CN103177259B - Color lump recognition methods - Google Patents
Color lump recognition methods Download PDFInfo
- Publication number
- CN103177259B CN103177259B CN201310125145.XA CN201310125145A CN103177259B CN 103177259 B CN103177259 B CN 103177259B CN 201310125145 A CN201310125145 A CN 201310125145A CN 103177259 B CN103177259 B CN 103177259B
- Authority
- CN
- China
- Prior art keywords
- color lump
- color
- image
- profile
- lump
- 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)
Abstract
The invention provides a kind of color lump recognition methods and device, this color lump recognition methods and device are by carrying out rim detection to the image of needs identification, the profile that extracts color lump in the image that needs identification is determined the shape of color lump, the position that obtains color lump according to the center-of-mass coordinate of the profile calculating color lump of the color lump extracting, the color of choosing a some position in color lump is determined the color of color lump. Therefore it can identify shape, position and the color of color lump simultaneously, thereby can more comprehensively extract the information of color lump.
Description
Technical field
The present invention relates to a kind of image processing field, particularly relate to a kind of color lump recognition methods.
Background technology
Color lump identification is the pith of image recognition, has very wide application prospect, can be applicable to machineThe place identification of device people match, automatic sorting and the classification etc. of mail parcel. But, conventional color lump identificationMethod can only be identified the color of color lump, and cannot obtain the more information of polychrome piece, the shape of for example color lump and positionPut information. Cause like this color lump information that extracts comprehensive not, poor to the identification of color lump.
Summary of the invention
Based on this, be necessary to provide a kind of color lump of shape, position and the color that can simultaneously identify color lump to knowOther method.
A kind of color lump recognition methods, comprises the following steps: color lump shape recognition, the image of needs identification is enteredRow rim detection, extracts the profile of color lump in the image that needs identification, determines the shape of color lump; Color lump positionIdentification, according to the center-of-mass coordinate of the profile calculating color lump of the color lump extracting; The identification of color lump color, chooses shapeAny point in the definite color lump of shape according to the image of the needs identification of described any some positionColor is determined the color of color lump.
Therein in an embodiment, before the step of described color lump shape recognition, also comprise needs identificationImage carries out the illumination compensation step of illumination compensation processing.
In an embodiment, the step of described color lump shape recognition comprises the following steps therein: will need to knowOther image is converted to gray level image; The greyscale image transitions of changing is become to bianry image; Utilize rim detectionDescribed bianry image is carried out to regional analysis, and filter small size region, extract the profile of described bianry image;Extract the summit number of the profile of described bianry image, judge the shape of color lump.
In an embodiment, the summit number of the profile of the described bianry image of described extraction, judges look thereinThe step of the shape of piece comprises the steps: the profile of described bianry image to justify detection, described in judgementWhether profile is circular; If so, finish the judgement of the shape to described profile; Otherwise, extract described wheelWide summit number, judges the shape of color lump according to the summit number of profile.
Therein in an embodiment, described described profile is justified to detection, judge that whether described profile isCircular step adopts Hough transformation algorithm to carry out.
In an embodiment, the step of described color lump location recognition comprises the steps: according to extraction thereinTo the profile of color lump determine region of search; Calculate zeroth order square and the first moment of the Hu square of described region of search;Determine the center-of-mass coordinate of color lump according to the zeroth order square of the Hu square calculating and first moment.
In an embodiment, the step of described color lump color identification is to carry out in hsv color space therein.
In an embodiment, in described color lump color identification step, the barycenter institute that chooses color lump is in place thereinThe color of putting is determined the color of color lump.
In an embodiment, before described color lump color identification step, also comprise and adopt mean filter side thereinThe step that method is carried out denoising to the image of needs identification.
A kind of color lump recognition device, comprising: color lump shape recognition module, and for the image of needs identification is enteredRow rim detection, extracts the profile of color lump in the image that needs identification, determines the shape of color lump; Color lump positionIdentification module, for calculating the center-of-mass coordinate of color lump according to the profile of the color lump extracting; The identification of color lump colorModule, for any point in the definite color lump of selected shape and according to described any some positionNeed the color of the image of identification to determine the color of color lump.
Above-mentioned color lump recognition methods and device, by the image of needs identification is carried out to rim detection, extract and needIn the image of identification, the profile of color lump is determined the shape of color lump, calculates look according to the profile of the color lump extractingThe center-of-mass coordinate of piece is obtained the position of color lump, chooses the color of a some position in color lump and determinesThe color of color lump. Therefore it can identify shape, position and the color of color lump simultaneously, thereby can be more completeThe information of the extraction color lump of face.
Brief description of the drawings
Fig. 1 is the flow chart of the color lump recognition methods of an embodiment;
Fig. 2 is the flow chart of the color lump shape recognition of color lump recognition methods shown in Fig. 1;
Fig. 3 is the profile of the bianry image that obtains by rim detection of the extraction of color lump recognition methods shown in Fig. 1Summit number, judge the flow chart of the shape of color lump;
Fig. 4 is the flow chart of the color lump location recognition of color lump recognition methods shown in Fig. 1;
Fig. 5 is the module map of the color lump recognition device of an embodiment;
Fig. 6 is the concrete module map of color lump recognition device shown in Fig. 5.
Detailed description of the invention
Please refer to Fig. 1, an embodiment provides a kind of color lump recognition methods. This color lump recognition methods comprises followingStep:
Step S110, illumination compensation, carries out illumination compensation processing to the image of needs identification.
In the image of needs identification herein, can comprise multiple color lumps, the image of needs identification be carried out hereinIllumination compensation processing is in order to eliminate the impact of illumination variation on recognition result.
In this embodiment, illumination compensation is processed the histogram equalization method that adopts. Traditional histogram equalizationChange method is the dynamic range by increasing gradation of image value, increases contrast, so that image has is larger anti-Poor. Histogram equalization method to be applied to RGB (redgreenblue, RGB) coloured image increase hereinPersistent erection, its step is as follows: first will need coloured image to be processed (herein for needing the image of identification) to representBecome three-dimensional array a: Zrgb＝[ZrZgZb]T, wherein Zr，Zg，ZbRepresent respectively red component image, green pointSpirogram picture and blue component image, each component image can be regarded as a gray level image; Then utilizeHistogram equalization method is respectively to Zr，Zg，ZbProcess, obtain the output image component S after strengtheningr，Sg，Sb; Finally by the output image component S after strengtheningr，Sg，SbAfter three component composition illumination compensationsOutput image. So just complete illumination compensation processing.
Step S120, color lump shape recognition, carries out rim detection to the image of needs identification, extracts and needs to knowThe profile of color lump in other image, determines the shape of color lump. After the shape of color lump is determined, color lump also just canFrom the image of needs identification, distinguish. Can conveniently determine like this position and the color of color lump. Please joinExamine Fig. 2, this step S120, the step of color lump shape recognition is further comprising the steps:
Step S121, is converted to gray level image by the image of needs identification.
Step S122, becomes bianry image by the greyscale image transitions of changing. By the image conversion of needs identificationOne-tenth bianry image can simplified image, the conveniently processing to image.
Step S123, utilizes rim detection to carry out regional analysis to bianry image, and filters small size region,Extract the profile of bianry image. Rim detection is herein tolerance, detection and the location to variation of image grayscale.Rim detection can adopt Canny edge detection algorithm, carrys out process decision chart by the maximum of asking signal functionPicture edge pixel point. Rim detection mainly comprises the following steps: first use Gaussian filter smoothed binary image;Then by the finite difference of single order local derviation assign to amplitude and the direction of compute gradient; Then carry out non-to gradient magnitudeMaximum suppresses; Finally detect with dual threshold algorithm and be connected edge. So just can obtain bianry imageEdge. After bianry image edge extracting, the area of the profile that calculating bianry image edge surrounds, with removal faceThe long-pending profile that is less than setting value, realizes small size region and filters. So just can be by whole needed color lumpExtract, and obtain the profile of required color lump.
Step S124, the summit number of the profile of the bianry image that extraction obtains by rim detection, judges lookThe shape of piece. Please refer to Fig. 3, this step S124 specifically comprises the steps: the profile of bianry image to enterRow circle detects, and judges whether this profile is circular; If so, finish the judgement of the shape to this profile; No, extract the summit number of this profile, judge the shape of color lump according to the summit number of profile. If profile topCounting out to meet is greater than 2 and be less than 5 condition, filters this profile and finishes shape decision, if this is taken turnsWide summit number equals 4, judges the quadrangle that is shaped as of color lump, if this profile summit number equals 3,Judge the triangle that is shaped as of color lump. In this embodiment, the shape of color lump comprises circle, triangle and fourLimit shape. But herein not as limit, by suitable step is changed and can make this color lump identification sideMethod can be identified the more color lump of Multiple Shape. In addition, profile is justified to detection herein, judge that whether profile isCircular step adopts Hough transformation algorithm to carry out.
Step S130, color lump location recognition, according to the center-of-mass coordinate of the profile calculating color lump of the color lump extracting.Please refer to Fig. 4, this step S130, color lump location recognition step mainly comprises the steps:
Step S131, determines region of search according to the profile of the color lump extracting.
Step S132, calculates zeroth order square and the first moment of the Hu square of this region of search. Set up an office (x, y) for certain searchesLocation of pixels in rope region, I (x, y) is the pixel value that point (x, y) is located, the zeroth order square of this region of searchM00With first moment M01，M10As follows respectively:
Step S133, determines the centroid position of color lump according to the zeroth order square of the Hu square calculating and first moment.Just can obtain this region of search, the i.e. matter of color lump according to zeroth order square and the first moment of the Hu square calculatingThe position of the heart is: (xc,yc)＝(M10/M00,M01/M00). In this embodiment, the position of color lump barycenter alsoIt is the position of color lump.
Step S140, color lump color identification, any point in the definite color lump of selected shape according to thisThe color of the image of the needs identification of a some position is determined the color of color lump. Carrying out step S140,Before the step of color lump color identification, can adopt mean filter method to carry out denoising to the image of needs identification.In addition, in this embodiment, while carrying out the identification of color lump color, be the color of choosing the barycenter position of color lumpDetermine the color of color lump. Step S140, the step of color lump color identification is at HSV (huesaturationValue, form and aspect saturation degree) carry out in color space. Step S120, the step of color lump shape recognition canDetermine the shape of color lump, also can filter out some unwanted pictures in the image that needs identification simultaneously. ThisSample, the shape of color lump just can be by color lump position, shape and color in the time carrying out the identification of color lump color after determiningBe mapped. And there will not be the not corresponding situation of color lump position, shape and color. For example, there is coordinateIt is the Green triangle shape and color of coordinate (2,7) position that the circular color lump of redness of (3,5) position carries out after color identificationThe color of piece, green.
Concrete, the RGB coloured image of the image that need to carry out color identification after illumination compensation is converted toHsv color space. In hsv color space, H, S, V represent respectively color, saturation degree, brightness.Color H is by given around the anglec of rotation of V axle, and scope is 0-360. Detect the color of color lump centroid position, willIt is as color lump color. Then the H value at color lump barycenter place is set in the neighborhood window centered by this pointHave a mean value of H value, to realize the object of color lump image denoising. In the present embodiment by colorModel is divided into six parts, and wherein, if the scope of the value of the color H obtaining is 0-60, color lump color isRed; If the scope of the value of H is 60-120, color lump color is yellow; If the scope of the value of H is 120-180,Color lump color is green; If the scope of the value of H is 180-240, color lump color is cyan; If H'sThe scope of value is 240-300, and color lump color is blue; If the scope of the value of H is 300-360, color lumpColor is pinkish red. In other embodiments, color model can be cut apart more detailed. So just can obtainArrive the color of the barycenter of color lump, also just obtained the color of color lump.
Just can obtain shape, position and the colouring information of color lump by above step.
Please refer to Fig. 5, one embodiment of the present of invention provide a kind of color lump recognition device. This color lump recognition deviceComprise illumination compensation module 210, color lump shape recognition module 220, color lump location identification module 230 and color lumpColor identification module 240. Illumination compensation module 210 is carried out illumination compensation processing to the image of needs identification. LookBlock-shaped identification module 220, for the image of needs identification is carried out to rim detection, extracts the figure that needs identificationThe profile of color lump in picture, determines the shape of color lump. Color lump location identification module 230 is extracted for basisThe profile of color lump calculates the center-of-mass coordinate of color lump. Color lump color identification module 240 is determined for selected shapeAny point in color lump and according to this color of the image of the needs identification of any some position determineThe color of color lump.
Please refer to Fig. 6, further, color lump shape recognition module 220 comprise greyscale image transitions module 221,Bianry image modular converter 222, profile extraction module 223 and summit extraction module 224. Greyscale image transitionsModule 221 is for being converted to gray level image by the image of needs identification. Bianry image modular converter 222 forThe greyscale image transitions of changing is become to bianry image. Convert the image of needs identification to bianry image passableSimplified image, the convenient processing to image. Profile extraction module 223 is for utilizing rim detection to binary mapPicture carries out regional analysis, and filters small size region, extracts the profile of bianry image. Summit extraction module 224Be used for the summit number of the profile that extracts the bianry image obtaining by rim detection, judge the shape of color lump.
Please refer to Fig. 6, color lump location identification module 230 comprises region of search determination module 231, computing module232 and centroid position determination module 233. Wherein, region of search determination module 231 extracts for basisThe profile of color lump is determined region of search. Computing module 232 is for calculating the zeroth order square of Hu square of this region of searchAnd first moment. Zeroth order square and the first moment of the Hu square that centroid position determination module 233 calculates for basisDetermine the centroid position of color lump.
Above-mentioned color lump recognition methods and device, by the image of needs identification is carried out to rim detection, extract and needIn the image of identification, the profile of color lump is determined the shape of color lump, calculates look according to the profile of the color lump extractingThe center-of-mass coordinate of piece is obtained the position of color lump, chooses the color of a some position in color lump and determinesThe color of color lump. Therefore it can identify shape, position and the color of color lump simultaneously, thereby can be more completeThe information of the extraction color lump of face.
The above embodiment has only expressed several embodiment of the present invention, and it describes comparatively concrete and detailed,But can not therefore be interpreted as the restriction to the scope of the claims of the present invention. It should be pointed out that for this areaThose of ordinary skill, without departing from the inventive concept of the premise, can also make some distortion andImprove, these all belong to protection scope of the present invention. Therefore, the protection domain of patent of the present invention should be with appendedClaim is as the criterion.
Claims (6)
1. a color lump recognition methods, is characterized in that, comprises the following steps:
Adopt histogram equalization method to carry out illumination compensation processing to the image of needs identification, comprising:
First coloured image to be processed need is expressed as to three-dimensional array a: Zrgb＝[ZrZgZb]T, wherein Zr，Zg，ZbRepresent respectively red component image, green component image and blue component image, each component image canTo regard a gray level image as; Then utilize histogram equalization method respectively to Zr，Zg，ZbLocateReason, obtains the output image component S after strengtheningr，Sg，Sb; Finally by the output image component S after strengtheningr，Sg，SbOutput image after three component composition illumination compensations;
Color lump shape recognition, comprising:
The image of needs identification is converted to gray level image;
The greyscale image transitions of changing is become to bianry image;
Utilize rim detection to carry out regional analysis to described bianry image, comprising:
Use Gaussian filter smoothed binary image;
By the finite difference of single order local derviation assign to amplitude and the direction of compute gradient;
Gradient magnitude is carried out to non-maximum inhibition; With
With dual threshold algorithm detect and be connected edge, obtain the edge of bianry image;
Calculate the area of the profile that surrounds of bianry image edge, be less than the profile of setting value to remove area, withFilter small size region, extract the profile of described bianry image; And
Extract the summit number of the profile of described bianry image, judge the shape of color lump;
Color lump location recognition, according to the center-of-mass coordinate of the profile calculating color lump of the color lump extracting;
The identification of color lump color, comprising:
The RGB coloured image of the image that need to carry out color identification after illumination compensation is converted to HSV faceThe colour space;
The H value at color lump barycenter place be set in the neighborhood window centered by this color lump barycenter had a HThe mean value of value; And
Color model is divided into six parts, wherein, if the scope of the value of the color H obtaining is 0-60,Color lump color is red; If the scope of the value of H is 60-120, color lump color is yellow; If the value of HScope is 120-180, and color lump color is green; If the scope of the value of H is 180-240, color lump colorFor cyan; If the scope of the value of H is 240-300, color lump color is blue; If the scope of the value of H is300-360, color lump color is pinkish red.
2. color lump recognition methods according to claim 1, is characterized in that, the described two-value of described extractionThe summit number of the profile of image, judges that the step of the shape of color lump comprises the steps:
Profile to described bianry image is justified detection, judges whether described profile is circular;
If so, finish the judgement of the shape to described profile;
Otherwise, extract the summit number of described profile, judge the shape of color lump according to the summit number of profile.
3. color lump recognition methods according to claim 2, is characterized in that, described described profile is enteredRow circle detects, and judges whether described profile is that circular step adopts Hough transformation algorithm to carry out.
4. color lump recognition methods according to claim 1, is characterized in that, described color lump location recognitionStep comprise the steps:
Determine region of search according to the profile of the color lump extracting;
Calculate zeroth order square and the first moment of the Hu square of described region of search;
Determine the center-of-mass coordinate of color lump according to the zeroth order square of the Hu square calculating and first moment.
5. color lump recognition methods according to claim 1, is characterized in that, knows in described color lump colorIn other step, the color of choosing the barycenter position of color lump is determined the color of color lump.
6. color lump recognition methods according to claim 1, is characterized in that, knows in described color lump colorBefore other step, also comprise the step that adopts mean filter method to carry out denoising to the image of needs identification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310125145.XA CN103177259B (en) | 2013-04-11 | 2013-04-11 | Color lump recognition methods |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310125145.XA CN103177259B (en) | 2013-04-11 | 2013-04-11 | Color lump recognition methods |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103177259A CN103177259A (en) | 2013-06-26 |
| CN103177259B true CN103177259B (en) | 2016-05-18 |
Family
ID=48637100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310125145.XA Active CN103177259B (en) | 2013-04-11 | 2013-04-11 | Color lump recognition methods |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103177259B (en) |
Families Citing this family (25)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105082141A (en) * | 2015-05-08 | 2015-11-25 | 上海交通大学 | Execution modules, magic cube resolving robot and use method for magic cube resolving robot |
| CN104821000B (en) * | 2015-05-22 | 2017-12-22 | 京东方科技集团股份有限公司 | Color-identifying system, color recognition method and display device |
| CN105138616A (en) * | 2015-08-10 | 2015-12-09 | 闻泰通讯股份有限公司 | Method convenient for searching for image |
| CN105243664B (en) * | 2015-10-08 | 2018-07-17 | 东南大学 | A kind of wheeled mobile robot fast-moving target tracking method of view-based access control model |
| CN105513086B (en) * | 2016-01-26 | 2019-02-19 | 上海葡萄纬度科技有限公司 | A kind of Educational toy external member and its localization method is detected based on the matched magic square of shape |
| CN106096599B (en) * | 2016-04-28 | 2019-03-26 | 浙江工业大学 | A kind of inside truck positioning method based on painting color lump |
| CN107491714B (en) * | 2016-06-13 | 2022-04-05 | 中科晶锐(苏州)科技有限公司 | Intelligent robot and target object identification method and device thereof |
| CN106295656B (en) * | 2016-08-03 | 2017-09-15 | 徐庆 | Image outline characteristic extraction method and device based on image color lump content |
| CN106325750A (en) * | 2016-08-26 | 2017-01-11 | 曹蕊 | Character recognition method and system applied in terminal equipment |
| CN107336251A (en) * | 2016-09-20 | 2017-11-10 | 苏州小璐机器人有限公司 | A kind of control method and system of robot queue |
| CN106485231A (en) * | 2016-10-21 | 2017-03-08 | 上海未来伙伴机器人有限公司 | A kind of target color lump extracting method and device |
| CN106855949A (en) * | 2016-11-29 | 2017-06-16 | 广州视源电子科技股份有限公司 | The recognition methods of mark point and device |
| CN106780531B (en) * | 2016-12-22 | 2019-10-15 | 河海大学 | A kind of image edge extraction method based on color lump point feature |
| CN106964563B (en) * | 2017-04-28 | 2019-02-22 | 福州大学 | A kind of plastic bottle method for sorting |
| CN106984556B (en) * | 2017-04-28 | 2019-05-10 | 福州大学 | A kind of plastic bottle intelligent sorting system and its working method |
| CN108152278B (en) * | 2017-11-22 | 2020-07-24 | 沈阳普泽众康医药科技有限公司 | Urine detection method and device |
| CN108510477B (en) * | 2018-03-15 | 2021-08-24 | 深圳市飞点健康管理有限公司 | Method and device for positioning color block of test paper |
| CN108596232B (en) * | 2018-04-16 | 2022-03-08 | 杭州睿珀智能科技有限公司 | Automatic insole classification method based on shape and color characteristics |
