CN113689397A - Workpiece circular hole feature detection method and workpiece circular hole feature detection device - Google Patents
Workpiece circular hole feature detection method and workpiece circular hole feature detection device Download PDFInfo
- Publication number
- CN113689397A CN113689397A CN202110966633.8A CN202110966633A CN113689397A CN 113689397 A CN113689397 A CN 113689397A CN 202110966633 A CN202110966633 A CN 202110966633A CN 113689397 A CN113689397 A CN 113689397A
- Authority
- CN
- China
- Prior art keywords
- image data
- template
- point information
- workpiece
- actual
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 84
- 238000000034 method Methods 0.000 claims abstract description 42
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 31
- 238000012545 processing Methods 0.000 claims abstract description 30
- 238000005457 optimization Methods 0.000 claims abstract description 11
- 238000000605 extraction Methods 0.000 claims description 27
- 238000003860 storage Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000011960 computer-aided design Methods 0.000 description 41
- 230000008569 process Effects 0.000 description 12
- 238000004364 calculation method Methods 0.000 description 10
- 230000004044 response Effects 0.000 description 9
- 238000009792 diffusion process Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 238000005286 illumination Methods 0.000 description 7
- 230000035945 sensitivity Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000012549 training Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000009466 transformation Effects 0.000 description 5
- 238000011897 real-time detection Methods 0.000 description 4
- 238000004519 manufacturing process Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 3
- 238000013139 quantization Methods 0.000 description 3
- 230000004075 alteration Effects 0.000 description 2
- 230000003321 amplification Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003199 nucleic acid amplification method Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000001133 acceleration Effects 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a workpiece round hole feature detection method and a workpiece round hole feature detection device, wherein the workpiece round hole feature detection method comprises the following steps: acquiring camera calibration parameters and CAD model information of a standard workpiece; generating template image data according to the camera calibration parameters and the CAD model information; extracting the characteristics of the template image data to obtain template characteristic point information; acquiring actual image data of a workpiece to be detected; extracting the features of the actual image data to obtain actual feature point information; matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information; performing iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain processed target template feature point information; and determining the characteristic value of the target circular hole according to the processed characteristic point information of the target template. By adopting the method, the detection of the circular hole of the workpiece can be realized, and the detection precision can be improved.
Description
Technical Field
The invention relates to the technical field of workpiece detection, in particular to a workpiece round hole feature detection method and a workpiece round hole feature detection device.
Background
In the field of industrial vision, many challenges are still faced with high precision target detection. The template matching algorithm plays an important role in target detection because the template matching does not need a large amount of training set and training time and can process targets with few textures.
In the related art, one is a template matching algorithm based on feature points, but the algorithm often cannot achieve a real-time detection effect for increasing robustness. One is a template matching algorithm based on object contour features, which uses the chamfer distance to measure the difference between the template picture contour and the object contour to be measured, but which is extremely sensitive to the occlusion of the object. In addition, the commonly used template matching algorithm is also very sensitive to illumination variation, noise, etc., and the time length of the algorithm is increased when a plurality of templates exist. The other method is a real-time detection method based on gradient response for a non-texture object, the method adopts binary representation of gradient direction characteristics, and simultaneously utilizes cache parallelism of a modern computer, the matching algorithm based on the shape can detect objects with multiple classes and few textures in real time, but the method cannot accurately estimate the scale of the object, and meanwhile, the matching precision of the algorithm is still to be improved. In addition, for the template matching algorithm, the specific operation is as follows: firstly, photographing each gesture of an identified object, carrying out template training on the photographed image, and storing the trained template in a database; in the identification stage, the tested pictures are compared with templates in the database one by one, and the template matching position with the maximum similarity is selected as the real coordinate of the object on the image.
Disclosure of Invention
The present invention is directed to solving at least one of the problems of the prior art. Therefore, an object of the present invention is to provide a workpiece circular hole feature detection method, by which a workpiece circular hole can be detected and detection accuracy can be improved.
The embodiment of the second aspect of the invention provides a workpiece circular hole feature detection device.
A third aspect of the present invention provides a computer storage medium.
The embodiment of the fourth aspect of the invention provides a workpiece circular hole feature detection system.
In order to solve the above problem, a workpiece round hole feature detection method according to an embodiment of the first aspect of the present invention includes: acquiring camera calibration parameters and CAD (Computer Aided Design) model information of a standard workpiece; generating template image data according to the camera calibration parameters and the CAD model information, wherein the template image data is the image data of the round hole in the standard workpiece; extracting the characteristics of the template image data to obtain template characteristic point information; acquiring actual image data of a workpiece to be detected, wherein the actual image data is image data of a target round hole in the workpiece to be detected; performing feature extraction on the actual image data to obtain actual feature point information; matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information; performing iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain processed target template feature point information; and determining the characteristic value of the target round hole according to the processed characteristic point information of the target template.
