CN110069966A - A kind of image rapid identification method based on robot - Google Patents
A kind of image rapid identification method based on robot Download PDFInfo
- Publication number
- CN110069966A CN110069966A CN201810070017.2A CN201810070017A CN110069966A CN 110069966 A CN110069966 A CN 110069966A CN 201810070017 A CN201810070017 A CN 201810070017A CN 110069966 A CN110069966 A CN 110069966A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- target
- robot
- color
- 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
- 238000000034 method Methods 0.000 title claims abstract description 28
- 230000011218 segmentation Effects 0.000 claims abstract description 17
- 238000003909 pattern recognition Methods 0.000 claims abstract description 6
- 238000001228 spectrum Methods 0.000 claims description 6
- 238000012986 modification Methods 0.000 claims description 3
- 230000004048 modification Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 description 9
- 230000006870 function Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000002310 reflectometry Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/24—Aligning, centring, orientation detection or correction of the image
- G06V10/245—Aligning, centring, orientation detection or correction of the image by locating a pattern; Special marks for positioning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
-
- 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
- G06T2207/10024—Color image
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Manipulator (AREA)
Abstract
The present invention relates to a kind of image rapid identification method based on robot, the robot is provided with pattern recognition device, and the image rapid identification method based on robot includes: the image information of the target of described image identification device acquisition tracking;Image region segmentation is carried out to the target image information, initial seed point pixel is generated to the color of the target of tracking;Established pixel tree is grown by starting seed point pixel by level traversal, region growing is completed;Export recognition result.Image rapid identification method provided by the invention based on robot can quickly find target object, and can relatively accurately be partitioned into target and its edge, not substantially increased the speed and accuracy of robot identification target by the interference of background substantially.
Description
Technical field
The present invention relates to robotic technology field more particularly to a kind of image rapid identification methods based on robot.
Background technique
The development of information technology, information technology are widely used in people's lives mode and working method.If any
Existing information technology is more effectively utilized, people never stop exploring.
Robot field be collection computer, machinery, sensing technology, the information processing technology, image procossing and identification technology,
Language identification and processing technique, control technology and the communication technology etc. are in the system of one.
Currently, most intelligent robot has a vision collecting function, and view-based access control model acquires function, and robot may be implemented pair
The image recognition of current object to be identified.Common image recognition processes are to pre-save sample image data, are carrying out image
The image data of object to be identified is done into matching search to obtain corresponding with all sample image datas saved when identification
Sample image data, using the corresponding iamge description of sample image data as recognition result.
In above-mentioned image recognition processes, key point is of sample image data Yu object image data to be identified
Match.But since the image data of sample image data and object to be identified is all planar image data, sample image data
The only image recording of some angle of object, when carrying out Image Acquisition to object from another angle, even same object,
The image data of acquisition is also different from the sample image data finally obtained.
Therefore, how quickly tracking target to be separated from background, is improved to the accurate of tracking target identification
Rate becomes those skilled in the art's technical problem urgently to be resolved.
Summary of the invention
The object of the present invention is to provide a kind of image rapid identification methods based on robot, will can quickly track target
It is separated from background, improves the accuracy rate to tracking target identification.
To achieve the goals above, the present invention provides a kind of image rapid identification method based on robot, the machines
Device people is provided with pattern recognition device and color identification device, and the image rapid identification method based on robot includes:
Step S01, described image identification device intercepts target image in advance, and is filtered to it;
Step S02, frequency spectrum threshold value is preset, selection target vision intermediate frequency spectrum is more than the region of preset threshold;If being more than, step is executed
Otherwise rapid S03 continues to execute step S02;
Step S03, image region segmentation is carried out to the target image information, initial seed is generated to the color of the target of tracking
Point pixel;
Step S04, the described color identification device according to the HSL color value of seed point pixel, by the target image color of interception with
Preset color is compared, and setting compares threshold value and thens follow the steps S05 if being more than, and otherwise continues to execute step S04;
Step S05, two labels are added to each pixel in image-region, one is for judging whether pixel grows
Growth label, another be for judge pixel whether be edge edge label, under initial situation, two labels are
0;
Step S06, it establishes a sub-pixel point queue and empties, in pretreated image, it is highest to choose gray value
Point is used as sub-pixel point, sub-pixel point is included in queue, and two labels of the sub-pixel point are set to 0;
Step S07, the sub-pixel point in queue is chosen, the label of the adjacent pixel of sub-pixel point is judged, if adjacent
The growth label of pixel be 0, then judge whether the gray value of the adjacent pixel meets preset growth conditions, if raw
Long label is non-zero, then does not consider the point;If the adjacent pixel meets growth conditions, the adjacent pixel is added
Enqueue, and two labels of the adjacent pixel are set to 0;
Step S08, so circulation is until all pixels are by judgement;
Step S09, whether the edge label for judging sub-pixel point is 0, if the edge label of sub-pixel point is then set 0,
If otherwise without modification;Delete the header element in queue;
Step S10, established pixel tree is grown by starting seed point pixel by level traversal;Export recognition result.
Preferably, in step S01, described image identification device is camera.
Preferably, in step S01, described image identification device is omni-directional vison sensor, the omni-directional vison sensor
It is made of panorama reflecting mirror and camera.
It is preferably, described that image region segmentation is carried out to the target image information in step S03, comprising:
Step S031, gray level thresholding segmentation is carried out to the target image information, obtains the bianry image of picture;
Step S032, the minimum circumscribed rectangle of the picture is calculated according to the bianry image of the picture and obtains the external square
The two-dimensional coordinate of shape obtains including outside the minimum of the picture by the two dimensional coordinate map into corresponding RGB color image
Connect rectangle;
Step S033, color segmentation is carried out to the minimum circumscribed rectangle, obtains the color bianry image of minimum circumscribed rectangle;
Step S034, mesh is partitioned into from described image information by the bianry image of the picture and the color bianry image
Logo image, then the target image is divided into several image subsets.
It preferably, further include the target to tracking before the progress image region segmentation to the target image information
Colorfulness threshold value is set, the colorfulness of image-region is located in the threshold range, is just split to the image-region.
Preferably, in step S03, the color of the target of described pair of tracking generates initial seed point pixel, comprising: to image
It is pre-processed, choosing color angle value in image is the pixel of preset value as sub-pixel point.
Image rapid identification method provided by the invention based on robot can quickly find target object, and energy
It is relatively accurately partitioned into target and its edge, is not substantially increased the speed of robot identification target by the interference of background substantially
Degree and accuracy.
Detailed description of the invention
Fig. 1 is that a kind of process of embodiment of the image rapid identification method provided by the invention based on robot is illustrated
Figure.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application is carried out clearly and completely
Description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.Based on this Shen
Please in embodiment, every other reality obtained by those of ordinary skill in the art without making creative efforts
Example is applied, shall fall in the protection scope of this application.
Referring to FIG. 1, Fig. 1 is a kind of embodiment of the image rapid identification method provided by the invention based on robot
Flow diagram.
As shown in Figure 1, the present invention provides a kind of image rapid identification method based on robot, the robot setting
There are pattern recognition device and color identification device, the image rapid identification method based on robot includes:
Step S01, described image identification device intercepts target image in advance, and is filtered to it;
Step S02, frequency spectrum threshold value is preset, selection target vision intermediate frequency spectrum is more than the region of preset threshold;If being more than, step is executed
Otherwise rapid S03 continues to execute step S02;
Step S03, image region segmentation is carried out to the target image information, initial seed is generated to the color of the target of tracking
Point pixel;
Step S04, the described color identification device according to the HSL color value of seed point pixel, by the target image color of interception with
Preset color is compared, and setting compares threshold value and thens follow the steps S05 if being more than, and otherwise continues to execute step S04;
Step S05, two labels are added to each pixel in image-region, one is for judging whether pixel grows
Growth label, another be for judge pixel whether be edge edge label, under initial situation, two labels are
0;
Step S06, it establishes a sub-pixel point queue and empties, in pretreated image, it is highest to choose gray value
Point is used as sub-pixel point, sub-pixel point is included in queue, and two labels of the sub-pixel point are set to 0;
Step S07, the sub-pixel point in queue is chosen, the label of the adjacent pixel of sub-pixel point is judged, if adjacent
The growth label of pixel be 0, then judge whether the gray value of the adjacent pixel meets preset growth conditions, if raw
Long label is non-zero, then does not consider the point;If the adjacent pixel meets growth conditions, the adjacent pixel is added
Enqueue, and two labels of the adjacent pixel are set to 0;
Step S08, so circulation is until all pixels are by judgement;
Step S09, whether the edge label for judging sub-pixel point is 0, if the edge label of sub-pixel point is then set 0,
If otherwise without modification;Delete the header element in queue;
Step S10, established pixel tree is grown by starting seed point pixel by level traversal;Export recognition result.
Image rapid identification method provided by the invention based on robot can quickly find target object, and energy
It is relatively accurately partitioned into target and its edge, is not substantially increased the speed of robot identification target by the interference of background substantially
Degree and accuracy.
In preferred scheme, described image identification device can be camera, and the figure of tracking target is obtained by camera
As information.
In preferred scheme, described image identification device can also be omni-directional vison sensor, the omni-directional vison sensing
Device is made of panorama reflecting mirror and camera.Panorama reflecting mirror is the conical surface with certain curvature, being capable of 360 ° omni-directional visitor
Ground is seen by scene reflectivity to camera, the image information of available more accurate tracking target.
It is described that image region segmentation is carried out to the target image information, comprising: to the target figure in preferred scheme
As information progress gray level thresholding segmentation, the bianry image of picture is obtained;The figure is calculated according to the bianry image of the picture
The minimum circumscribed rectangle of piece and the two-dimensional coordinate for obtaining the boundary rectangle, the two dimensional coordinate map is color to corresponding RGB
In chromatic graph picture, obtain include the picture minimum circumscribed rectangle;Color segmentation is carried out to the minimum circumscribed rectangle, is obtained most
The color bianry image of small boundary rectangle;Believed by the bianry image and the color bianry image of the picture from described image
It is partitioned into target image in breath, then the target image is divided into several image subsets.
It further include to tracking before the progress image region segmentation to the target image information in preferred scheme
The colorfulness of target threshold value is set, the colorfulness of image-region is located in the threshold range, just to image-region progress
Segmentation.In actual scene, due to the variation of the external environmental factors such as illumination condition, when the value range of each color is prosperous not
It is fixed, it the phenomenon that often will appear juxtaposition between the color value of various colors, is set by the colorfulness of the target to tracking
Set threshold value, the colorfulness of image-region is located in the threshold range, is just split to the image-region, could correctly by
The target tracked carries out the identification of color.
In preferred scheme, the color of the target of described pair of tracking generates initial seed point pixel, comprising: carries out to image
Pretreatment, choosing color angle value in image is the pixel of preset value as sub-pixel point.
Structure, feature and effect of the invention, the above institute are described in detail based on the embodiments shown in the drawings
Only presently preferred embodiments of the present invention is stated, but the present invention does not limit the scope of implementation as shown in the drawings, it is all according to structure of the invention
Think made change or equivalent example modified to equivalent change, when not going beyond the spirit of the description and the drawings,
It should all be within the scope of the present invention.
Claims (6)
1. a kind of image rapid identification method based on robot, the robot is provided with pattern recognition device and color is known
Other device, which is characterized in that the image rapid identification method based on robot includes:
Step S01, described image identification device intercepts target image in advance, and is filtered to it;
Step S02, frequency spectrum threshold value is preset, selection target vision intermediate frequency spectrum is more than the region of preset threshold;If being more than, step is executed
Otherwise rapid S03 continues to execute step S02;
Step S03, image region segmentation is carried out to the target image information, initial seed is generated to the color of the target of tracking
Point pixel;
Step S04, the described color identification device according to the HSL color value of seed point pixel, by the target image color of interception with
Preset color is compared, and setting compares threshold value and thens follow the steps S05 if being more than, and otherwise continues to execute step S04;
Step S05, two labels are added to each pixel in image-region, one is for judging whether pixel grows
Growth label, another be for judge pixel whether be edge edge label, under initial situation, two labels are
0;
Step S06, it establishes a sub-pixel point queue and empties, in pretreated image, it is highest to choose gray value
Point is used as sub-pixel point, sub-pixel point is included in queue, and two labels of the sub-pixel point are set to 0;
Step S07, the sub-pixel point in queue is chosen, the label of the adjacent pixel of sub-pixel point is judged, if adjacent
The growth label of pixel be 0, then judge whether the gray value of the adjacent pixel meets preset growth conditions, if raw
Long label is non-zero, then does not consider the point;If the adjacent pixel meets growth conditions, the adjacent pixel is added
Enqueue, and two labels of the adjacent pixel are set to 0;
Step S08, so circulation is until all pixels are by judgement;
Step S09, whether the edge label for judging sub-pixel point is 0, if the edge label of sub-pixel point is then set 0,
If otherwise without modification;Delete the header element in queue;
Step S10, established pixel tree is grown by starting seed point pixel by level traversal;Export recognition result.
2. the image rapid identification method according to claim 1 based on robot, which is characterized in that in step S01, institute
Stating pattern recognition device is camera.
3. the image rapid identification method according to claim 1 based on robot, which is characterized in that in step S01, institute
Stating pattern recognition device is omni-directional vison sensor, and the omni-directional vison sensor is made of panorama reflecting mirror and camera.
