
| CN104394377B (en) | A kind of fuzzy abnormal recognition methods of monitoring image and device | |
| CN105828065B (en) | A kind of video pictures overexposure detection method and device | |
| CN103093179B (en) | Video strip quantitative calculation method | |
| CN110021133B (en) | All-weather fire-fighting fire patrol early-warning monitoring system and fire image detection method | |
| CN105740792B (en) | Object detection method and device | |
| CN104732236B (en) | A kind of crowd's abnormal behaviour intelligent detecting method based on layered shaping | |
| JP2010097265A (en) | Smoke detecting apparatus | |
| CN104539936A (en) | System and method for monitoring snow noise of monitor video | |
| CN104539899A (en) | Method for distinguishing monitoring camera abnormal situation through video signal | |
| CN109905670A (en) | A kind of multi-stage platform monitoring system | |
| CN110660025A (en) | Industrial monitoring video image sharpening method based on GAN network | |
| CN110210530A (en) | Intelligent control method, device, equipment, system and storage medium based on machine vision | |
| CN109030499A (en) | A kind of device and method preventing defect counts repeat count suitable for target defect continuous on-line detection | |
| CN104657997B (en) | A kind of lens shift detection method and device | |
| CN106385580B (en) | Video jitter detection method based on gradation of image distribution characteristics | |
| CN103095966A (en) | Video jitter quantization method and video jitter quantization device | |
| JP2015158796A (en) | Object detection device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |