CN101251373A - Method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture - Google Patents
Method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture Download PDFInfo
- Publication number
- CN101251373A CN101251373A CNA2008101028812A CN200810102881A CN101251373A CN 101251373 A CN101251373 A CN 101251373A CN A2008101028812 A CNA2008101028812 A CN A2008101028812A CN 200810102881 A CN200810102881 A CN 200810102881A CN 101251373 A CN101251373 A CN 101251373A
- Authority
- CN
- China
- Prior art keywords
- image
- pixel
- similarity
- dimensional dimension
- point pixel
- 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
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention relates to a quick detection method for a microstructure three-dimensional stereo image, in particular to a method of establishing a matching relation through a stereo image pair and calculating the length, width and height of the microstructure through an object image inverse engineering model. The method comprises the following main steps of: preprocessing the images, picking up feature points in the images, matching the features, carrying out the optimization and error correction for matching results, carrying out the object image inverse engineering and outputting sizes of the three dimensions. The method which belongs to the non-contact type measurement without damaging the surface of a sample can ensure the measurement precision of the system. Simultaneously the method improves the detection speed through generating a stereo surface, thereby bringing about a simpler and quicker operating flow. The method is a computer-aided measuring method with high efficiency.
Description
Technical field
The present invention relates to a kind of micro-structure three-dimensional dimension stereoscopic picture detection method, particularly relate to and adopt stereo-picture to setting up matching relationship, by the counter method of asking Model Calculation microstructure length, width and height of image.
Background technology
The micro-structure three-dimensional dimension detection is the important directions in the MEMS (micro electro mechanical system).The range scale of microstructure is about several microns-tens microns, and volume is very little, and its three-dimensional dimension estimates it is necessary in microstructure processing and the production.Microstructure is a bigger scope, can be micro element, as raceway groove, boss etc., can also be the combined system that constitutes by in a plurality of micro elements.Microstructure generally will just can be finished through multiple working procedure, be subjected to the influence of external factor such as processing conditions, groping the processing technology stage, often need to estimate length, width and the height of microstructure, according to testing result, constantly the adjusting process parameter is till reaching the rational technique index then.
At present, occurred multiple size detecting method, scanning electron microscope detection method and atomic force microscopy are two kinds of microstructure size detection methods using always, and this two class methods accuracy of detection is higher, but belong to contact measurement method, micro-structure surface is had destruction.And the operating process of these two class methods is very complicated, finishes a measuring process and need spend the long time, is not easy to grope the quick measurement in processing technology stage.
Interferometric method, triangulation, fringe projection method also have been used for the measurement of micro-structure three-dimensional dimension, these methods belong to contactless measurement, micro-structure surface is not had destruction, but its detection resolution and precision are on the low side, are suitable for the bigger micro-structure three-dimensional dimension of yardstick and measure.These methods are analyzed the fringe spacing in the image, frequency information, and accuracy of detection is subjected to the influence of outside auxiliary optics obvious, and operating process is comparatively complicated.
Summary of the invention
At the existing problem of above-mentioned existing micro-structure three-dimensional dimension detection method, the present invention releases the contactless method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture that is suitable for small scale, its purpose is to adopt stereo-picture to setting up matching relationship, by anti-Model Calculation microstructure length, width and the height asked of image, provide length, width and the height number of microstructure region of interest automatically.
To achieve these goals, the present invention has taked following technical scheme.Right by the microscopic stereovision system acquisition stereo-picture that optical stereo microscope and two CCD cameras are formed, and send computing machine to, by the computing machine stereoscopic image to handling and mating, by the anti-3 D stereo surface that obtains microstructure of asking of image, thus the length of definite region of interest, width and height.Described micro-structure three-dimensional dimension stereoscopic picture detection method may further comprise the steps:
1, pretreatment image
Import left image and right image to computing machine, carry out filtering, target dividing processing by computing machine then, distinguish object pixel and background pixel in the image, object pixel is set to black, background pixel is set to white, obtains left and right sides pretreatment image and left and right sides target image.
2, the unique point pixel in the extraction image
At left target image and right target image, adopt general feature extracting method to extract unique point pixel in left image and the right image, write down the image coordinate of unique point pixel in the image coordinate of unique point pixel in the left image and the right image.
The feature extracting method that adopts comprises local maximum entropy method (MEM), Wavelet Transform, Sober Operator Method and the gray threshold method asked.
3, characteristic matching
With the unique point pixel in the left image is sub pixel, and corresponding region of search is set in right image, and the region of search method to set up is: if the image coordinate of sub pixel in left image is (w
1 l, w
2 l), in right image with coordinate (w
1 l, w
2 l) corresponding pixel is the center, rectangular region [w
1 l± d
1, w
2 l± d
2] be region of search, wherein 1≤d
1≤ 10,0≤d
2≤ 4.
All reference point pixels in the traversal search zone are calculated the similarity between sub pixel and the reference point pixel, obtain the similarity set of region of search correspondence, and set is added up to similarity, obtains the regional similarity of region of search correspondence.Statistical method comprises the method for asking maximum similarity, and average method is asked in set to similarity.
The similarity threshold of regional similarity and setting relatively, judge whether there is the match point pixel in this region of search, if regional similarity greater than similarity threshold, then the match point pixel of current seed pixel exists, otherwise the match point pixel does not exist.Travel through all sub pixels in the left image,, in right image, find out the match point pixel of all sub pixel correspondences according to identical search strategy.Similarity threshold is got the value in the interval [0.6,0.9].
4, matching result optimization and error correction
Adopt outer polar curve constraint condition and consistency constraint principle that matching result is optimized and error correction, further get rid of wrong match point pixel and mistake match point pixel in the match point collection of pixels, the image coordinate of output match point pixel.
5, image is counter asks
Image coordinate and parallax with sub pixel in the image of the left and right sides and corresponding match point pixel are independent variable, adopt the anti-model of asking of image in the computing machine binocular stereo vision principle, calculate the volume coordinate set of the object space reference point of all sub pixels and match point pixel correspondence.If the image coordinate of certain sub pixel is (w
1 l, w
2 l), corresponding match point pixel coordinate is (w
1 r, w
2 r), parallax is defined as w
1 l-w
1 rThe anti-model of asking of image that adopts comprises pinhole imaging system model, projection model and amblyopia difference microscopic stereovision model.
The anti-coordinate set of the object space reference point of model acquisition of asking of image is handled, adopted the triangulation method in the graphics that three dimensions is put moulding, obtain space curved surface.
6, output three-dimensional dimension
Select the tested zone of three-dimensionalreconstruction face, export the distance of this regional length, width and short transverse, this distance is tested three-dimensional dimension.
Micro-structure three-dimensional dimension method for quick involved in the present invention has adopted handles stereo-picture to setting up the method that matching relationship calculates three-dimensional size, this metering system belongs to non-contact measurement, do not destroy sample surfaces, and can guarantee the measuring accuracy of system, simultaneously, improve detection speed by generating three-dimensional surface, operating process is simpler and more direct, is the very high computer aided measurement method of a kind of efficient.
Description of drawings
Fig. 1 is the method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture synoptic diagram that the present invention relates to
Fig. 2 is the feature matching method synoptic diagram that the present invention relates to
Description of symbols in the accompanying drawing
S11, import left image
S12, import right image
S21, pretreatment image
Target in S31, the identification left and right sides image
The left target image of S32, output
The right target image of S33, output
S41, extract left characteristics of image
S42, extract right characteristics of image
S51, characteristic matching
S52, matching result optimization and error correction
S53, the output of match point image coordinate
S54, left image characteristic point
The similarity of S55, calculating seed and reference point
All unique points in S56, the traversing graph picture
S57, coupling finish
S58, selection reference point
S59, right images match zone is set
S591, storage similarity result of calculation
S592, judge whether to have traveled through all pixels in the matching area
Whether maximum similarity is greater than the threshold value that is provided with in S593, the judgement set
S594, give up this sub pixel
S595, reference point is set is match point
S61, image is counter asks
The 3 dimensional drawing of S71, generation
The selection of S81, measured zone
S91, output three-dimensional dimension: length, width and height
Embodiment
Now in conjunction with the accompanying drawings the present invention is further elaborated.Fig. 1 and Fig. 2 show the process flow diagram of the method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture that the present invention relates to, and as shown in the figure, method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture may further comprise the steps:
1, pretreatment image
Import left image S11 and right image S12 to computing machine, adopt coloured image vector median filter method to carry out filtering, obtain filtered coloured image IAL and IAR.Then coloured image is adopted and cut apart target and image background in the following method.
(1), image I AL (IAR) is converted into gray level image IBL (IBR) at all pixels among the view picture coloured image IAL (IAR).Method for transformation is: the red, green, blue gray-scale value of the pixel among the gray level image IBL (IBR) equates, and equals the sum average value of the red, green, blue gray-scale value of respective pixel among the image I AL (IAR).
(2) at all pixels among the gray level image IBL (IBR), image I BL (IBR) is converted into split image ICL (ICR), dividing method is: judge that whether grey scale pixel value among the gray level image IBL (IBR) is greater than preset threshold TM, if set up greater than relation, then be the background pixel among the image I CL (ICR), be set to white, if be false greater than relation, then be the object pixel among the image I CL (ICR), be set to black, TM gets 100.
(3) image I CL (ICR) is carried out area filtering, the pseudo-object pixel of filtering obtains filtered image I DL (IDR).Adopt the method for point by point search, seek each isolated black picture element group that distributes among the image I CL (ICR), and write down the black picture element number of each black picture element group, if the black picture element number of this black picture element group is greater than preset threshold, this threshold value gets 200, then be the target label pixel, otherwise be pseudo-target label pixel, be set to white.
(4) color of image recovers.The color value of the black objects marked pixels among the image I DL (IDR) is set to the pixel color value of the corresponding position among the image I AL (IAR), and image I AL (IAR) is pretreated image S32 (S33).
2, the unique point pixel in the extraction image
At the left target image S32 of output and the right target image S33 of output, the feature extracting method of local maximum entropy is asked in employing, if the gray-scale value of the current pixel among target image S32 and the S33 is f (i, j), (i j) is the image coordinate of current pixel, selects the m * n neighborhood of this pixel, m, n are set to 8, calculate the entropy of the pixel in this zone:
H
I, jFor (i, j) to the entropy of deserved 8 neighborhoods, k, p are the image coordinate of the pixel in 8 fields, if H
I, j>0.05, be the unique point pixel then, keep original color, otherwise be background pixel, be set to white, thus, the right image characteristic point pixel S42 of left image characteristic point pixel S41 that obtains extracting and extraction.
3, characteristic matching
At the left characteristics of image pixel S41 that extracts and the right characteristics of image pixel S42 of extraction, with the current unique point among the S41 is sub pixel, the corresponding position is provided with right images match region S 59 in S42, and method to set up is: if the image coordinate of sub pixel in left image is (w
1 l, w
2 l), in right image with coordinate (w
1 l, w
2 l) corresponding pixel is the center, rectangular region [w
1 l± 10, w
2 l± 10] as the region of search.
Select the reference point pixel S58 among the S59, all reference point pixels in the traversal search zone, calculate the similarity between sub pixel and all the reference point pixels, storage similarity result of calculation S591, and the similarity that constitutes right images match region S 59 is gathered the general formula below similarity calculating is used:
ρ is the similarity value of left image sub pixel and right image reference point pixel, g
LkBe the red, green, blue gray-scale value of certain sub pixel correspondence, g
RkBe the red, green, blue gray-scale value of certain reference point pixel correspondence, k=1,2,3,
Similarity set is added up, is adopted the method for asking maximum similarity, if maximum similarity greater than preset threshold 0.6, the right image reference point pixel of this similarity correspondence is the match point pixel, otherwise, no match point pixel in the right images match region S 59.Traveled through all sub pixels in the left image,, in right image, found out the match point pixel of all sub pixels according to identical search strategy.
4, matching result optimization and error correction
Adopt outer polar curve constraint condition and consistency constraint principle that matching result is optimized and error correction, further get rid of wrong match point pixel and mistake match point pixel in the coupling.Based on outer polar curve constraint condition, it is the variable zone of vertical image coordinate at center that the match point pixel of sub pixel in right image should be positioned at the sub pixel, vertically the domain of walker of image coordinate is ± 5 pixels, if the difference of vertical image coordinate of match point pixel and the vertical image coordinate of sub pixel exceeds this scope, be wrong match point pixel, otherwise be correct match point pixel.
Based on the consistency constraint principle, exist the match point pixel of the sub pixel of priority position also to have identical relation in the left image, if the image coordinate of match point pixel does not satisfy this relation, then be mistake match point pixel.
5, image is counter asks
Image coordinate and parallax with sub pixel in the image of the left and right sides and corresponding match point pixel are independent variable, adopt amblyopia difference microscopic stereovision model in the computing machine binocular stereo vision principle, calculate the volume coordinate set of the object space reference point of all sub pixels and match point pixel correspondence.If the image coordinate of certain sub pixel is (w
1 l, w
2 l), corresponding match point pixel coordinate is (w
1 r, w
2 r), parallax is defined as w
1 l-w
1 r
The anti-coordinate set of the object space reference point of model acquisition of asking of image is handled, adopted the triangulation method in the graphics that three dimensions is put moulding, obtain space curved surface.
6, output three-dimensional dimension
Select the tested zone of three-dimensionalreconstruction face, export the distance of this regional length, width and short transverse, this distance is tested three-dimensional dimension.
To one skilled in the art, clearly, the present invention can make multiple improvement and variation, if fall into appending claims and the scope that is equal in, these improvement of the present invention and variation are just contained in the present invention.
Claims (8)
1. method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture is characterized in that: adopt microscopic stereovision system acquisition stereo-picture right, and send computing machine to, the computing machine stereoscopic image is to handling and mating, and concrete treatment step is as follows:
1) pretreatment image;
Left image and right image to image pair are handled respectively, obtain left and right sides pretreatment image and left and right sides target image;
2) the unique point pixel in the extraction image
Extract the unique point pixel in left target image and the right target image, write down the image coordinate of unique point pixel in the image coordinate of unique point pixel in the left image and the right image;
3) characteristic matching
With the unique point pixel in the left image is sub pixel, corresponding region of search is set in right image seeks its match point;
4) matching result optimization and error correction;
5) image is counter asks;
Image coordinate and parallax with sub pixel in the image of the left and right sides and corresponding match point pixel are independent variable, adopt the anti-model of asking of image, calculate the volume coordinate set of the object space reference point of all sub pixels and match point pixel correspondence, adopt the triangulation method in the graphics that three dimensions is put moulding, obtain space curved surface;
6) output three-dimensional dimension
Select the tested zone of three-dimensionalreconstruction face, export the distance of this regional length, width and short transverse, this distance is tested three-dimensional dimension.
2. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 1, it is characterized in that: the pre-service described in the step 1 comprises carries out filtering, target dividing processing to left image and right image, distinguish object pixel and background pixel in the image, object pixel is set to black, background pixel is set to white, obtains left and right sides pretreatment image and left and right sides target image.
3. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 1 is characterized in that: the method for the extract minutiae pixel described in the step 2 is for asking local maximum entropy method (MEM) or Wavelet Transform or Sober Operator Method or gray threshold method.
4. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 1, it is characterized in that: the feature matching method described in the step 3 is: with the unique point pixel in the left image is sub pixel, corresponding region of search is set in right image, all reference point pixels in the traversal search zone, calculate the similarity between sub pixel and the reference point pixel, obtain the similarity set of region of search correspondence, set is added up to similarity, obtains the regional similarity of region of search correspondence; The similarity threshold of regional similarity and setting relatively, if regional similarity greater than similarity threshold, then the match point pixel of current seed pixel exists, otherwise the match point pixel does not exist; Travel through all sub pixels in the left image,, in right image, find out the match point pixel of all sub pixel correspondences according to identical searching method.
5. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 4 is characterized in that: the described method that set is added up to similarity is the method for maximum similarity.
6. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 4 is characterized in that: the interval of described similarity threshold is [0.6,0.9].
7. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 1, it is characterized in that: adopt outer polar curve constraint condition and consistency constraint principle that matching result is optimized and error correction in the described step 4, get rid of wrong match point pixel and mistake match point pixel in the match point collection of pixels.
8. a kind of method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture according to claim 1 is characterized in that: the anti-model of asking of the image described in the step 5 is pinhole imaging system model or projection model or amblyopia difference microscopic stereovision model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2008101028812A CN101251373A (en) | 2008-03-28 | 2008-03-28 | Method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CNA2008101028812A CN101251373A (en) | 2008-03-28 | 2008-03-28 | Method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101251373A true CN101251373A (en) | 2008-08-27 |
Family
ID=39954883
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CNA2008101028812A Pending CN101251373A (en) | 2008-03-28 | 2008-03-28 | Method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101251373A (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102187367A (en) * | 2008-10-20 | 2011-09-14 | 本田技研工业株式会社 | Vehicle periphery monitoring device |
CN102333234A (en) * | 2011-10-28 | 2012-01-25 | 清华大学 | Binocular stereo video state information monitoring method and device |
CN102621143A (en) * | 2012-01-21 | 2012-08-01 | 天津大学 | Dual-view angle multiple bubble corresponding point matching method based on multi-dimensional threshold value constraint |
CN104915957A (en) * | 2015-05-29 | 2015-09-16 | 何再兴 | Matching rectification method for improving three dimensional visual sense identification precision of industrial robot |
CN105092607A (en) * | 2015-08-27 | 2015-11-25 | 浙江大学 | Method for evaluating surface defects of spherical optical components |
CN106152946A (en) * | 2015-03-31 | 2016-11-23 | 酷派软件技术(深圳)有限公司 | A kind of method of Measuring Object length and terminal |
CN106907988A (en) * | 2017-02-27 | 2017-06-30 | 北京工业大学 | The micro- visual modeling method of basic data matrix |
CN108020158A (en) * | 2016-11-04 | 2018-05-11 | 浙江大华技术股份有限公司 | A kind of three-dimensional position measuring method and device based on ball machine |
CN109539996A (en) * | 2018-12-12 | 2019-03-29 | 安徽省农业科学院水稻研究所 | A kind of rice stigma measurement of length method |
CN109919247A (en) * | 2019-03-18 | 2019-06-21 | 北京石油化工学院 | Characteristic point matching method, system and equipment in harmful influence stacking binocular ranging |
CN110825080A (en) * | 2019-10-31 | 2020-02-21 | 华南农业大学 | Orchard path visual navigation method, system and medium based on fuzzy control algorithm |
CN110926330A (en) * | 2018-09-20 | 2020-03-27 | 发那科株式会社 | Image processing apparatus, image processing method, and program |
US10861177B2 (en) | 2015-11-11 | 2020-12-08 | Zhejiang Dahua Technology Co., Ltd. | Methods and systems for binocular stereo vision |
-
2008
- 2008-03-28 CN CNA2008101028812A patent/CN101251373A/en active Pending
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102187367B (en) * | 2008-10-20 | 2017-03-08 | 本田技研工业株式会社 | Vehicle periphery monitoring device |
CN102187367A (en) * | 2008-10-20 | 2011-09-14 | 本田技研工业株式会社 | Vehicle periphery monitoring device |
CN102333234A (en) * | 2011-10-28 | 2012-01-25 | 清华大学 | Binocular stereo video state information monitoring method and device |
CN102621143A (en) * | 2012-01-21 | 2012-08-01 | 天津大学 | Dual-view angle multiple bubble corresponding point matching method based on multi-dimensional threshold value constraint |
CN102621143B (en) * | 2012-01-21 | 2013-10-09 | 天津大学 | Dual-view angle multiple bubble corresponding point matching method based on multi-dimensional threshold value constraint |
CN106152946B (en) * | 2015-03-31 | 2019-08-02 | 酷派软件技术(深圳)有限公司 | A kind of method and terminal measuring object length |
CN106152946A (en) * | 2015-03-31 | 2016-11-23 | 酷派软件技术(深圳)有限公司 | A kind of method of Measuring Object length and terminal |
CN104915957A (en) * | 2015-05-29 | 2015-09-16 | 何再兴 | Matching rectification method for improving three dimensional visual sense identification precision of industrial robot |
CN104915957B (en) * | 2015-05-29 | 2017-10-27 | 何再兴 | A kind of matching antidote for improving industrial robot 3D vision accuracy of identification |
CN105092607A (en) * | 2015-08-27 | 2015-11-25 | 浙江大学 | Method for evaluating surface defects of spherical optical components |