According to the workpiece round hole feature detection method, the template image data can be generated through camera calibration parameters and CAD model information of a standard workpiece, compared with a traditional template matching algorithm, the template image data is generated by referring to the CAD model information, namely three-dimensional data information, so that detection of the standard workpiece in any shape can be expanded flexibly, detection accuracy can be improved, and sensitivity to illumination can be reduced.
In some embodiments, the generating template image data from the camera calibration parameters and CAD model information includes: performing multi-angle projection on the CAD model information by using the camera calibration parameters to obtain initial template image data; scaling the initial template image data according to a preset radius range to obtain initial template image data under different scales; and taking the initial image data under the different scales as final template image data.
In some embodiments, the template feature point information includes feature point coordinates, feature point gradient values, and feature point gradient directions.
In some embodiments, the scaling the initial template image data according to a preset radius range to obtain the initial template image data at different scales includes: acquiring the standard radius of the target circular hole according to the CAD model information; determining the preset radius range according to the standard radius and a preset zooming interval; and in the preset radius range, carrying out scale scaling on the initial template image data according to a preset step length to obtain the initial template image data under different scales.
In some embodiments, feature extraction is performed on the actual image data, including: performing image preprocessing on the actual image data to obtain preprocessed actual image data; amplifying the processed actual image data by a spline difference method to obtain amplified actual image data; and performing feature extraction on the amplified actual image data.
In some embodiments, matching the template feature point information with the actual feature point information to obtain target template feature point information includes: and matching the template characteristic point information with the actual characteristic point information by a template search matching method to obtain target template characteristic point information.
In some embodiments, the characteristic values of the target circular hole include a radius of the target circular hole, center coordinates, and a normal to a plane in which the target circular hole lies.
The embodiment of the second aspect of the invention provides a workpiece circular hole feature detection device, which comprises: the first acquisition module is used for acquiring camera calibration parameters and CAD model information of a standard workpiece; the template generation module is used for generating template image data according to the camera calibration parameters and the CAD model information, wherein the template image data is the image data of the round hole in the standard workpiece; the first feature extraction module is used for extracting features of the template image data to obtain template feature point information; the second acquisition module is used for acquiring actual image data of the workpiece to be detected, wherein the actual image data is image data of a target circular hole in the workpiece to be detected; the second feature extraction module is used for extracting features of the actual image data to obtain actual feature point information; the matching module is used for matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information; the processing module is used for carrying out iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain processed target template feature point information; and the determining module is used for determining the characteristic value of the target circular hole according to the processed characteristic point information of the target template.
According to the workpiece round hole feature detection device provided by the embodiment of the invention, the template image data can be generated through the camera calibration parameters and the CAD model information of the standard workpiece, compared with a template matching algorithm, the embodiment of the invention generates the template image data by referring to the CAD model information, namely three-dimensional data information, so that the detection of the standard workpiece in any shape can be flexibly expanded, the detection accuracy can be improved, and the sensitivity to illumination can be reduced.
A third aspect of the present invention provides a computer storage medium, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the workpiece circular hole feature detection method described in the above embodiments.
An embodiment of a fourth aspect of the present invention provides a workpiece circular hole feature detection system, including: the image acquisition module is used for acquiring actual image data of the workpiece to be detected; in the workpiece circular hole feature detection device in the above embodiment, the workpiece circular hole feature detection device is connected to the image acquisition module.
According to the workpiece round hole feature detection system provided by the embodiment of the invention, the workpiece round hole feature detection device provided by the embodiment is adopted, so that the workpiece round hole can be detected, the detection precision can be improved, and the sensitivity to illumination can be reduced.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a flow chart of a method for detecting a workpiece circular hole feature according to one embodiment of the invention;
FIG. 2 is a schematic diagram of gradient direction partitioning according to one embodiment of the present invention;
fig. 3(a) - (c) are schematic diagrams illustrating the diffusion of the gradient direction of the feature point to the neighborhood according to an embodiment of the present invention;
FIGS. 4 (d) - (i) are schematic diagrams of initial template image data according to one embodiment of the present invention;
FIG. 5 is a block diagram of an apparatus for detecting a circular hole feature of a workpiece according to an embodiment of the present invention;
FIG. 6 is a block diagram of a workpiece hole feature detection system according to an embodiment of the invention.
Reference numerals:
a workpiece circular hole feature detection device 10; a workpiece circular hole feature detection system 20;
a first acquisition module 1; a template generation module 2; a first feature extraction module 3; a second obtaining module 4; a second feature extraction module 5; a matching module 6; a processing module 7; a determination module 8; an image acquisition module 9.
Detailed Description
Embodiments of the present invention will be described in detail below, the embodiments described with reference to the drawings being illustrative, and the embodiments of the present invention will be described in detail below.
In order to solve the above problem, an embodiment of the first aspect of the present invention provides a method for detecting a workpiece circular hole feature, which can detect a workpiece circular hole, and can improve detection accuracy and reduce sensitivity to light.
Referring to fig. 1, a workpiece round hole feature detection method according to an embodiment of the present invention is described, and as shown in fig. 1, the workpiece round hole feature detection method at least includes steps S1 to S8.
And step S1, acquiring camera calibration parameters and CAD model information of the standard workpiece.
The camera calibration parameters include internal parameters of the camera, such as focal length and pixel size, and external parameters of the camera, such as position and rotation direction of the camera. The CAD model information of the standard workpiece comprises coordinate information of each point in the standard workpiece, orientation information of the standard workpiece, standard radius of a round hole in the standard workpiece, standard center coordinates of the center of the round hole, a standard normal direction of a plane where the round hole is located and the like.
In the embodiment, it can be understood that, when a workpiece to be detected is identified, template image data and actual image data of the workpiece to be detected should be acquired by the same camera so as to meet the matching degree of the workpiece circular hole feature detection, and therefore, after the camera calibration is actually completed, corresponding camera calibration parameters can be input into related computer processing equipment, and the computer processing equipment generates subsequent template image data according to the camera calibration parameters. And when the workpiece to be detected is actually produced, the CAD model information of the corresponding standard workpiece is preset, namely the workpiece to be detected is produced by the CAD model information of the standard workpiece, so that the CAD model information of the standard workpiece can be input into relevant computer processing equipment. Therefore, the computer processing equipment can acquire the needed camera calibration parameters and the CAD model information of the standard workpiece according to the information input by the user.
And step S2, generating template image data according to the camera calibration parameters and the CAD model information, wherein the template image data is the image data of the round hole in the standard workpiece.
In the embodiment, the existing template matching algorithm can detect the characteristics of the circular holes of the workpiece only through the steps of image acquisition, template training, template matching and the like, unlike the existing template matching algorithm, the embodiment of the invention does not need to carry out the processes of image acquisition and template training when acquiring the template image data, but automatically generates template image data directly with camera calibration parameters and CAD model information of the standard workpiece, that is, taking CAD model information of a standard workpiece as an actual workpiece, representing the camera after actual calibration by camera calibration parameters, thereby simulating the operation of shooting the standard workpiece by the camera and processing the image data after the simulated shooting operation, thus, the image data after the analog shooting operation and the image data after the analog shooting operation are processed are the generated template image data. Meanwhile, compared with the template in the existing template matching algorithm which is obtained by training collected two-dimensional image data, the template image data is generated by referring to CAD (computer-aided design) model information, namely three-dimensional data information, of the standard workpiece, so that the template image data of the standard workpiece in any shape can be flexibly expanded in the template image data generation process, and therefore the generated template image data are richer, the flexibility of subsequent detection is improved, the accuracy of subsequent detection and identification is improved, and the sensitivity to illumination is reduced.
Step S3, feature extraction is performed on the template image data to obtain template feature point information.
The template feature point information may be understood as information of a point in the template image data that can best reflect the feature of the circular hole in the standard workpiece, and if the minimum edge point in the template image data is taken as a feature point, the information of the minimum edge point is the template feature point information.
In the embodiment, considering that the storage space required for storing the template image data is large and the template image data cannot directly participate in the calculation of the similarity metric, the embodiment of the present invention performs feature extraction on the template image data to store the extracted template feature point information, for example, the extracted template feature point information may be stored in yml format, thereby reducing the occupation of the storage space.
In the embodiment, it can be understood that, in step S2, the number of template image data generated according to the camera calibration parameters and the CAD model information is large, and therefore, when performing feature extraction on the template image data, feature extraction needs to be performed on each template image data respectively to obtain template feature point information corresponding to each template image data.
The feature extraction is performed on each template image data, a pyramid structure can be adopted for each template image data, that is, each template image data is scaled in an equal proportion to obtain the template image data after being scaled in an equal proportion corresponding to each template image data, each original template image data and the corresponding template image data after being scaled in an equal proportion form a pyramid structure, the feature extraction is further performed on each original template image data and the corresponding template image data after being scaled in an equal proportion, and the template feature point information extracted from each original template image data and the template feature point information extracted from the corresponding template image data after being scaled in an equal proportion are used as the template image template feature point information finally extracted from the original template image data.
In some embodiments, the template feature point information includes feature point coordinates, feature point gradient values, and feature point gradient directions.
Step S4, acquiring actual image data of the workpiece to be detected, where the actual image data is image data of the target circular hole in the workpiece to be detected.
In an embodiment, when the workpiece to be detected is identified, the calibrated camera may be used to shoot the workpiece to be detected, and the shot actual image data may be transmitted to the relevant computer processing device.
In step S5, feature extraction is performed on the actual image data to obtain actual feature point information, so as to match the actual feature point information with the template feature point information obtained earlier.
In an embodiment, the actual feature point information may include feature point coordinates, feature point gradient values, and feature point gradient directions.
And step S6, matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information.
The target template feature point information may be understood as the template feature point information with the highest similarity to the actual feature point information among all the template feature point information.
In an embodiment, the feature point information may be an image gradient direction descriptor, and the image gradient direction descriptors corresponding to all the template feature point information are matched with the image gradient direction descriptors corresponding to the actual feature point information one by one to obtain the target template feature point information.
Specifically, when template feature point information is matched with actual feature point information, edge gradient information of an image is used as feature matching, and the matching step comprises similarity measurement, gradient direction quantization, direction diffusion, pre-response graph processing and linear storage.
And the similarity measurement is to calculate similarity aiming at cosine values between the direction of the template characteristic points and the direction of the actual image edge points in the matching process. In the calculation process, a calculation mode of efficiently measuring the similarity is adopted, for each gradient direction on an object to be measured, namely actual characteristic point information, the most similar direction in the template characteristic point information corresponding to the input template image data is searched in the neighborhood of the relevant gradient position, and the calculation formula is as follows.
Wherein ori (I, T) is a direction angle of a feature point T in the actual image data I, ori (T, R) is a direction angle of a feature point R in the template image data T, P is a set of template feature point information in the template image data, c is a current matching position of the template image data T on the actual image data I, and R (c + R) is a domain area of the actual image data at c + R where the size is the template image data.
Furthermore, for the gradient direction vectorization, the edge of the image can be calculated by using a Sobel operator, the gradient direction and the gradient copy of the edge can be calculated at the same time, and the non-maximum suppression processing can be adopted for the gradient amplitude of each edge point. Quantifying the gradient direction to n by the above algorithm0Equal regions and using a 2-ary coding.
For example, referring to fig. 2, 0 ° -180 ° is divided into 5 equally divided regions, and when the quantization direction points to a certain section, the corresponding position is set to 1, and the other positions are set to 0, such as when the quantized gradient is located in the third section, i.e., 72 ° -108 °, the binary representation thereof is 00100. For another example, 0 ° to 180 ° may be divided into 16 equal divisions, two bytes (a binary string with a length of 16) are used to represent the direction of the gradient, the value of the modified similarity function is 0 to 8, and meanwhile, since two bytes are used to store the cumulative sum, the maximum number of template feature points may be set to [65535/8 ]: 8191, and thus, for a circular hole with a large aperture, by appropriately setting a larger number of feature points, the matching may achieve a more accurate and more robust circular hole feature detection effect.
Further, for directional diffusion, in order to make the quantization of the gradient more robust to noise, the gradient direction of each feature point is assigned to other positions in the periphery in the calculation, for example, the gradient direction of each feature point in fig. 3(a) is assigned to other positions in the 3 × 3 neighborhood of the periphery shown in fig. 3 (b). And, in order to make the subsequent calculation more efficient, a binary string may be used to represent the gradient image after diffusion, and for each position, if the gradient direction of itself and the gradient direction of the neighborhood diffusion occur, the corresponding position of the binary string is 1, as shown in fig. 3(c), a binary image formed in a binary string manner in fig. 3(b), that is, a diffusion image.
Further, for the preprocessed response graph, when the number of the gradient directions is n0Then, pre-response values of each direction on the actual image data are calculated in advance to form n0A pre-response map. When matching is carried out, only the corresponding pre-response graph needs to be searched according to the gradient direction of the characteristic points, so that the similarity value on the actual image data can be obtained, and the matching efficiency is further improved in an off-line calculation mode. Specifically, n is first calculated0Combining the similarity values of each direction and all gradients to form a lookup table, and further combining each edge point in the diffusion image with n0The similarity values are calculated inversely for each gradient to form n0A pre-response map.
For example, assume n0At 8, the similarity measure has only five values of 0, 0.3, 0.7, 0.9 and 1, wherein the similarity function is calculated as follows.
ε(I，T，t)＝maxt∈T|cos(ori(T，r)-ori(I，t))|
Finally, for linear storage, in order to enable computer processing equipment to read the Cache Hit Rate (Cache Hit Rate) of the memory information, the numerical value of the pre-response graph is stored again in a more friendly manner of Cache (Cache memory) reading during calculation, and the numerical value of the pre-response graph is read by taking T as a step pitch, so that the linear storage is completed, and the matching efficiency is improved.
In addition, a search strategy of an image pyramid may be adopted in matching, for example, taking a two-layer image pyramid as an example, the resolution of the top-layer image pyramid is 1/2 of the resolution of the original image, and a search strategy with a step size of 4 is used at the top layer; and the bottom layer image pyramid is used as an original image, and a search strategy with the step length of 2 is used at the bottom layer. In a specific search, a primary position of the best match is searched at the top layer, and then a more accurate matching position is searched in the field of the primary position by using a smaller step size in the bottom layer image. Therefore, the matching can achieve a more accurate and more robust circular hole feature detection effect.
In the embodiment, when the template feature point information is matched with the actual feature point information, a multi-processor program design method for OpenMP parallel acceleration can be introduced, so that a multi-thread processing mode is adopted, the operation efficiency of the algorithm can be improved, and the method is particularly suitable for the condition of simultaneously detecting a plurality of round holes.
And step S7, carrying out iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain the processed target template feature point information.
In an embodiment, the iterative optimization alignment algorithm is a point set-to-point set registration method, and specifically, there are two point sets: and in the calculation, the target point set is unchanged, the source point set is rotated and translated, even the source point set is subjected to scale transformation, so that the transformed source point set is superposed with the target point set as much as possible, and the transformation process is point set registration. The output of the correspondence of the two sets of points is typically a rigid transformation matrix: representing rotation and translation, which is applied to the source data set, the result is a complete match to the target data set. Therefore, in the embodiment of the invention, the target template feature point information is used as a source point set, the actual feature point information is used as a target point set, and a point-to-surface iterative optimization alignment algorithm is adopted, namely, the square sum of the distances from the points on the target template feature point information to the corresponding point tangent planes on the actual feature point information is minimized, so that the position fine adjustment is carried out on the target template feature point information through the iterative optimization alignment algorithm, the feature value of the matched target circular hole can reach the sub-pixel precision, even the detection error is less than 0.01mm, meanwhile, the image data rotation parameters are not needed to be considered in the process, the detection speed and the detection stability are improved, and the effect of real-time detection is realized.
Step S8, determining a feature value of the target circular hole according to the processed target template feature point information, that is, the processed target template feature point information corresponds to the actual feature point information, so that the feature value of the circular hole in the processed target template feature point information is the feature value of the target circular hole in the actual feature point information.
In some embodiments, the characteristic values of the target circular aperture include a radius of the target circular aperture, center coordinates, and a normal to a plane in which the target circular aperture lies. Therefore, the radius in the processed target template characteristic point information is the radius of the target round hole in the workpiece to be detected, and the center coordinate of the round hole in the processed target template characteristic point information is the center coordinate of the target round hole in the workpiece to be detected.
According to the workpiece round hole feature detection method, the template image data can be generated through camera calibration parameters and CAD model information of a standard workpiece, compared with a traditional template matching algorithm, the template image data is generated by referring to the CAD model information, namely three-dimensional data information, so that detection of the standard workpiece in any shape can be expanded flexibly, detection accuracy can be improved, and sensitivity to illumination can be reduced.
In some embodiments, the camera calibration parameters are used for performing multi-angle projection on the CAD model information to obtain initial template image data, that is, template pictures of the CAD model information at different viewing angles of the camera are obtained as the initial template image data, as shown in (d) - (i) of fig. 4, the initial template image data at different viewing angles are respectively represented, for example, for each hole in the CAD model, the CAD center coordinate of the round hole is used as the center, the CAD model radius of the round hole is used as the radius, the CAD normal of the plane where the round hole is located is used as the rotating shaft, one point is collected at intervals of 0.01 degrees, 360 degrees are annularly collected, 36000 dense points are collected in total, 36000 dense points are projected to the image plane where the round hole camera is observed by using the camera calibration parameters to form the boundary information of the template, and finally, the initial template image data can be obtained by using the flood filling algorithm; furthermore, considering that a certain difference exists between the actual radius of the circular hole on the workpiece and the radius of the corresponding circular hole in the CAD model during actual production, the embodiment of the invention scales the initial template image data according to the preset radius range to obtain the initial template image data under different scales; finally, the initial image data under different scales are used as final template image data, so that the final template image data are generated under the condition that different visual angles and actual production errors of the camera are considered, and the accuracy of subsequent detection of the workpiece to be detected is improved.
In some embodiments, a standard radius of the target circular hole is obtained according to the CAD model information, wherein the radius meeting the manufacturing requirement of the circular hole on the workpiece when the workpiece is produced is the standard radius; furthermore, a preset radius range is determined according to the standard radius and a preset scaling interval, wherein the preset scaling interval can be understood as an error scaling interval allowed for the actually produced workpiece, and can be set according to the actual situation without limitation; and in the preset radius range, carrying out scale scaling on the initial template image data according to a preset step length so as to obtain the initial template image data under different scales. Therefore, on the basis of the initial template image data under the multi-scale, when the workpiece is identified subsequently, the rigid workpiece with any scale can be identified, and the flexibility of detection is improved.