4. the image rapid identification method according to claim 1-3 based on robot, which is characterized in that step
It is described that image region segmentation is carried out to the target image information in S03, comprising:
Step S031, gray level thresholding segmentation is carried out to the target image information, obtains the bianry image of picture;
Step S032, the minimum circumscribed rectangle of the picture is calculated according to the bianry image of the picture and obtains the external square
The two-dimensional coordinate of shape obtains including outside the minimum of the picture by the two dimensional coordinate map into corresponding RGB color image
Connect rectangle;
Step S033, color segmentation is carried out to the minimum circumscribed rectangle, obtains the color bianry image of minimum circumscribed rectangle;
Step S034, mesh is partitioned into from described image information by the bianry image of the picture and the color bianry image
Logo image, then the target image is divided into several image subsets.
5. the image rapid identification method according to claim 4 based on robot, which is characterized in that described to the mesh
It further include that threshold value is arranged to the colorfulness of the target of tracking before logo image information carries out image region segmentation, image-region
Colorfulness is located in the threshold range, is just split to the image-region.
6. the image rapid identification method according to claim 1 based on robot, which is characterized in that in step S03, institute
It states and initial seed point pixel is generated to the color of the target of tracking, comprising: image is pre-processed, colorfulness in image is chosen
Value is the pixel of preset value as sub-pixel point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810070017.2A CN110069966A (en) | 2018-01-24 | 2018-01-24 | A kind of image rapid identification method based on robot |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810070017.2A CN110069966A (en) | 2018-01-24 | 2018-01-24 | A kind of image rapid identification method based on robot |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110069966A true CN110069966A (en) | 2019-07-30 |
Family
ID=67365592
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810070017.2A Pending CN110069966A (en) | 2018-01-24 | 2018-01-24 | A kind of image rapid identification method based on robot |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110069966A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113418509A (en) * | 2021-05-20 | 2021-09-21 | 中国农业科学院烟草研究所(中国烟草总公司青州烟草研究所) | Automatic target-aiming detection device and detection method for agriculture |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104978558A (en) * | 2014-04-11 | 2015-10-14 | 北京数码视讯科技股份有限公司 | Target identification method and device |
-
2018
- 2018-01-24 CN CN201810070017.2A patent/CN110069966A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104978558A (en) * | 2014-04-11 | 2015-10-14 | 北京数码视讯科技股份有限公司 | Target identification method and device |
Non-Patent Citations (1)
Title |
---|
谭宝成等: "全自主移动机器人视觉系统图像分割方法研究", 《西安工业大学学报》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113418509A (en) * | 2021-05-20 | 2021-09-21 | 中国农业科学院烟草研究所(中国烟草总公司青州烟草研究所) | Automatic target-aiming detection device and detection method for agriculture |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wang et al. | Semantic segmentation of crop and weed using an encoder-decoder network and image enhancement method under uncontrolled outdoor illumination | |
CN111797653B (en) | Image labeling method and device based on high-dimensional image | |
WO2017155602A1 (en) | Systems and methods for normalizing an image | |
WO2013088175A1 (en) | Image processing method | |
CN110021029B (en) | Real-time dynamic registration method and storage medium suitable for RGBD-SLAM | |
CN107844736A (en) | iris locating method and device | |
JP7142420B2 (en) | Image processing device, learning method, trained model, image processing method | |
CN110443247A (en) | A kind of unmanned aerial vehicle moving small target real-time detecting system and method | |
CN112883881B (en) | Unordered sorting method and unordered sorting device for strip-shaped agricultural products | |
CN107527368A (en) | Three-dimensional attitude localization method and device based on Quick Response Code | |
CN101996317B (en) | Method and device for identifying markers in human body | |
CN108274476B (en) | Method for grabbing ball by humanoid robot | |
CN114387505A (en) | Hyperspectral and laser radar multi-modal remote sensing data classification method and system | |
CN113989604A (en) | Tire DOT information identification method based on end-to-end deep learning | |
CN101447027A (en) | Binaryzation method of magnetic code character area and application thereof | |
CN113689365B (en) | Target tracking and positioning method based on Azure Kinect | |
CN113469195B (en) | Target identification method based on self-adaptive color quick point feature histogram | |
CN116843581B (en) | Image enhancement method, system, device and storage medium for multi-scene graph | |
CN110069966A (en) | A kind of image rapid identification method based on robot | |
CN111310703B (en) | Identity recognition method, device, equipment and medium based on convolutional neural network | |
CN112489049A (en) | Mature tomato fruit segmentation method and system based on superpixels and SVM | |
Li et al. | [Retracted] Image Processing and Recognition Algorithm Design in Intelligent Imaging Device System | |
Wu | Review of theory and methods of image segmentation | |
CN109918974A (en) | A kind of robot target recognition methods | |
CN110070553A (en) | A kind of matching process of robot vision image |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190730 |
|
RJ01 | Rejection of invention patent application after publication |