US10861177B2 (en) | 2015-11-11 | 2020-12-08 | Zhejiang Dahua Technology Co., Ltd. | Methods and systems for binocular stereo vision |
CN108020158A (en) * | 2016-11-04 | 2018-05-11 | 浙江大华技术股份有限公司 | A kind of three-dimensional position measuring method and device based on ball machine |
CN106907988B (en) * | 2017-02-27 | 2019-03-22 | 北京工业大学 | The micro- visual modeling method of basic data matrix |
CN106907988A (en) * | 2017-02-27 | 2017-06-30 | 北京工业大学 | The micro- visual modeling method of basic data matrix |
CN110926330A (en) * | 2018-09-20 | 2020-03-27 | 发那科株式会社 | Image processing apparatus, image processing method, and program |
CN110926330B (en) * | 2018-09-20 | 2022-12-02 | 发那科株式会社 | Image processing apparatus, image processing method, and program |
US11625842B2 (en) | 2018-09-20 | 2023-04-11 | Fanuc Corporation | Image processing apparatus and image processing method |
CN109539996A (en) * | 2018-12-12 | 2019-03-29 | 安徽省农业科学院水稻研究所 | A kind of rice stigma measurement of length method |
CN109919247A (en) * | 2019-03-18 | 2019-06-21 | 北京石油化工学院 | Characteristic point matching method, system and equipment in harmful influence stacking binocular ranging |
CN110825080A (en) * | 2019-10-31 | 2020-02-21 | 华南农业大学 | Orchard path visual navigation method, system and medium based on fuzzy control algorithm |
CN110825080B (en) * | 2019-10-31 | 2020-11-13 | 华南农业大学 | Orchard path visual navigation method, system and medium based on fuzzy control algorithm |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101251373A (en) | Method for rapidly detecting micro-structure three-dimensional dimension stereoscopic picture | |
CN108596975A (en) | A kind of Stereo Matching Algorithm for weak texture region | |
CN107248159A (en) | A kind of metal works defect inspection method based on binocular vision | |
CN109615654B (en) | Method for measuring corrosion depth and area of inner surface of drainage pipeline based on binocular vision | |
CN101398886A (en) | Rapid three-dimensional face identification method based on bi-eye passiveness stereo vision | |
CN102607455B (en) | Three-dimensional measurement method for microstructure based on optical microscope and variable illumination | |
CN103727930A (en) | Edge-matching-based relative pose calibration method of laser range finder and camera | |
CN104833679A (en) | Three-dimensional reverse calibration and detection method for microscopic defects | |
CN101246595A (en) | Multi-view point data splitting method of optical three-dimensional scanning system | |
CN104121902A (en) | Implementation method of indoor robot visual odometer based on Xtion camera | |
CN101739683B (en) | Image segmentation and multithread fusion-based method and system for evaluating depth of single image | |
CN102798349A (en) | Three-dimensional surface extraction method based on equal-gray line search | |
CN102621143B (en) | Dual-view angle multiple bubble corresponding point matching method based on multi-dimensional threshold value constraint | |
CN108399630B (en) | Method for quickly measuring distance of target in region of interest in complex scene | |
KR100438212B1 (en) | Extraction method for 3-dimensional spacial data with electron microscope and apparatus thereof | |
CN111415378B (en) | Image registration method for automobile glass detection and automobile glass detection method | |
CN106446832B (en) | Video-based pedestrian real-time detection method | |
CN105783785A (en) | Wavelet-ridge phase extraction method | |
CN105716530A (en) | Method for measuring geometric dimension of vehicle based on binocular stereoscopic vision | |
CN115601423A (en) | Edge enhancement-based round hole pose measurement method in binocular vision scene | |
CN106056599B (en) | A kind of object recognition algorithm and device based on Object Depth data | |
Teoh et al. | Ground plane detection for autonomous vehicle in rainforest terrain | |
Xiao et al. | A segment-based stereo matching method with ground control points | |
CN112966594A (en) | Environment sensing method | |
CN105374040A (en) | Large mechanical workpiece stereo matching method based on vision measurement |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20080827 |