CN105282542B - The detection method and system of abnormal striped in a kind of video image - Google Patents
The detection method and system of abnormal striped in a kind of video image Download PDFInfo
- Publication number
- CN105282542B CN105282542B CN201510688768.7A CN201510688768A CN105282542B CN 105282542 B CN105282542 B CN 105282542B CN 201510688768 A CN201510688768 A CN 201510688768A CN 105282542 B CN105282542 B CN 105282542B
- Authority
- CN
- China
- Prior art keywords
- gray
- striped
- abnormal
- scale map
- straight line
- 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
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention provides the detection method and system of abnormal striped in a kind of video image, comprises the following steps:Read current video frame images;The gray-scale map of the gray-scale map, the gray-scale map of G passages and channel B of R passages is obtained respectively;Choosing the gray-scale map of any passage carries out Hough transform and obtains in the gray-scale map of selected passage to meet pre-conditioned straight line;Judge whether the quantity of straight line meets straight line amount threshold and when the straight line amount threshold is met, confirm there is abnormal striped in the gray-scale map of selected passage;Repeat the above steps, detect in other two gray-scale maps of passage with the presence or absence of abnormal striped;At least one when having abnormal striped in gray-scale map, the gray-scale map of G passages and the gray-scale map of channel B in R passages, confirms there is abnormal striped in current video frame images.Detection accuracy of the present invention is high, and real-time is good, in can be widely applied to the abnormal bar detection of the video image in existing intelligent safety and defence system.
Description
Technical field
The present invention relates to image technique field, the more particularly to detection of video image and administrative skill field, specially
The detection method and system of abnormal striped in a kind of video image.
Background technology
Video (Video) refer to a series of static images caught in the way of electric signal, noted down, being processed, being stored,
Transmission and the various technologies reappeared.Continuous image change it is per second more than more than 24 frames (frame) picture when, according to the persistence of vision
Principle, human eye cannot distinguish the tableaux of single width;Smooth continuous visual effect is appeared to be, so continuous picture is called
Video.Video technique is developed for television system, but has developed into a variety of forms now with profit consumption
Person gets off videograph.The prosperity of network technology also promotes the record fragment of video to be present in Yin Te in the form of streaming media
On net and can by computer receive with play.Video belongs to different technologies from film, and the latter is will be dynamic using photography
Image capturing is a series of still photo.
Rapid development of the intelligent security protection technology with the development of science and technology with progressive and 21st century information technology has been stepped
A brand-new field is entered, the boundary between intelligent security protection technology and computer progressively disappears.Technology of Internet of things
Popularization and application so that the security protection in city is developed from past simple security protection system to city integratedization system, the peace in city
Anti- project covers numerous fields, have Street Community, building construction, bank post office, road monitoring, motor vehicles, police,
Mobile object, ship etc..Especially for important place, such as:Airport, harbour, water, electricity and gas factory, bridge dam, river course, subway etc.
Place, comprehensive stereoscopic protective can be set up after introducing technology of Internet of things by means such as wireless mobile, track and localizations.
At present, the monitoring device of intelligent security guard is increasingly popularized, and constituted monitoring network of being networked by monitoring device is also got over
Come huger, this is accomplished by carrying out the working condition of monitoring device automatic detection and the abnormal conditions to occurring are alarmed.
In the case of exception occurs in monitor video picture, there is striped exception in video pictures so that the identification of target becomes in image
Difficulty, influences beholder's visual experience.If this kind of exception is not timely detected and processes, monitoring system can be caused at nobody
In the case of keeping an eye on, in invalid monitor state.The method that existing detection method is used is cumbersome, it is impossible to reach real-time detection simultaneously
And Detection accuracy is not high.
The content of the invention
The shortcoming of prior art in view of the above, it is an object of the invention to provide abnormal striped in a kind of video image
Detection method and system, for solve video image in the prior art occur striped it is abnormal when cannot precise real-time detection ask
Topic.
In order to achieve the above objects and other related objects, the present invention provides a kind of detection side of abnormal striped in video image
Method, the detection method of abnormal striped is comprised the following steps in the video image:1) current video frame images are read;2) to institute
Stating current video frame images carries out RGB channel separating treatment, and gray-scale map, the gray-scale map and B of G passages of R passages are obtained respectively
The gray-scale map of passage;3) choosing the gray-scale map of any passage carries out Hough transform and in the gray-scale map of passage selected by obtaining
Meet pre-conditioned straight line;4) judge whether the quantity of the straight line meets straight line amount threshold and in the number of the straight line
When amount meets the straight line amount threshold, confirm there is abnormal striped in the gray-scale map of selected passage;Repeat step 3) and
Step 4), detect in other two gray-scale maps of passage with the presence or absence of abnormal striped;5) gray-scale map in the R passages, the G
In the gray-scale map of the gray-scale map of passage and the channel B during at least one presence exception striped, the current frame of video is confirmed
There is abnormal striped in image.
Preferably, in the R passages in gray-scale map, the gray-scale map of the G passages and the gray-scale map of the channel B at least
When there is abnormal striped in one gray-scale map of passage, the detection method of abnormal striped also includes in the video image:Judge
Whether the continuously display time of the abnormal striped reaches time threshold, and is reached in the continuously display time of the abnormal striped
During the time threshold, then confirm there is abnormal striped in the current video frame images.
Preferably, the detection method of abnormal striped also includes in the video image:Confirming the current frame of video
Alarm is carried out when there is abnormal striped in image.
Preferably, choosing the gray-scale map of any passage carries out Hough transform and seizes the gray-scale map of the passage for taking selected
In the pre-conditioned straight line that meets specifically include:Choosing the gray-scale map of any passage carries out rim detection, obtains rim detection
Image;The edge-detected image to obtaining carries out Hough straight-line detections, obtains Hough matrixes;According to the Hough squares
Meet pre-conditioned straight line in the gray-scale map of the selected passage of battle array acquisition.
Preferably, it is described to meet pre-conditioned straight line and be specially:It is pre- that the length of the straight line and angle of inclination meet
If length and angle of inclination.
To achieve the above object, the present invention also provides a kind of detecting system of abnormal striped in video image, the video
The detecting system of abnormal striped includes in image:Image reading module, for reading current video frame images;Gray scale artwork
Block, is connected with described image read module, for the described current video frame images to being read in described image read module
RGB channel separating treatment is carried out, the gray-scale map of the gray-scale map, the gray-scale map of G passages and channel B of R passages is obtained respectively;Straight line is obtained
Modulus block, is connected with the gray scale module, and the gray-scale map for choosing any passage carries out Hough transform and obtains selected
Passage gray-scale map in meet pre-conditioned straight line;Judge to confirm module, be connected with the straight line acquisition module, be used for
Judge whether the quantity of the straight line meets straight line amount threshold and the quantity in the straight line meets the straight line amount threshold
When, confirm there is abnormal striped in the gray-scale map of selected passage;Abnormal striped module, judges to confirm module phase with described
Even, exist in the gray-scale map of the gray-scale map of the R passages, the gray-scale map of the G passages and the channel B at least one
During abnormal striped, confirm there is abnormal striped in the current video frame images.
Preferably, the detecting system of abnormal striped also includes in the video image:Duration judge module, it is and described
Judge to confirm that module is connected, for the gray-scale map of the gray-scale map in the R passages, the gray-scale map of the G passages and the channel B
In at least one passage gray-scale map in judge whether the continuously display time of the abnormal striped reaches when there is abnormal striped
Time threshold;When the continuously display time of the abnormal striped time threshold is reached, the abnormal striped module is true again
Recognize in the current video frame images and there is abnormal striped.
Preferably, the detecting system of abnormal striped also includes in the video image:Alarm module, with the exception
Striped module is connected, for carrying out alarm when there is abnormal striped in confirming the current video frame images.
Preferably, the straight line acquisition module includes:Edge-detected image unit, the gray-scale map for choosing any passage
Rim detection is carried out, edge-detected image is obtained;Matrix unit, is connected with the edge-detected image unit, for obtaining
The edge-detected image carry out Hough straight-line detections, obtain Hough matrixes;Straight line acquiring unit, with the matrix unit
It is connected, for meeting pre-conditioned straight line in the gray-scale map that selected passage is obtained according to the Hough matrixes.
Preferably, it is described to meet pre-conditioned straight line and be specially in the straight line acquiring unit:The length of the straight line
The default length met with angle of inclination and angle of inclination.
As described above, in a kind of video image of the invention abnormal striped detection method and system, with following beneficial
Effect:
The present invention carries out RGB channel separating treatment by current video frame images, and the gray scale of R passages is obtained respectively
The edge and these edges that striped is presented in the gray-scale map of figure, the gray-scale map of G passages and channel B are the spies of continuous straight line
Levy, the detection of straight line is carried out by carrying out Hough transform to former stripe pattern, analysis detects the individual of the straight line in each gray-scale map
Number and angle of inclination, so as to detect that the abnormal situation of striped occurs in video image, solve in the prior art due to electromagnetism
Interference etc. caused by reason video image there is striped exception, system cannot precise real-time detection problem.Present invention detection is accurate
True property is high, and real-time is good, in can be widely applied to the abnormal bar detection of the video image in existing intelligent safety and defence system.
Brief description of the drawings
Fig. 1 is shown as the schematic flow sheet of the detection method of abnormal striped in video image of the invention.
Fig. 2 is shown as the structural representation of the detecting system of abnormal striped in video image of the invention.
Component label instructions
The detecting system of abnormal striped in 1 video image
11 image reading modules
12 gray scale modules
13 straight line acquisition modules
14 judge to confirm module
15 abnormal striped modules
16 duration judge modules
17 alarm modules
S11~S15 steps
Specific embodiment
Embodiments of the present invention are illustrated below by way of specific instantiation, those skilled in the art can be by this specification
Disclosed content understands other advantages of the invention and effect easily.The present invention can also be by specific realities different in addition
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints with application, without departing from
Various modifications or alterations are carried out under spirit of the invention.
The purpose of the present embodiment is the detection method and system for providing abnormal striped in a kind of video image, for solving
Certainly in the prior art video image occur striped it is abnormal when cannot precise real-time detection problem.Described in detail below implementation
The abnormal detection method of striped and the principle of system and implementation method, make those skilled in the art not in a kind of video image of example
Creative work is needed to be appreciated that the detection method and system of abnormal striped in a kind of video image of the present embodiment.
The present embodiment provides a kind of detection method of abnormal striped in video image, is examined based on Hough transform and edge
The video pictures striped method for detecting abnormality of survey, the present embodiment can accurately detect that the abnormal situation of striped occurs in video, judge
Go out the angle of striped, and meet real-time, in may apply to existing intelligent safety and defence system.Specifically, as shown in figure 1, institute
The detection method for stating abnormal striped in video image is comprised the following steps.
Step S11, reads current video frame images.
The current video frame images are carried out RGB channel separating treatment by step S12, and the gray scale of R passages is obtained respectively
The gray-scale map of figure, the gray-scale map of G passages and channel B.
Step S13, choosing the gray-scale map of any passage carries out Hough transform and in the gray-scale map of passage selected by obtaining
Meet pre-conditioned straight line.
Step S14, judges whether the quantity of the straight line meets straight line amount threshold and in the quantity satisfaction of the straight line
During the straight line amount threshold, confirm there is abnormal striped in the gray-scale map of selected passage.
Repeat step S13 and step S14, detects in other two gray-scale maps of passage with the presence or absence of abnormal striped.
Step S15, in the gray-scale map, the gray-scale map of the G passages and the gray-scale map of the channel B in the R passages extremely
Few one when there is abnormal striped, confirms there is abnormal striped in the current video frame images.
Step S11 to step S15 is described in detail below.
The detection method of abnormal striped is specifically according to striped in video pictures in the video image that the present embodiment is provided
The edge of presentation and these edges are the features of continuous straight line, are carried out directly by carrying out Hough transform to former stripe pattern
The detection of line, the number and angle uniformity of the straight line that analysis is detected carry out detection image and the abnormal situation of striped occur, this
Invention can carry out alert process to there is the abnormal control point of striped.
Step S11, reads current video frame images.
Specifically, in the present embodiment, the duration that striped exception occurs in initialization video is T=0, is read current
Video frame images, such as labeled as [ft(i,j)]m×n, wherein, t represents detection striped abnormal moment, i.e. t frame video images
Picture, the size of image frame is m × n.
The current video frame images are carried out RGB channel separating treatment by step S12, and the gray scale of R passages is obtained respectively
The gray-scale map of figure, the gray-scale map of G passages and channel B.Even if multiple are various for the situation of image formation striped, may be only in image
One Color Channel produces striped, is still able to detect striped exception according to the inventive method.
If occurring in that the regular striped of arrangement in video pictures, to judge image whether there is striped, actually judge
The edge line in the absence of marshalling is deposited in image.But if from the point of view of the gray-scale map of whole video image, overall video figure
Not streaky feature in the gray-scale map of picture, therefore video image is divided into tri- passages of R, G, B in the present embodiment enters respectively
Row detection, obtains tri- gray-scale maps of passage of R, G, B.
Specifically, in the present embodiment, the striped abnormality strip=0 of initialisation image, and R, G, B tri- leads to
Striped the abnormality stripR=0, stripG=0, stripB=0 of the respective gray-scale map in road;Because the gray-scale map to image enters
Row observation, it is found that it not necessarily has streak feature, so to image [ft(i,j)]m×nChannel separation is carried out, R, G, B tri- is obtained
Respective gray-scale map [the f of individual passagetR(i,j)]m×n, [ftG(i,j)]m×n, [ftB(i,j)]m×n。
Step S13, choosing the gray-scale map of any passage carries out Hough transform and in the gray-scale map of passage selected by obtaining
Meet pre-conditioned straight line.
Specifically, in the present embodiment, choose the gray-scale map of any passage carry out Hough transform and seizing take it is selected
The pre-conditioned straight line that meets in the gray-scale map of passage is specifically included:
First, the gray-scale map for choosing any passage carries out rim detection, obtains edge-detected image.
Specifically, for example first to channel B gray level image [ftB(i,j)]m×nRim detection is carried out, edge image is obtained
[gtB(i,j)]m×n, the process of rim detection is as follows:
Wherein, i=0,1 ..., m-1, j=0,1 ..., n-4, th be judgment threshold.For image size m × n=384
The frame of video of × 288 resolution ratio, preferably empirical value are th=15.
Then Hough (Hough) straight-line detection is carried out to the edge-detected image for obtaining, obtains Hough matrixes;Tool
Body ground, in the present embodiment, to image [gtB(i,j)]m×nHough straight-line detections are carried out, Hough matrixes are obtained.
The Hough transform of image to implement process as follows:
It is as follows that the rectangular co-ordinate of image cathetus is converted to polar conversion formula:
γ=X × cos (θ)+Y × sin (θ);Wherein, the row of X representative images, the row of Y representative images, θ represents straight line L's
Angle of inclination, γ represents the former vertical line distance to straight line L of image.
Institute on straight line L can a little correspond to identical γ and θ.According to above-mentioned principle, Hough changes are carried out to image
Change.Because the value on Hough matrixes varies, for convenience of watching, it is normalized to [0,255].The horizontal seat of Hough matrixes
Mark represents angle, θ, scope [- 90,89];Ordinate represents radius γ, and scope is [0, S].The computing formula of wherein S is with reference to public
Formula:
Then pre-conditioned straight line is met in the gray-scale map of the passage according to selected by the Hough matrixes are obtained.
Wherein, it is described to meet pre-conditioned straight line and be specially:Default length that the length of the straight line and angle of inclination meet and
Angle of inclination.
Specifically, in the present embodiment, by Hough matrixes, straight line information linesB is obtained.Comprising satisfaction in linesB
The angle of inclination of the straight line of threshold requirement length, unit is radian.Then linesB is analyzed, the striped for calculating channel B image is sentenced
Disconnected factor strip.
Step S14, judges whether the quantity of the straight line meets straight line amount threshold and in the quantity satisfaction of the straight line
During the straight line amount threshold, confirm there is abnormal striped in the gray-scale map of selected passage.
Specifically, in the present embodiment, the length s of linesB is calculated, the radian in linesB is switched to angle
LinesBtheta, conversion formula such as formula:
LinesBtheta (p)=linesB (p) * 180/pi
Wherein, p=0,1 ..., s-1, pi represent for the radian of 180 ° of angle.
By the angle in linesBtheta, point 8 regions are counted, the region straight line number after being counted
regnum.The computational methods of regnum are according to equation below:
Wherein, k=0,1 ..., 7, p=0,1 ..., s-1, region 0 represent [- 90 °, -67.5 °), region 1 represent [-
67.5 °, -45 °), region 2 represent [- 45 °, -22.5 °), region 3 represent [- 22.5 °, 0 °), region 4 represent [0 °, 22.5 °),
Region 5 represent [22.5 °, 45 °), region 6 represent [45 °, 67.5 °), region 7 represent [67.5 °, 90 °).
The maximum M in regnum is calculated, judges whether M meets formula:M≥MTh, wherein, MThIt is judgment threshold, if
There is abnormal striped in the gray-scale map for being the passage for then confirming selected, the bar of channel B gray-scale map is set in the present embodiment
Line abnormality stripB=1, otherwise, stripB=0.
Repeat step S13 and step S14, detects in other two gray-scale maps of passage with the presence or absence of abnormal striped.It is i.e. right
R, G passage gray level image repeat step S13 and step S14, obtain the striped abnormality stripG of R, G passage gray-scale map,
stripR。
Step S15, in the gray-scale map, the gray-scale map of the G passages and the gray-scale map of the channel B in the R passages extremely
Few one when there is abnormal striped, confirms there is abnormal striped in the current video frame images.
Specifically, in the present embodiment, two field picture [f is calculatedt(i,j)]m×nStriped abnormality strip, computing formula
It is as follows:
If strip=1, two field picture [ft(i,j)]m×nThere is striped exception, otherwise, do not exist.
Additionally, in the present embodiment, gray-scale map, the gray-scale map of the G passages and the channel B in the R passages
When there is abnormal striped in the gray-scale map of at least one passage in gray-scale map, the detection method of abnormal striped in the video image
Also include:Judge whether the continuously display time of the abnormal striped reaches time threshold, and continuing in the abnormal striped
When the display time reaches the time threshold, then confirm there is abnormal striped in the current video frame images.
Specifically, if T >=ThT, wherein ThTIt is duration judgment threshold, is calculated according to the frame per second of 25 frame per second, uses
Family can set according to demand, for example, if the setting duration does not allow more than 1 second, ThT=25 × 1=25, then show to deposit
Persistently blocking, otherwise, then show be probably it is of short duration block, and do not block currently.
In the present embodiment, the detection method of abnormal striped also includes in the video image:Confirm it is described current
Alarm is carried out when there is abnormal striped in video frame images, striped exception persistently occurs in specific exportable video image
State outcome, such as text prompt or voice message.
To realize the detection method of abnormal striped in above-mentioned video image, the present embodiment correspondence is provided in a kind of video image
The detecting system of abnormal striped, specifically, as shown in Fig. 2 the detecting system 1 of abnormal striped includes in the video image:Figure
As read module 11, gray scale module 12, straight line acquisition module 13 judges to confirm module 14 and abnormal striped module 15.
Described image read module 11 is used to read current video frame images.In the present embodiment, initialization video goes out
The existing striped abnormal duration is T=0, reads current video frame images, such as labeled as [ft(i,j)]m×n, wherein, t
Detection striped abnormal moment, i.e. t frame video images picture are represented, the size of image frame is m × n.
The gray scale module 12 is connected with described image read module 11, for reading in described image read module 11
The described current video frame images for taking carry out RGB channel separating treatment, and gray-scale map, the gray scale of G passages of R passages are obtained respectively
The gray-scale map of figure and channel B.Even if image forms the situation of striped, and multiple are various, may only in image, some Color Channel be produced
Carded sliver line, system of the invention is still able to detect striped exception.
If occurring in that the regular striped of arrangement in video pictures, to judge image whether there is striped, actually judge
The edge line in the absence of marshalling is deposited in image.But if from the point of view of the gray-scale map of whole video image, overall video figure
Not streaky feature in the gray-scale map of picture, therefore video image is divided into tri- passages of R, G, B in the present embodiment enters respectively
Row detection, obtains tri- gray-scale maps of passage of R, G, B.
Specifically, in the present embodiment, the striped abnormality strip=0 of the initialisation image of gray scale module 12,
And striped the abnormality stripR=0, stripG=0, stripB=0 of the respective gray-scale map of tri- passages of R, G, B;Due to right
The gray-scale map of image is observed, it is found that it not necessarily has streak feature, so the gray scale module 12 is to image [ft
(i,j)]m×nChannel separation is carried out, tri- passages of R, G, B each gray-scale map [f is obtainedtR(i,j)]m×n, [ftG(i,j)]m×n,
[ftB(i,j)]m×n。
The straight line acquisition module 13 is connected with the gray scale module 12, and the gray-scale map for choosing any passage is carried out
Meet pre-conditioned straight line in the gray-scale map of Hough transform and the selected passage of acquisition.
Specifically, in the present embodiment, the straight line acquisition module 13 includes:Edge-detected image unit, matrix unit
And straight line acquiring unit (not shown).
The gray-scale map that the edge-detected image unit is used to choose any passage carries out rim detection, obtains rim detection
Image.
Specifically, for example first to channel B gray level image [ftB(i,j)]m×nRim detection is carried out, edge image is obtained
[gtB(i,j)]m×n, the process of rim detection is as follows:
Wherein, i=0,1 ..., m-1, j=0,1 ..., n-4, th be judgment threshold.For image size m × n=384
The frame of video of × 288 resolution ratio, preferably empirical value are th=15.
The matrix unit is connected with the edge-detected image unit, for entering to the edge-detected image for obtaining
Row Hough straight-line detections, obtain Hough matrixes;The straight line acquiring unit is connected with the matrix unit, for according to described
Meet pre-conditioned straight line in the gray-scale map of the selected passage of Hough matrixes acquisition.
Then Hough (Hough) straight-line detection is carried out to the edge-detected image for obtaining, obtains Hough matrixes;Tool
Body ground, in the present embodiment, to image [gtB(i,j)]m×nHough straight-line detections are carried out, Hough matrixes are obtained.
The Hough transform of image to implement process as follows:
It is as follows that the rectangular co-ordinate of image cathetus is converted to polar conversion formula:
γ=X × cos (θ)+Y × sin (θ);Wherein, the row of X representative images, the row of Y representative images, θ represents straight line L's
Angle of inclination, γ represents the former vertical line distance to straight line L of image.
Institute on straight line L can a little correspond to identical γ and θ.According to above-mentioned principle, Hough changes are carried out to image
Change.Because the value on Hough matrixes varies, for convenience of watching, it is normalized to [0,255].The horizontal seat of Hough matrixes
Mark represents angle, θ, scope [- 90,89];Ordinate represents radius γ, and scope is [0, S].The computing formula of wherein S is with reference to public
Formula:
Then pre-conditioned straight line is met in the gray-scale map of the passage according to selected by the Hough matrixes are obtained.
Wherein, it is described to meet pre-conditioned straight line and be specially in the straight line acquiring unit:The length of the straight line and angle of inclination
The default length for meeting and angle of inclination.
Specifically, in the present embodiment, by Hough matrixes, straight line information linesB is obtained.Comprising satisfaction in linesB
The angle of inclination of the straight line of threshold requirement length, unit is radian.Then linesB is analyzed, the striped for calculating channel B image is sentenced
Disconnected factor strip.
It is described judge confirm module 14 be connected with the straight line acquisition module 13, for judge the straight line quantity whether
Meet straight line amount threshold and when the quantity of the straight line meets the straight line amount threshold, confirm the ash of selected passage
There is abnormal striped in degree figure.
Specifically, in the present embodiment, in the judgement confirms module 14, the length s of linesB is calculated, linesB
In radian switch to angle linesBtheta, conversion formula such as formula:
LinesBtheta (p)=linesB (p) * 180/pi
Wherein, p=0,1 ..., s-1, pi represent for the radian of 180 ° of angle.
By the angle in linesBtheta, point 8 regions are counted, the region straight line number after being counted
regnum.The computational methods of regnum are according to equation below:
Wherein, k=0,1 ..., 7, p=0,1 ..., s-1, region 0 represent [- 90 °, -67.5 °), region 1 represent [-
67.5 °, -45 °), region 2 represent [- 45 °, -22.5 °), region 3 represent [- 22.5 °, 0 °), region 4 represent [0 °, 22.5 °),
Region 5 represent [22.5 °, 45 °), region 6 represent [45 °, 67.5 °), region 7 represent [67.5 °, 90 °).
The maximum M in regnum is calculated, judges whether M meets formula:M≥MTh, wherein, MThIt is judgment threshold, if
There is abnormal striped in the gray-scale map for being the passage for then confirming selected, the bar of channel B gray-scale map is set in the present embodiment
Line abnormality stripB=1, otherwise, stripB=0.
Using the straight line acquisition module 13 and it is described judge confirm module 14 detect other two gray-scale maps of passage in
With the presence or absence of abnormal striped.Obtain striped the abnormality stripG, stripR of R, G passage gray-scale map.
The abnormal striped module 15 judges to confirm that module 14 is connected with described, for the gray-scale map in the R passages, institute
When stating in the gray-scale map of G passages and the gray-scale map of the channel B at least one and there is abnormal striped, the current video is confirmed
There is abnormal striped in two field picture.
Specifically, in the present embodiment, in the abnormal striped module 15, two field picture [f is calculatedt(i,j)]m×nBar
Line abnormality strip, computing formula is as follows:
If strip=1, two field picture [ft(i,j)]m×nThere is striped exception, otherwise, do not exist.
Additionally, in the present embodiment, the detecting system 1 of abnormal striped also includes in the video image:Duration is sentenced
Disconnected module 16.The duration judge module 16 judges to confirm that module 14 is connected with described, for the gray scale in the R passages
When there is abnormal striped in the gray-scale map of at least one passage in the gray-scale map of figure, the gray-scale map of the G passages and the channel B
Judge whether the continuously display time of the abnormal striped reaches time threshold;Reached in the continuously display time of the abnormal striped
During to the time threshold, the abnormal striped module 15 confirms there is abnormal striped in the current video frame images again.
Specifically, if T >=ThT, wherein ThTIt is duration judgment threshold, is calculated according to the frame per second of 25 frame per second, uses
Family can set according to demand, for example, if the setting duration does not allow more than 1 second, ThT=25 × 1=25, then show to deposit
Persistently blocking, otherwise, then show be probably it is of short duration block, and do not block currently.
In the present embodiment, the detecting system 1 of abnormal striped also includes in the video image:Alarm module 17,
The alarm module 17 is connected with the abnormal striped module 15, for being deposited in the current video frame images are confirmed
Alarm is carried out in abnormal striped.It is different that striped persistently occurs in specific described exportable video image of alarm module 17
Normal state outcome, such as text prompt or voice message.
In sum, the present invention carries out RGB channel separating treatment by current video frame images, R is obtained respectively and is led to
The gray-scale map in road, the gray-scale map of G passages and striped is presented in the gray-scale map of channel B edge and these edges are continuous straight
The feature of line, the detection of straight line is carried out by carrying out to former stripe pattern Hough transform, and analysis detects straight in each gray-scale map
The number of line and angle of inclination, so as to detect that the abnormal situation of striped occurs in video image, solve in the prior art by
There is striped exception in video image caused by the reasons such as electromagnetic interference, system cannot precise real-time detection problem.The present invention
Detection accuracy is high, and real-time is good, can be widely applied to the abnormal striped of the video image in existing intelligent safety and defence system
In detection.So, the present invention effectively overcomes various shortcoming of the prior art and has high industrial utilization.
The above-described embodiments merely illustrate the principles and effects of the present invention, not for the limitation present invention.It is any ripe
The personage for knowing this technology all can carry out modifications and changes under without prejudice to spirit and scope of the invention to above-described embodiment.Cause
This, those of ordinary skill in the art is complete with institute under technological thought without departing from disclosed spirit such as
Into all equivalent modifications or change, should be covered by claim of the invention.
Claims (8)
1. in a kind of video image abnormal striped detection method, it is characterised in that:The inspection of abnormal striped in the video image
Survey method is comprised the following steps:
1) current video frame images are read;
2) RGB channel separating treatment is carried out to the current video frame images, respectively obtains the gray-scale map of R passages, G passages
The gray-scale map of gray-scale map and channel B;
3) satisfaction that choosing the gray-scale map of any passage carries out Hough transform and obtain in the gray-scale map of selected passage is preset
The straight line of condition;
4) judge whether the quantity of the straight line meets straight line amount threshold and the quantity in the straight line meets the straight line number
During amount threshold value, confirm there is abnormal striped in the gray-scale map of selected passage;
Repeat step 3) and step 4), detect in other two gray-scale maps of passage with the presence or absence of abnormal striped;
5) at least one is present in gray-scale map in the R passages, the gray-scale map of the G passages and the gray-scale map of the channel B
During abnormal striped, confirm there is abnormal striped in the current video frame images;
At least one passage in gray-scale map, the gray-scale map of the G passages and the gray-scale map of the channel B in the R passages
When there is abnormal striped in gray-scale map, the detection method of abnormal striped also includes in the video image:Judge the abnormal bar
Whether the continuously display time of line reaches time threshold, and reaches the time threshold in the continuously display time of the abnormal striped
During value, then confirm there is abnormal striped in the current video frame images.
2. in video image according to claim 1 abnormal striped detection method, it is characterised in that:The video image
The detection method of middle abnormal striped also includes:
Alarm is carried out when there is abnormal striped in confirming the current video frame images.
3. in video image according to claim 1 abnormal striped detection method, it is characterised in that:Choose any passage
Gray-scale map carry out the pre-conditioned straight line that meets that Hough transform and seizing taken in the gray-scale map of selected passage and specifically wrap
Include:
Choosing the gray-scale map of any passage carries out rim detection, obtains edge-detected image;
The edge-detected image to obtaining carries out Hough straight-line detections, obtains Hough matrixes;
Meet pre-conditioned straight line in the gray-scale map of the passage according to selected by the Hough matrixes are obtained.
4. in video image according to claim 3 abnormal striped detection method, it is characterised in that:It is described to meet default
The straight line of condition is specially:Default length and angle of inclination that the length of the straight line and angle of inclination meet.
5. in a kind of video image abnormal striped detecting system, it is characterised in that:The inspection of abnormal striped in the video image
Examining system includes:
Image reading module, for reading current video frame images;
Gray scale module, is connected with described image read module, for described current to what is read in described image read module
Video frame images carry out RGB channel separating treatment, the gray-scale map of R passages, the gray-scale map of G passages and channel B are obtained respectively
Gray-scale map;
Straight line acquisition module, is connected with the gray scale module, and the gray-scale map for choosing any passage carries out Hough transform simultaneously
Meet pre-conditioned straight line in the gray-scale map of the passage selected by obtaining;
Judge to confirm module, be connected with the straight line acquisition module, whether the quantity for judging the straight line meets straight line number
Measure threshold value and when the quantity of the straight line meets the straight line amount threshold, exist in the gray-scale map for confirming selected passage
Abnormal striped;
Abnormal striped module, judges to confirm that module is connected, for the gray-scale map in the R passages, the ash of the G passages with described
At least one when having abnormal striped in the gray-scale map of degree figure and the channel B, confirms to be deposited in the current video frame images
In abnormal striped;
The detecting system of abnormal striped also includes in the video image:
Duration judge module, judges to confirm that module is connected, for the gray-scale map in the R passages, the G passages with described
Gray-scale map and the channel B gray-scale map at least one passage gray-scale map in judge the exception when there is abnormal striped
Whether the continuously display time of striped reaches time threshold;
When the continuously display time of the abnormal striped time threshold is reached, the abnormal striped module confirms described again
There is abnormal striped in current video frame images.
6. in video image according to claim 5 abnormal striped detecting system, it is characterised in that:The video image
The detecting system of middle abnormal striped also includes:
Alarm module, is connected with the abnormal striped module, for existing in the current video frame images are confirmed
Alarm is carried out during abnormal striped.
7. in video image according to claim 5 abnormal striped detecting system, it is characterised in that:The straight line is obtained
Module includes:
Edge-detected image unit, the gray-scale map for choosing any passage carries out rim detection, obtains edge-detected image;
Matrix unit, is connected with the edge-detected image unit, for carrying out Hough to the edge-detected image for obtaining
Straight-line detection, obtains Hough matrixes;
Straight line acquiring unit, is connected with the matrix unit, the ash for obtaining selected passage according to the Hough matrixes
Meet pre-conditioned straight line in degree figure.
8. in video image according to claim 7 abnormal striped detecting system, it is characterised in that:The straight line is obtained
It is described to meet pre-conditioned straight line and be specially in unit:The default length that the length of the straight line and angle of inclination meet
And angle of inclination.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510688768.7A CN105282542B (en) | 2015-10-21 | 2015-10-21 | The detection method and system of abnormal striped in a kind of video image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510688768.7A CN105282542B (en) | 2015-10-21 | 2015-10-21 | The detection method and system of abnormal striped in a kind of video image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105282542A CN105282542A (en) | 2016-01-27 |
CN105282542B true CN105282542B (en) | 2017-06-30 |
Family
ID=55150755
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510688768.7A Active CN105282542B (en) | 2015-10-21 | 2015-10-21 | The detection method and system of abnormal striped in a kind of video image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105282542B (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110896470A (en) * | 2018-09-13 | 2020-03-20 | 广州弘度信息科技有限公司 | Video fringe detection method |
CN110298869A (en) * | 2019-07-08 | 2019-10-01 | 威创集团股份有限公司 | A kind of video analysis method and system |
CN110674763B (en) * | 2019-09-27 | 2022-02-11 | 国网四川省电力公司电力科学研究院 | Transmission channel tower image identification method and system based on symmetry inspection |
CN112399150B (en) * | 2020-11-16 | 2023-04-11 | 广州顶新时代电子系统科技有限公司 | Method for optimizing imaging picture of monitoring camera |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103093179B (en) * | 2011-10-28 | 2016-06-15 | 浙江大华技术股份有限公司 | Video strip quantitative calculation method |
CN104243969B (en) * | 2013-06-20 | 2017-11-24 | 中兴通讯股份有限公司 | Image jail-bar detection approach and device |
CN104282011B (en) * | 2013-07-04 | 2018-05-25 | 浙江大华技术股份有限公司 | The method and device of interference stripes in a kind of detection video image |
CN104657975B (en) * | 2014-05-13 | 2017-10-24 | 武汉科技大学 | A kind of method of video image travers Disturbance Detection |
CN104766297B (en) * | 2014-10-08 | 2016-03-02 | 江苏省电力公司电力科学研究院 | A kind of electric power video image striped fault detection method combined based on spatial domain and time-domain analysis |
-
2015
- 2015-10-21 CN CN201510688768.7A patent/CN105282542B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN105282542A (en) | 2016-01-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105282542B (en) | The detection method and system of abnormal striped in a kind of video image | |
CN107367310A (en) | A kind of river level remote monitoring method based on computer vision | |
CN104021388B (en) | Barrier during backing automatic detection and method for early warning based on binocular vision | |
CN105675623B (en) | It is a kind of based on the sewage color of sewage mouth video and the real-time analysis method of flow detection | |
CN105453153B (en) | Traffic lights detects | |
CN105702048B (en) | Highway front truck illegal road occupation identifying system based on automobile data recorder and method | |
CN103778786B (en) | A kind of break in traffic rules and regulations detection method based on remarkable vehicle part model | |
CN105913093B (en) | A kind of template matching method for Text region processing | |
CN109147368A (en) | Intelligent driving control method device and electronic equipment based on lane line | |
CN104766086B (en) | The monitoring and managing method and system of a kind of way mark | |
CN107392885A (en) | A kind of method for detecting infrared puniness target of view-based access control model contrast mechanism | |
EP2851841A2 (en) | System and method of alerting a driver that visual perception of pedestrian may be difficult | |
CN109284674A (en) | A kind of method and device of determining lane line | |
CN109064495A (en) | A kind of bridge floor vehicle space time information acquisition methods based on Faster R-CNN and video technique | |
TW201120814A (en) | Method for determining if an input image is a foggy image, method for determining a foggy level of an input image and cleaning method for foggy images | |
CN110390306A (en) | Detection method, vehicle and the computer readable storage medium of right angle parking stall | |
CN103456024B (en) | A kind of moving target gets over line determination methods | |
CN102867175B (en) | Stereoscopic vision-based ATM (automatic teller machine) machine behavior analysis method | |
US11315374B2 (en) | Detection of fraudulent access at control gates | |
CN107948465A (en) | A kind of method and apparatus for detecting camera and being disturbed | |
CN109635758A (en) | Wisdom building site detection method is dressed based on the high altitude operation personnel safety band of video | |
CN107301375A (en) | A kind of video image smog detection method based on dense optical flow | |
CN111325048A (en) | Personnel gathering detection method and device | |
CN107977596A (en) | A kind of car plate state identification method and device | |
CN109409191A (en) | A kind of zebra stripes vehicle evacuation detection method and system based on machine learning |
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 |