For example, assuming that the preset zoom interval is 0.95 to 1.05 and the standard radius is 5cm, it may be determined that the preset radius range is 5cm × 0.95 to 5cm × 1.05, that is, 4.75cm to 5.25cm, and further, within the preset radius range of 4.75cm to 5.25cm, the initial template image data is scaled by a preset step size of 0.002, that is, the initial template image data corresponding to radii of 4.75cm, 4.752cm, 4.754cm and 4.756cm … … 5.25.25 cm are respectively obtained, so that the initial template image data under different scales may be obtained.
In some embodiments, the image preprocessing is performed on the actual image data, for example, the actual image data is subjected to linear contrast enhancement first to increase image edge contrast, and then the gaussian filtering and laplacian operator are used to perform smooth denoising and sharpening on the actual image data to reduce image noise information and store the edge information of the image, so that the preprocessed actual image data is obtained through the image preprocessing process, and the detection effect can be improved; furthermore, image interception processing is carried out on the actual image data after preprocessing so as to obtain the actual image data after interception, namely, the image data part of the non-target circular hole in the actual image data is removed so as to improve the detection precision; further, the processed actual image data is amplified through a spline difference method to obtain the amplified actual image data, so that subsequent detection is facilitated, and particularly, the detection on a small target circular hole is facilitated; and performing feature extraction on the amplified actual image data. Therefore, the actual image data is subjected to the processing process and then subjected to feature extraction, and the detection accuracy is improved.
In an embodiment, when performing image preprocessing, the canny edge algorithm may be used to extract edge information.
In some embodiments, the template feature point information is matched with the actual feature point information by a template search matching method to obtain the target template feature point information. The template searching and matching method adopts the gradient information of a color image as the basis of template matching, so that the center coordinates of the circular holes can reach the pixel-level precision, and in addition, the method can also realize sliding window searching on a picture by thousands of templates, thereby achieving the effect of real-time detection.
In an embodiment, a sliding window scoring mechanism may be improved in the matching process, so that the matching achieves a more accurate and more robust circular hole feature detection effect, for example, when the gradient directions of corresponding feature points in the template image data and the actual image data are consistent, the score of the feature point is 4; and when the difference between the corresponding characteristic points in the template image data and the actual image data is 1, the score of the characteristic point is 1, and by using the scoring mechanism, the score sum of all the characteristic points is finally calculated, and the interval of 0-100 is normalized.
For explanation, the calculation of the center coordinates of the target circular hole is performed, for example, assuming that the minimum x coordinate of the template feature point after extraction in the x direction in the template image data is tlxThe minimum y coordinate in the y direction is tlyWhen the template image data is subjected to image interception processing, the adopted interception matrix is as follows:
the invalid area in the template image data can be removed by carrying out image interception processing on the template image data, so that the sliding window space is larger during matching.
Furthermore, after the matching of the actual feature point information is completed, under the condition that the intercepted actual image data is not amplified, the coordinate of the pixel point of which the upper left corner position in the template image data is located in the actual image data is assumed to be [ match]TAnd the rigid transformation matrix obtained after the iterative optimization alignment algorithm is transformation, the central coordinate P of the target circular holecenter：
Wherein, (cx, cy) is the center coordinates of the circular hole in the feature point information of the target template, that is, the pixel coordinates of the projection of the center of the circular hole in the CAD model in the template image data.
In addition, when the intercepted actual image data, that is, the image data of interest is subjected to the amplification processing, assuming that the amplification factor of the intercepted actual image data is n, the center coordinate of the target circular hole in the actual image data is finally obtained as [ P ]center.x/n+ROI.x，Pcenter.y/n+ROI.y]TAnd (ROI.x, ROI.y) is the coordinates of pixel points of which the upper left corner positions in the interested image data are positioned in the actual image data.
It should be noted that, when the workpiece to be detected is actually photographed, the photographed image data includes a plurality of round holes on the workpiece to be detected, and in order to avoid interference on detection of each round hole, the actual image data is subjected to image capture processing to obtain the interested image data for each round hole, i.e., the captured actual image data, so that the interested image data is matched with the template image data, and the detection accuracy is improved.
In summary, according to the workpiece circular hole feature detection method provided by the embodiment of the invention, the template image data is obtained through the camera calibration parameters and the CAD model information of the standard workpiece, namely, the three-dimensional vision technology and the camera calibration technology are combined, and the rotation parameters are not required to be considered during matching.
According to a second aspect of the present invention, the apparatus 10 for detecting a workpiece circular hole feature includes, as shown in fig. 3, a first obtaining module 1, a template generating module 2, a first feature extracting module 3, a second obtaining module 4, a second feature extracting module 5, a matching module 6, a processing module 7, and a determining module 8.
The first acquisition module 1 is used for acquiring camera calibration parameters and CAD model information of a standard workpiece; the template generation module 2 is used for generating template image data according to the camera calibration parameters and the CAD model information, wherein the template image data is the image data of the round hole in the standard workpiece; the first feature extraction module 3 is used for extracting features of the template image data to obtain template feature point information; the second obtaining module 4 is used for obtaining actual image data of the workpiece to be detected, wherein the actual image data is image data of a target circular hole in the workpiece to be detected; the second feature extraction module 5 is configured to perform feature extraction on actual image data to obtain actual feature point information; the matching module 6 is used for matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information; the processing module 7 is used for performing iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain processed target template feature point information; the determining module 8 is configured to determine a feature value of the target circular hole according to the processed target template feature point information.
It should be noted that a specific implementation manner of the workpiece circular hole feature detection apparatus 10 according to the embodiment of the present invention is similar to a specific implementation manner of the workpiece circular hole feature detection method according to any of the above-mentioned embodiments of the present invention, and please refer to the description of the method portion specifically, and details are not repeated here in order to reduce redundancy.
According to the workpiece round hole feature detection device 10, the template image data can be generated through camera calibration parameters and CAD model information of a standard workpiece, compared with a traditional template matching algorithm, the embodiment of the invention generates the template image data by referring to the CAD model information, namely three-dimensional data information, so that the detection of the standard workpiece in any shape can be flexibly expanded, the detection accuracy can be improved, and the sensitivity to illumination can be reduced.
In some embodiments, the template generating module 2 is further configured to perform multi-angle projection on the CAD model information with the camera calibration parameters to obtain initial template image data; scaling the initial template image data according to a preset radius range to obtain initial template image data under different scales; and taking the initial template image data and the initial image data under different scales as final template image data.
In some embodiments, the template generating module 2 is further configured to obtain a standard radius of the target circular hole according to the CAD model information; determining a preset radius range according to the standard radius and a preset zooming interval; and in the preset radius range, carrying out scale scaling on the initial template image data according to a preset step length so as to obtain the initial template image data under different scales.
In some embodiments, the second feature extraction module 5 is further configured to perform image preprocessing on the actual image data to obtain preprocessed actual image data; amplifying the processed actual image data by a spline difference method to obtain amplified actual image data; and performing feature extraction on the amplified actual image data.
In some embodiments, the matching module 6 is further configured to match the template feature point information with the actual feature point information by a template search matching method to obtain target template feature point information.
In a third aspect, the present invention provides a computer storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the workpiece circular hole feature detection method provided in the foregoing embodiments.
In a fourth aspect of the present invention, a workpiece circular hole feature detection system is provided, and as shown in fig. 4, the workpiece circular hole feature detection system 20 includes an image acquisition module 9 and the workpiece circular hole feature detection apparatus 10 provided in the foregoing embodiment.
The image acquisition module 9 is used for acquiring actual image data of a workpiece to be detected, and the workpiece round hole feature detection device 10 is connected with the image acquisition module 9.
It should be noted that a specific implementation manner of the workpiece circular hole feature detection system 20 based on deep learning according to the embodiment of the present invention is similar to a specific implementation manner of the workpiece circular hole feature detection apparatus 10 according to any of the above embodiments of the present invention, and please refer to the description of the workpiece circular hole feature detection apparatus 10 specifically, and details are not repeated here in order to reduce redundancy.
In some embodiments, the image acquisition module 9 comprises a plurality of CCD (charge coupled device) cameras.
According to the workpiece circular hole feature detection system 20 provided by the embodiment of the invention, the workpiece circular hole feature detection device 10 provided by the embodiment can be used for detecting a workpiece circular hole, and the detection precision can be improved.
In the description of this specification, any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of custom logic functions or processes, and alternate implementations are included within the scope of the preferred embodiment of the present invention in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the embodiments of the present invention.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present invention may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention, and that variations, modifications, substitutions and alterations can be made to the above embodiments by those of ordinary skill in the art within the scope of the present invention.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an illustrative embodiment," "an example," "a specific example," or "some examples" or the like mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
Claims (10)
1. A workpiece circular hole feature detection method is characterized by comprising the following steps:
acquiring camera calibration parameters and CAD model information of a standard workpiece;
generating template image data according to the camera calibration parameters and the CAD model information, wherein the template image data is the image data of the round hole in the standard workpiece;
extracting the characteristics of the template image data to obtain template characteristic point information;
acquiring actual image data of a workpiece to be detected, wherein the actual image data is image data of a target round hole in the workpiece to be detected;
performing feature extraction on the actual image data to obtain actual feature point information;
matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information;
performing iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain processed target template feature point information;
and determining the characteristic value of the target round hole according to the processed characteristic point information of the target template.
