CN105574842A - Match side line video detection method - Google Patents
Match side line video detection method Download PDFInfo
- Publication number
- CN105574842A CN105574842A CN201410623320.2A CN201410623320A CN105574842A CN 105574842 A CN105574842 A CN 105574842A CN 201410623320 A CN201410623320 A CN 201410623320A CN 105574842 A CN105574842 A CN 105574842A
- Authority
- CN
- China
- Prior art keywords
- axis
- value
- point
- straight line
- straight
- 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
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to a match side line video detection method, belongs to the technical field of a video detection method, and particularly relates to the match side line video detection method. The invention provides the match side line video detection method capable of maintaining good edge information. The method comprises the following steps of 1, firstly, performing edge detection on an image to be processed; 2, scanning a binary image obtained in the first step, and when a non-zero pixel point is met, calculating the polar coordinate value corresponding to the point; 3, traversing the whole counter, ignoring all value results smaller than the threshold, and merging similar results; and 4, determining straight lines in an original image by a reverse projection method, firstly calculating the end points of the straight lines conforming to the conditions in the counter in the first step, and determining the projection direction of the pixel points near the straight lines according to the parameter values.
Description
Technical field
The invention belongs to video detecting method technical field, particularly relate to a kind of match sideline video detecting method.
Background technology
Football forbidden zone is divided into large forbidden zone and goal area.Goalmouth is named again in goal area, two ends, competition area are far from the line at 5.50m place inside goal, and in field, respectively draw a long line vertical with end line for 5.5m, its one end connects with goal, another end points is interconnected and parallel with end line, and the area within the scope of these three lines and end line is exactly goal area.Large forbidden zone is two ends, competition area apart from the end line at 16.5m place inside ball 35 door pillar, each picture long 16.5m in field the line vertical with end line, and another end points of these two lines is interconnected and parallel with end line, the region that these three lines and end line are encircled a city.Forbidden zone in this paper is detected mainly for large forbidden zone, because the forbidden zone lines rule of all competition area is identical, so the angle of lines number in the some pictures of football video and lines can determine this whether picture comprising forbidden zone to a great extent.The angle will Hough transform being used to extract number of lines in picture and lines herein, re-uses Bayes classifier to distinguish whether comprise forbidden zone.
Summary of the invention
The present invention is exactly for the problems referred to above, provides a kind of match sideline video detecting method keeping good marginal information.
For achieving the above object, the present invention includes following steps.
1) first image to be processed is carried out rim detection.
2) bianry image obtained in the previous step is scanned, often run into the pixel of a non-zero, just calculate this polar value corresponding to point.
3) travel through whole counter, will wherein be less than the value result Ignore All of thresholding, and merge similar result merging.
4) Inverse Projection is adopted to determine the straight line in former figure; The end points of qualified straight line in the first computing counter of the first step; Size according to the value of parameter decides the direction of pixel projection near straight line.
5) after whole subpoints has all marked, scanning x-axis, if there is one " 1 ", and in several positions of later in relation, major part is also " 1 ", then think that of set is the horizontal ordinate of required straight-line segment starting point.
As a kind of preferred version, of the present invention 2) bianry image obtained in the previous step is scanned, often run into the pixel of a non-zero, just calculate this polar value corresponding to point, then the counter corresponding to calculated polar value is added 1.
As another kind of preferred version, of the present invention 4) adopt Inverse Projection to determine the straight line in former figure; The end points of qualified straight line in the first computing counter of the first step; Size according to the value of parameter decides the direction of pixel projection near straight line, and larger value projects to y-axis, and less value then projects to x-axis; To invest the point of x-axis, if the distance between these point and straight lines is enough little, then the position in the x-axis of correspondence will puts 1, otherwise set to 0.
5) after whole subpoints has all marked, scanning x-axis, if there is one " 1 ", and major part is also " 1 " in several positions of later in relation, then think that of set is the horizontal ordinate of required straight-line segment starting point, obtain its ordinate bringing straight-line equation into, obtain an end points of required line segment.
Beneficial effect of the present invention.
The present invention carries out pre-service, denoising, the good marginal information of maintenance to image, then rim detection is carried out with sobel operator, obtain the obvious binary image in line edge, forbidden zone, recycling Hough transform obtains the information of lines, makes it to detect forbidden zone camera lens finally by training Bayes classifier.
Embodiment
The present invention includes following steps.
1) first image to be processed is carried out rim detection.
2) bianry image obtained in the previous step is scanned, often run into the pixel of a non-zero, just calculate this polar value corresponding to point.
3) travel through whole counter, will wherein be less than the value result Ignore All of thresholding, and merge similar result merging.
4) Inverse Projection is adopted to determine the straight line in former figure; The end points of qualified straight line in the first computing counter of the first step; Size according to the value of parameter decides the direction of pixel projection near straight line.
5) after whole subpoints has all marked, scanning x-axis, if there is one " 1 ", and in several positions of later in relation, major part is also " 1 ", then think that of set is the horizontal ordinate of required straight-line segment starting point.
Described 2) bianry image obtained in the previous step is scanned, often run into the pixel of a non-zero, just calculate this polar value corresponding to point, then the counter corresponding to calculated polar value is added 1.
Described 4) Inverse Projection is adopted to determine the straight line in former figure; The end points of qualified straight line in the first computing counter of the first step; Size according to the value of parameter decides the direction of pixel projection near straight line, and larger value projects to y-axis, and less value then projects to x-axis; To invest the point of x-axis, if the distance between these point and straight lines is enough little, then the position in the x-axis of correspondence will puts 1, otherwise set to 0.
5) after whole subpoints has all marked, scanning x-axis, if there is one " 1 ", and major part is also " 1 " in several positions of later in relation, then think that of set is the horizontal ordinate of required straight-line segment starting point, obtain its ordinate bringing straight-line equation into, obtain an end points of required line segment.
Report invention, after carrying out edge treated to image, utilizes region, hough change detection forbidden zone white line to determine edge, forbidden zone.The principle of Hough transform utilizes the global characteristics of image to connect the concrete shape of edge pixel, forms the method for continuous print smooth edges.Hough transform, by the mapping of parameter space and original image and cumulative, realizes the identification to known analytic expression curve.It is mainly used in the straight line or curve that detect bianry image, and it can detect the target of known form.Its advantage is that antijamming capability is strong, under the condition that signal to noise ratio (S/N ratio) is low, can detect straight line.
Disaggregated model based on Bayes statistical method is a kind of typical sorting technique.Bayes' theorem, as the core formula in bayesian theory, is the theoretical foundation of Bayesian learning method.It estimates the posterior probability of event by excavation event itself and the prior imformation of sampled data, ensure that the utilization factor of prior imformation, improves classify accuracy.
Why the accuracy that video forbidden zone is detected, more than 85%, also has some undetected, and be because when occurring in forbidden zone, possible scene switches relatively more frequent, or far and near shot transition, causes the linear feature of forbidden zone not obvious, thus causes judgement to occur mistake.
Above content is the further description done the present invention in conjunction with concrete preferred implementation; can not assert that specific embodiment of the invention is confined to these explanations; for general technical staff of the technical field of the invention; without departing from the inventive concept of the premise; some simple deduction or replace can also be made, all should be considered as belonging to the protection domain that claims that the present invention submits to are determined.
Claims (3)
1.
a kind of match sideline video detecting method, is characterized in that comprising the following steps:
1) first image to be processed is carried out rim detection;
2) bianry image obtained in the previous step is scanned, often run into the pixel of a non-zero, just calculate this polar value corresponding to point;
3) travel through whole counter, will wherein be less than the value result Ignore All of thresholding, and merge similar result merging;
4) Inverse Projection is adopted to determine the straight line in former figure; The end points of qualified straight line in the first computing counter of the first step; Size according to the value of parameter decides the direction of pixel projection near straight line;
5) after whole subpoints has all marked, scanning x-axis, if there is one " 1 ", and in several positions of later in relation, major part is also " 1 ", then think that of set is the horizontal ordinate of required straight-line segment starting point.
2.
a kind of match sideline video detecting method according to claim 1, it is characterized in that described 2) bianry image obtained in the previous step is scanned, often run into the pixel of a non-zero, just calculate this polar value corresponding to point, then the counter corresponding to calculated polar value is added 1.
3.
according to claim 2, a kind of match sideline video detecting method, is characterized in that described 4) adopt Inverse Projection to determine the straight line in former figure; The end points of qualified straight line in the first computing counter of the first step; Size according to the value of parameter decides the direction of pixel projection near straight line, and larger value projects to y-axis, and less value then projects to x-axis; To invest the point of x-axis, if the distance between these point and straight lines is enough little, then the position in the x-axis of correspondence will puts 1, otherwise set to 0;
5) after whole subpoints has all marked, scanning x-axis, if there is one " 1 ", and major part is also " 1 " in several positions of later in relation, then think that of set is the horizontal ordinate of required straight-line segment starting point, obtain its ordinate bringing straight-line equation into, obtain an end points of required line segment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410623320.2A CN105574842A (en) | 2014-11-09 | 2014-11-09 | Match side line video detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410623320.2A CN105574842A (en) | 2014-11-09 | 2014-11-09 | Match side line video detection method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105574842A true CN105574842A (en) | 2016-05-11 |
Family
ID=55884936
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410623320.2A Pending CN105574842A (en) | 2014-11-09 | 2014-11-09 | Match side line video detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105574842A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI584228B (en) * | 2016-05-20 | 2017-05-21 | 銘傳大學 | Method of capturing and reconstructing court lines |
| CN115379215A (en) * | 2018-06-03 | 2022-11-22 | Lg电子株式会社 | Method for decoding, encoding and transmitting video signal and medium for storing video signal |
-
2014
- 2014-11-09 CN CN201410623320.2A patent/CN105574842A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI584228B (en) * | 2016-05-20 | 2017-05-21 | 銘傳大學 | Method of capturing and reconstructing court lines |
| CN115379215A (en) * | 2018-06-03 | 2022-11-22 | Lg电子株式会社 | Method for decoding, encoding and transmitting video signal and medium for storing video signal |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11643076B2 (en) | Forward collision control method and apparatus, electronic device, program, and medium | |
| KR20180042254A (en) | Systems and methods for object tracking | |
| CN103106659A (en) | Open area target detection and tracking method based on binocular vision sparse point matching | |
| CN103065325B (en) | A kind of method for tracking target based on the polymerization of color Distance geometry Iamge Segmentation | |
| CN104134222A (en) | Traffic flow monitoring image detecting and tracking system and method based on multi-feature fusion | |
| CN111160291B (en) | Human eye detection method based on depth information and CNN | |
| CN103605967A (en) | Subway fare evasion prevention system and working method thereof based on image recognition | |
| CN105260749A (en) | Real-time target detection method based on oriented gradient two-value mode and soft cascade SVM | |
| CN105303153A (en) | Vehicle license plate identification method and apparatus | |
| EP2813973A1 (en) | Method and system for processing video image | |
| CN105426868A (en) | Lane detection method based on adaptive region of interest | |
| CN104537688A (en) | Moving object detecting method based on background subtraction and HOG features | |
| CN104952256A (en) | Video information based method for detecting vehicles at intersection | |
| CN104616006A (en) | Surveillance video oriented bearded face detection method | |
| CN102610104B (en) | Onboard front vehicle detection method | |
| CN103413149A (en) | Method for detecting and identifying static target in complicated background | |
| Sun et al. | A robust lane detection method for autonomous car-like robot | |
| CN103745236A (en) | Texture image identification method and texture image identification device | |
| Wang et al. | Geometry constraints-based visual rail track extraction | |
| CN105574842A (en) | Match side line video detection method | |
| Wang et al. | Construction research on multi-threshold segmentation based on improved Otsu threshold method | |
| Marinas et al. | Detection and tracking of traffic signs using a recursive Bayesian decision framework | |
| Yinka et al. | Performance of drivable path detection system of autonomous robots in rain and snow scenario | |
| CN103426185A (en) | Method and device for adjusting target scale in pan-tilt-zoom (PTZ) tracking process | |
| Le et al. | Pedestrian lane detection in unstructured environments for assistive navigation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20160511 |