| CN108890692A (en) * | 2018-07-05 | 2018-11-27 | 陕西大中科技发展有限公司 | A kind of material color identification method for industrial robot vision's sorting |
| CN109345735A (en) * | 2018-10-12 | 2019-02-15 | 南京理工大学 | A kind of self-service machine commodity recognition method and system |
| CN109587337A (en) * | 2018-12-14 | 2019-04-05 | 江苏科技大学 | A kind of method and smart phone of quick detection mycotoxin |
| CN110103223A (en) * | 2019-05-27 | 2019-08-09 | 西安交通大学 | A kind of identification of view-based access control model follows barrier-avoiding method and robot automatically |
| CN110414333A (en) * | 2019-06-20 | 2019-11-05 | 平安科技（深圳）有限公司 | A kind of detection method and device of image boundary |
| CN111126414A (en) * | 2019-12-03 | 2020-05-08 | 武汉邦拓信息科技有限公司 | Method for identifying map blocks in decoration map based on deep learning |
| CN113470125B (en) * | 2021-07-21 | 2024-01-12 | 杭州万事利丝绸文化股份有限公司 | Method and system for carrying out color matching on patterns |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1721144A (en) * | 2004-07-13 | 2006-01-18 | 中国科学院自动化研究所 | A kind of fast tracking method and device based on color of object surface |
| CN102005040A (en) * | 2010-09-09 | 2011-04-06 | 吉林大学 | Target joining algorithm based on multi-feature fusion match |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8045826B2 (en) * | 2008-07-03 | 2011-10-25 | Seiko Epson Corporation | Detecting edges in a digital images |
-
2013
- 2013-04-11 CN CN201310125145.XA patent/CN103177259B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1721144A (en) * | 2004-07-13 | 2006-01-18 | 中国科学院自动化研究所 | A kind of fast tracking method and device based on color of object surface |
| CN102005040A (en) * | 2010-09-09 | 2011-04-06 | 吉林大学 | Target joining algorithm based on multi-feature fusion match |
Non-Patent Citations (1)
| Title |
|---|
| 一种基于纹理和颜色的目标跟踪方法;李巍等;《计算机仿真》;20110131;第28卷(第1期);273-276 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103177259A (en) | 2013-06-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103177259B (en) | Color lump recognition methods | |
| CN109377485B (en) | Machine vision detection method for instant noodle packaging defects | |
| CN102043950B (en) | Vehicle outline recognition method based on canny operator and marginal point statistic | |
| CN109558806B (en) | Method for detecting high-resolution remote sensing image change | |
| CN105354599B (en) | A kind of color identification method based on improved SLIC super-pixel segmentation algorithm | |
| CN102750535B (en) | Method and system for automatically extracting image foreground | |
| CN105046700B (en) | Fruit surface defect detection method and system based on gamma correction and color classification | |
| CN107945200B (en) | Image binarization segmentation method | |
| CN104866849A (en) | Food nutrition label identification method based on mobile terminal | |
| CN108229458A (en) | A kind of intelligent flame recognition methods based on motion detection and multi-feature extraction | |
| CN104574405B (en) | A kind of coloured image threshold segmentation method based on Lab space | |
| CN103971128A (en) | Traffic sign recognition method for driverless car | |
| CN103605977A (en) | Extracting method of lane line and device thereof | |
| CN101551853A (en) | Human ear detection method under complex static color background | |
| CN104766344B (en) | Vehicle checking method based on movement edge extractor | |
| CN103198319B (en) | For the blurred picture Angular Point Extracting Method under the wellbore environment of mine | |
| CN102147867B (en) | Method for identifying traditional Chinese painting images and calligraphy images based on subject | |
| CN103198315A (en) | License plate character segmentation algorithm based on character outline and template matching | |
| CN108734131B (en) | Method for detecting symmetry of traffic sign in image | |
| CN102737221B (en) | Method and apparatus for vehicle color identification | |
| CN102184404B (en) | Method and device for acquiring palm region in palm image | |
| CN107527343A (en) | A kind of agaricus bisporus stage division based on image procossing | |
| CN105701487A (en) | Vehicle logo positioning method | |
| CN102881160A (en) | Outdoor traffic sign identification method under low-illumination scene | |
| CN105893960A (en) | Road traffic sign detecting method based on phase symmetry |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |