CN105976396A - Method used for identifying moving object on conveyor belt in video image - Google Patents
Method used for identifying moving object on conveyor belt in video image Download PDFInfo
- Publication number
- CN105976396A CN105976396A CN201610271848.7A CN201610271848A CN105976396A CN 105976396 A CN105976396 A CN 105976396A CN 201610271848 A CN201610271848 A CN 201610271848A CN 105976396 A CN105976396 A CN 105976396A
- Authority
- CN
- China
- Prior art keywords
- edge contour
- contour path
- path
- image
- point
- 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
Classifications
-
- 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
Landscapes
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a method used for identifying a moving object on a conveyor belt in a video image. The method is characterized by carrying out target outline identification and modeling processing on a previous frame of image information; setting a start point coordinate and an end point coordinate of a detection area and an allowable displacement deviation value; taking out edge outline paths one by one, acquiring a center point and determining whether to enter into the detection area, and taking an identification which is entered into the detection area as a monitoring point and storing to a monitoring point queue; using a displacement of center points of previous and following frames of images to confirm the edge outline path which needs to be calculated, and at an end terminal of the detection area, and when the edge outline path which needs to be calculated passes by, identifying the edge outline path to be the edge outline path which needs to be counted. By using the method used for identifying the moving object on the conveyor belt in the video image, the moving object on the conveyor belt in the video image can be rapidly and effectively identified so that the identified moving object is calculated and analyzed.
Description
Technical field
The present invention relates to a kind of method identifying moving object, more specifically, particularly relate to one and exist for video image identification
The method of moving object on conveyer belt.
Background technology
Traditional belt feeder enumerator is to use infrared sensor to be fixed on the both sides of conveyer belt, and infrared ray, another is launched in one end
One end receives this infrared signal, when target item passes through to block infrared ray from centre, can count by flip-flop number.
According to the operation principle of above-mentioned enumerator, the detected the biggest part of target item identified, and pass in conveyer belt
It is single channel when sending, only target item is passed sequentially through singly infrared sensor and can identify counting one by one.If
Detected target item volume is the thinnest and quantity is the most, if when these target items, batch transmits on a moving belt, infrared
Sensor can not carry out batch identification to it, and traditional belt feeder enumerator can not be identified and calculate.Nowadays, with meter
Calculate the high speed development of machine technology, it would be highly desirable to invent a kind of method carrying out video image identification for the small objects of motion, to meet
The production of modern processing manufacturing industry needs.
Summary of the invention
It is an object of the invention to provide a kind of method for video image identification moving object on a moving belt, utilize the method
The moving object that can identify fast and effectively in video image on conveyer belt, in order to the moving object to identifying is united
Meter is analyzed.
Technical scheme is as follows:
A kind of method for video image identification moving object on a moving belt, including processing step as follows:
(1) after target outline identification, thing is obtained after extracting the image information of former frame in image queue from image memory device
Body closed contour image；
(2) the object closed contour image of traversal step (1) gained, judges whether edge contour path meets object shape one by one
Shape feature, stores the edge contour path meeting body form feature to image block queue；
(3) in every frame image information, direction of transfer along conveyer belt sets the detection starting point coordinate in region and end point coordinate,
Shooting speed and the transfer rate of transmission belt according to image information determine the displacement tolerance of central point between before and after two two field picture
Value；
(4) the edge contour path in the image block queue obtained after completing step (2) is taken out one by one, obtains edge wheel
The central point in wide path, it is judged that whether the central point in edge contour path enters detection region, if the central point in edge contour path
Have been enter into detecting region, then the central point in this edge contour path is designated control point and stores to control point queue, otherwise this limit
The central point of edge contour path does not deals with；
(5) step (1) and step (2) are repeated, it is thus achieved that the image block queue of next frame, and by edge wheel in image block queue
The central point asking edge contour path is taken out in wide path one by one, it is judged that whether the central point in edge contour path enters detection region,
If the central point in edge contour path has been enter into detecting region, then enter the next one and process step, otherwise jump out and from image block team
Row take next edge contour path centering point；
(6) take one by one with the control point in control point queue through having passed through the central point in the edge contour path of step (5) gained
Go out contrast, allow partially if the displacement difference in the central point in this edge contour path and control point queue between a certain control point is in displacement
Difference range, then enter the next one and process step, otherwise jump out and take next edge contour path centering from image block queue
Point；
(7) the end point coordinate of the central point and detection region that have passed through step (6) gained edge contour path is made comparisons,
If the center point coordinate in this edge contour path is not above detecting the end point coordinate in region, then by this edge contour path
Heart point coordinates value replaces control point coordinate figure, is otherwise the edge contour path needing to count by this edge contour Path Recognition.
Preferably, in step (2), minimum widith, minimum constructive height and image block are made in the described edge contour path extracted
Area filters and judges, meeting the edge contour path required is the edge contour path meeting object profile feature.
Preferably, minimum widith, minimum constructive height and image block area are made in the described edge contour path extracted and filter judgement,
Concrete processing method is: the edge contour path computing of extraction goes out pixel wide, pixels tall and closed contour area, then will
Calculate the most corresponding minimum pixel width with setting of the pixel wide in closed contour path, pixels tall and closed contour area
Angle value, minimum pixel height value and minimum image area value are made comparisons.
Preferably, in step (2), calculate contour of object width pixel value, object wheel respectively according to the true form of object
Wide height pixel value and contour of object area, described minimum pixel widths value is 0.8 times of contour of object width pixel value,
Described minimum pixel height value is 0.9 times of contour of object height pixel value, and described minimum image area value is contour of object
0.8 times of area.
Preferably, in step (3), described displacement tolerance value is that X-axis allows absolute value deviation 10 pixel, Y-axis
Allow overgauge 120 pixel.
Compared with prior art, the device have the advantages that into:
A kind of method for video image identification moving object on a moving belt of the present invention, first extracts from image memory device
Image queue carries out after the image information of former frame after objective contour identification obtaining object closed contour image.Judgment object one by one
Whether the edge contour path of closed contour image meets body form feature, stores the edge contour path met to image block
Queue.In image information, set the detection starting point coordinate in region and end point coordinate, and determine between before and after two two field picture
The displacement tolerance value of heart point.The edge contour path taken out one by one in image block queue obtains central point, it is judged that edge contour
Whether the central point in path enters detection region, if having been enter into detecting region, is then designated by the central point in this edge contour path
Control point stores to control point queue, and otherwise the central point in this edge contour path does not deals with.Obtain the image block team of next frame
Row, and the central point in edge contour path is sought in taking-up edge contour path one by one, it is judged that whether the central point in edge contour path enters
Enter to detect region, the edge contour path center point having been enter into detecting region contrasted one by one with the control point in control point queue,
And if in control point queue, displacement difference between a certain control point is in displacement tolerance value scope, then by this edge contour path
Central point compares with the end point in detection region, if being not above detecting the end point coordinate in region, then by this edge contour path
Center point coordinate value replace control point coordinate figure, be otherwise to need the edge contour path of counting by this edge contour Path Recognition.
The method for video image identification moving object on a moving belt of the present invention can identify fast and effectively in video image and pass
Send the moving object on band, in order to the moving object identified is carried out statistical analysis.
Accompanying drawing explanation
Fig. 1 is the control flow chart of the present invention.
Detailed description of the invention
Below in conjunction with detailed description of the invention, technical scheme is described in further detail, but does not constitute the present invention
Any restriction.
With reference to shown in Fig. 1, a kind of method for video image identification moving object on a moving belt of the present invention, including as follows
Process step:
(1) after target outline identification, thing is obtained after extracting the image information of former frame in image queue from image memory device
Body closed contour image, processes the carrying out the most one by one of the image information in image queue.
(2) the object closed contour image of traversal step (1) gained, judges whether edge contour path meets object shape one by one
Shape feature, stores the edge contour path meeting body form feature to image block queue.
Wherein, the described edge contour path extracted is made minimum widith, minimum constructive height and image block area and filters judgement, full
The edge contour path that foot requires is the edge contour path meeting object profile feature.The described edge contour path extracted is made
Minimum widith, minimum constructive height and image block area filter and judge, concrete processing method is: counted respectively in this closed contour path
Calculate pixel wide, pixels tall and closed contour area, then the pixel wide in closed contour path, pixel height will be calculated
Degree and the most corresponding minimum pixel widths value, minimum pixel height value and the minimum image area value with setting of closed contour area
Make comparisons.True form according to object calculates contour of object width pixel value, contour of object height pixel value and thing respectively
Body contour area, described minimum pixel widths value is 0.8 times of contour of object width pixel value, described minimum pixel height
Value is 0.9 times of contour of object height pixel value, described minimum image area value is contour of object area 0.8 times, will be
Pixel widths value, minimum pixel height value and minimum image area value are set as the certain of the pixel value of contour of object character pair
Multiple value time, identify object effect best.
(3) in every frame image information, direction of transfer along conveyer belt sets the detection starting point coordinate in region and end point coordinate,
Shooting speed and the transfer rate of transmission belt according to image information determine the displacement tolerance of central point between before and after two two field picture
Value.To the frame difference image produced with 60 frames/second speed and the transfer rate per second of conveyer belt, image block center can be calculated
Between point coordinates two frame, X-axis allows the absolute value deviation of 10 pixels, and Y-axis is the overgauge that the direction of motion allows 120 pixels, its
Middle permission X-axis deviation causes mainly due to the vibrations of camera head certainty of measurement or conveyer belt, and two frame deviation values of Y-axis
Relevant with the movement velocity of conveyer belt.
(4) the edge contour path in the image block queue obtained after completing step (2) is taken out one by one, obtains edge wheel
The central point in wide path, it is judged that whether the central point in edge contour path enters detection region, if the central point in edge contour path
Have been enter into detecting region, then the central point in this edge contour path is designated control point and stores to control point queue, otherwise this limit
The central point of edge contour path does not deals with.
(5) step (1) and step (2) are repeated, it is thus achieved that the image block queue of next frame, and by edge wheel in image block queue
The central point asking edge contour path is taken out in wide path one by one, it is judged that whether the central point in edge contour path enters detection region,
If the central point in edge contour path has been enter into detecting region, then enter the next one and process step, otherwise jump out and from image block team
Row take next edge contour path centering point.
(6) take one by one with the control point in control point queue through having passed through the central point in the edge contour path of step (5) gained
Go out contrast, allow partially if the displacement difference in the central point in this edge contour path and control point queue between a certain control point is in displacement
Difference range, then enter the next one and process step, otherwise jump out and take next edge contour path centering from image block queue
Point.
(7) the end point coordinate of the central point and detection region that have passed through step (6) gained edge contour path is made comparisons,
If the center point coordinate in this edge contour path is not above detecting the end point coordinate in region, then by this edge contour path
Heart point coordinates value replaces control point coordinate figure, is otherwise the edge contour path needing to count by this edge contour Path Recognition.
A kind of method for video image identification moving object on a moving belt of the present invention, first extracts from image memory device
Image queue carries out after the image information of former frame after objective contour identification obtaining object closed contour image.Judgment object one by one
Whether the edge contour path of closed contour image meets body form feature, stores the edge contour path met to image block
Queue.In image information, set the detection starting point coordinate in region and end point coordinate, and determine between before and after two two field picture
The displacement tolerance value of heart point.The edge contour path taken out one by one in image block queue obtains central point, it is judged that edge contour
Whether the central point in path enters detection region, if having been enter into detecting region, is then designated by the central point in this edge contour path
Control point stores to control point queue, and otherwise the central point in this edge contour path does not deals with.Obtain the image block team of next frame
Row, and the central point in edge contour path is sought in taking-up edge contour path one by one, it is judged that whether the central point in edge contour path enters
Enter to detect region, the edge contour path center point having been enter into detecting region contrasted one by one with the control point in control point queue,
And if in control point queue, displacement difference between a certain control point is in displacement tolerance value scope, then by this edge contour path
Central point compares with the end point in detection region, if being not above detecting the end point coordinate in region, then by this edge contour path
Center point coordinate value replace control point coordinate figure, be otherwise to need the edge contour path of counting by this edge contour Path Recognition.
The method for video image identification moving object on a moving belt of the present invention can identify fast and effectively in video image and pass
Send the moving object on band, in order to the moving object identified is carried out statistical analysis.
The foregoing is only presently preferred embodiments of the present invention, all made in the range of the spirit and principles in the present invention any amendment,
Equivalent and improvement etc., should be included within the scope of the present invention.
Claims (5)
1. the method for video image identification moving object on a moving belt, it is characterised in that include processing as follows step:
(1) after target outline identification, thing is obtained after extracting the image information of former frame in image queue from image memory device
Body closed contour image；
(2) the object closed contour image of traversal step (1) gained, judges whether edge contour path meets object shape one by one
Shape feature, stores the edge contour path meeting body form feature to image block queue；
(3) in every frame image information, direction of transfer along conveyer belt sets the detection starting point coordinate in region and end point coordinate,
Shooting speed and the transfer rate of transmission belt according to image information determine the displacement tolerance of central point between before and after two two field picture
Value；
(4) the edge contour path in the image block queue obtained after completing step (2) is taken out one by one, obtains edge wheel
The central point in wide path, it is judged that whether the central point in edge contour path enters detection region, if the central point in edge contour path
Have been enter into detecting region, then the central point in this edge contour path is designated control point and stores to control point queue, otherwise this limit
The central point of edge contour path does not deals with；
(5) step (1) and step (2) are repeated, it is thus achieved that the image block queue of next frame, and by edge wheel in image block queue
The central point asking edge contour path is taken out in wide path one by one, it is judged that whether the central point in edge contour path enters detection region,
If the central point in edge contour path has been enter into detecting region, then enter the next one and process step, otherwise jump out and from image block team
Row take next edge contour path centering point；
(6) take one by one with the control point in control point queue through having passed through the central point in the edge contour path of step (5) gained
Go out contrast, allow partially if the displacement difference in the central point in this edge contour path and control point queue between a certain control point is in displacement
Difference range, then enter the next one and process step, otherwise jump out and take next edge contour path centering from image block queue
Point；
(7) the end point coordinate of the central point and detection region that have passed through step (6) gained edge contour path is made comparisons,
If the center point coordinate in this edge contour path is not above detecting the end point coordinate in region, then by this edge contour path
Heart point coordinates value replaces control point coordinate figure, is otherwise the edge contour path needing to count by this edge contour Path Recognition.
Method for video image identification moving object on a moving belt the most according to claim 1, it is characterised in that
In step (2), minimum widith, minimum constructive height and image block area are made in the described edge contour path extracted and filter judgement,
Meeting the edge contour path required is the edge contour path meeting object profile feature.
Method for video image identification moving object on a moving belt the most according to claim 2, it is characterised in that
Minimum widith, minimum constructive height and image block area are made in the described edge contour path extracted and filters judgement, concrete processing method
For: the edge contour path computing of extraction is gone out pixel wide, pixels tall and closed contour area, then Guan Bi will be calculated
The pixel wide of contour path, pixels tall and closed contour area correspondence respectively and the minimum pixel widths value set, minimum image
Element height value and minimum image area value are made comparisons.
Method for video image identification moving object on a moving belt the most according to claim 3, it is characterised in that
In step (2), calculate contour of object width pixel value, contour of object height pixel value respectively according to the true form of object
And contour of object area, described minimum pixel widths value is 0.8 times of contour of object width pixel value, described minimum image
Element height value is 0.9 times of contour of object height pixel value, described minimum image area value is contour of object area 0.8 times.
Method for video image identification moving object on a moving belt the most according to claim 1, it is characterised in that
In step (3), described displacement tolerance value is that X-axis allows absolute value deviation 10 pixel, and Y-axis allows overgauge 120
Pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610271848.7A CN105976396A (en) | 2016-04-27 | 2016-04-27 | Method used for identifying moving object on conveyor belt in video image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610271848.7A CN105976396A (en) | 2016-04-27 | 2016-04-27 | Method used for identifying moving object on conveyor belt in video image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105976396A true CN105976396A (en) | 2016-09-28 |
Family
ID=56994051
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610271848.7A Pending CN105976396A (en) | 2016-04-27 | 2016-04-27 | Method used for identifying moving object on conveyor belt in video image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105976396A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529528A (en) * | 2016-09-30 | 2017-03-22 | 浙江宇视科技有限公司 | Method and equipment for identifying effective moving target |
| CN107103602A (en) * | 2017-04-19 | 2017-08-29 | 凌美芯（北京）科技有限责任公司 | A kind of intelligent material method of counting based on computer vision technique |
| CN108510539A (en) * | 2018-05-09 | 2018-09-07 | 中国矿业大学(北京) | A kind of coal production monitoring method based on image analysis |
| CN110070558A (en) * | 2019-04-11 | 2019-07-30 | 西安星舟天启智能装备有限责任公司 | It is a kind of that method and device is sorted based on the statistics of image procossing and Subarea detecting |
| CN110942481A (en) * | 2019-12-13 | 2020-03-31 | 西南石油大学 | Image processing-based vertical jump detection method |
| CN111487432A (en) * | 2020-04-02 | 2020-08-04 | 精英数智科技股份有限公司 | Belt speed measuring method, system and equipment and computer storage medium |
| CN111524151A (en) * | 2020-04-29 | 2020-08-11 | 深圳市铭特科技有限公司 | Object detection method and system based on background recognition |
| CN114441499A (en) * | 2022-04-11 | 2022-05-06 | 天津美腾科技股份有限公司 | Grade detection method and device, identification equipment, ore pulp grade instrument and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050259847A1 (en) * | 2004-01-29 | 2005-11-24 | Yakup Genc | System and method for tracking parcels on a planar surface |
| CN103150548A (en) * | 2013-01-31 | 2013-06-12 | 南京吉目希自动化科技有限公司 | Method for improving machine vision system identification accuracy |
| CN103426164A (en) * | 2013-06-09 | 2013-12-04 | 大连海事大学 | Scallop dimension calculating method based on Opencv image analysis and scallop sorting system |
| CN104574403A (en) * | 2015-01-12 | 2015-04-29 | 飞天诚信科技股份有限公司 | Intelligent cutting method |
-
2016
- 2016-04-27 CN CN201610271848.7A patent/CN105976396A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050259847A1 (en) * | 2004-01-29 | 2005-11-24 | Yakup Genc | System and method for tracking parcels on a planar surface |
| CN103150548A (en) * | 2013-01-31 | 2013-06-12 | 南京吉目希自动化科技有限公司 | Method for improving machine vision system identification accuracy |
| CN103426164A (en) * | 2013-06-09 | 2013-12-04 | 大连海事大学 | Scallop dimension calculating method based on Opencv image analysis and scallop sorting system |
| CN104574403A (en) * | 2015-01-12 | 2015-04-29 | 飞天诚信科技股份有限公司 | Intelligent cutting method |
Non-Patent Citations (2)
| Title |
|---|
| GRYCUK R 等: "From Single Image to List of Objects Based on Edge and Blob Detection", 《INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND SOFT COMPUTING》 * |
| 王荣 等: "视频流中的单向快速运动多目标可靠跟踪技术", 《电视技术》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106529528A (en) * | 2016-09-30 | 2017-03-22 | 浙江宇视科技有限公司 | Method and equipment for identifying effective moving target |
| CN107103602A (en) * | 2017-04-19 | 2017-08-29 | 凌美芯（北京）科技有限责任公司 | A kind of intelligent material method of counting based on computer vision technique |
| CN108510539A (en) * | 2018-05-09 | 2018-09-07 | 中国矿业大学(北京) | A kind of coal production monitoring method based on image analysis |
| CN110070558A (en) * | 2019-04-11 | 2019-07-30 | 西安星舟天启智能装备有限责任公司 | It is a kind of that method and device is sorted based on the statistics of image procossing and Subarea detecting |
| CN110942481A (en) * | 2019-12-13 | 2020-03-31 | 西南石油大学 | Image processing-based vertical jump detection method |
| CN110942481B (en) * | 2019-12-13 | 2022-05-20 | 西南石油大学 | Image processing-based vertical jump detection method |
| CN111487432A (en) * | 2020-04-02 | 2020-08-04 | 精英数智科技股份有限公司 | Belt speed measuring method, system and equipment and computer storage medium |
| CN111524151A (en) * | 2020-04-29 | 2020-08-11 | 深圳市铭特科技有限公司 | Object detection method and system based on background recognition |
| CN114441499A (en) * | 2022-04-11 | 2022-05-06 | 天津美腾科技股份有限公司 | Grade detection method and device, identification equipment, ore pulp grade instrument and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105976396A (en) | Method used for identifying moving object on conveyor belt in video image | |
| CA3094424C (en) | Safety monitoring and early-warning method for man-machine interaction behavior of underground conveyor belt operator | |
| CN106228546A (en) | Board card detection method and device | |
| CN111242108B (en) | Belt transfer point coal blockage identification method based on target detection | |
| CN108109137A (en) | The Machine Vision Inspecting System and method of vehicle part | |
| CN107063099A (en) | A kind of machinery manufacturing industry online quality monitoring method of view-based access control model system | |
| CN107705563B (en) | Laser radar-based continuous vehicle speed detection method | |
| CN109433641B (en) | Intelligent detection method for tablet capsule filling omission based on machine vision | |
| US20150145963A1 (en) | Stereo Camera | |
| CN113723325A (en) | Tool defect detection system for prefabricated parts | |
| CN104198752A (en) | High temperature steel billet motion state multi-rate detection method based on machine vision | |
| KR102218894B1 (en) | Device for Double Recognition Image Treatment for Conveyor Belt and Method for the Same | |
| CN114372991B (en) | Method and device for collecting and matching X-ray pictures of articles in security inspection process | |
| CN105320930A (en) | Object recognition apparatus | |
| CN115144399B (en) | Assembly quality detection method and device based on machine vision | |
| KR101214858B1 (en) | Moving object detecting apparatus and method using clustering | |
| CN110807354B (en) | Industrial assembly line product counting method | |
| CN110053827A (en) | A kind of cigarette packet three-dimensional appearance visible detection method and device | |
| CN105067836A (en) | Color sorter material automatic recognition and speed measurement algorithm | |
| CN114655655B (en) | Conveyor belt deviation detection method based on UNet network | |
| CN105975910A (en) | Technology processing method used for carrying out video identification on moving object and system thereof | |
| CN107153876A (en) | Machine vision shutter Self-tuning System intelligent control method based on Expert Rules | |
| US11748838B2 (en) | System and method for three-dimensional scan of moving objects longer than the field of view | |
| CN106778675B (en) | A kind of recognition methods of target in video image object and device | |
| CN105488448B (en) | A kind of differentiating method of bar code and Quick Response Code |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160928 |