2. The method for detecting the circular hole features of the workpiece according to claim 1, wherein the step of generating template image data according to the camera calibration parameters and the CAD model information comprises the steps of:
performing multi-angle projection on the CAD model information by using the camera calibration parameters to obtain initial template image data;
scaling the initial template image data according to a preset radius range to obtain initial template image data under different scales;
and taking the initial image data under the different scales as final template image data.
3. The method of claim 2, wherein the template feature point information comprises feature point coordinates, feature point gradient values, and feature point gradient directions.
4. The method for detecting the circular hole features of the workpiece according to claim 2, wherein the step of scaling the initial template image data according to a preset radius range to obtain the initial template image data under different scales comprises the following steps:
acquiring the standard radius of the target circular hole according to the CAD model information;
determining the preset radius range according to the standard radius and a preset zooming interval;
and in the preset radius range, carrying out scale scaling on the initial template image data according to a preset step length to obtain the initial template image data under different scales.
5. The workpiece circular hole feature detection method according to claim 1, wherein performing feature extraction on the actual image data comprises:
performing image preprocessing on the actual image data to obtain preprocessed actual image data;
image interception processing is carried out on the preprocessed actual image data to obtain intercepted actual image data;
amplifying the intercepted actual image data by a spline difference method to obtain amplified actual image data;
and performing feature extraction on the amplified actual image data.
6. The workpiece circular hole feature detection method according to claim 1, wherein matching the template feature point information with the actual feature point information to obtain target template feature point information comprises:
and matching the template characteristic point information with the actual characteristic point information by a template search matching method to obtain target template characteristic point information.
7. The method of claim 1, wherein the characteristic values of the target circular hole include a radius of the target circular hole, a center coordinate, and a normal to a plane in which the target circular hole is located.
8. A workpiece circular hole feature detection device is characterized by comprising:
the first acquisition module is used for acquiring camera calibration parameters and CAD model information of a standard workpiece;
the template generation module is used for generating template image data according to the camera calibration parameters and the CAD model information, wherein the template image data is the image data of the round hole in the standard workpiece;
the first feature extraction module is used for extracting features of the template image data to obtain template feature point information;
the second acquisition module is used for acquiring actual image data of the workpiece to be detected, wherein the actual image data is image data of a target circular hole in the workpiece to be detected;
the second feature extraction module is used for extracting features of the actual image data to obtain actual feature point information;
the matching module is used for matching the template characteristic point information with the actual characteristic point information to obtain target template characteristic point information;
the processing module is used for carrying out iterative optimization alignment algorithm processing on the target template feature point information by taking the actual feature point information as a reference to obtain processed target template feature point information;
and the determining module is used for determining the characteristic value of the target circular hole according to the processed characteristic point information of the target template.
9. A computer storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the workpiece hole feature detection method of any one of claims 1-7.
10. A workpiece circular hole feature detection system, comprising:
the image acquisition module is used for acquiring actual image data of the workpiece to be detected;
the workpiece circular hole feature detection device of claim 8, connected to the image acquisition module.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110966633.8A CN113689397A (en) | 2021-08-23 | 2021-08-23 | Workpiece circular hole feature detection method and workpiece circular hole feature detection device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110966633.8A CN113689397A (en) | 2021-08-23 | 2021-08-23 | Workpiece circular hole feature detection method and workpiece circular hole feature detection device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113689397A true CN113689397A (en) | 2021-11-23 |
Family
ID=78581351
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110966633.8A Pending CN113689397A (en) | 2021-08-23 | 2021-08-23 | Workpiece circular hole feature detection method and workpiece circular hole feature detection device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113689397A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114792373A (en) * | 2022-04-24 | 2022-07-26 | 广东天太机器人有限公司 | Visual identification spraying method and system of industrial robot |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101173856A (en) * | 2007-08-30 | 2008-05-07 | 上海交通大学 | Vehicle collision accident reappearance method based on phototopography and exterior profile deformation of car body |
| CN101546350A (en) * | 2009-05-07 | 2009-09-30 | 同济大学 | Matching method of bodywork closure member under line segment aggregate distance measurement |
| CN101698303A (en) * | 2009-09-11 | 2010-04-28 | 大连理工大学 | Automatic calibration method between three-dimensional laser and monocular vision |
| CN101763632A (en) * | 2008-12-26 | 2010-06-30 | 华为技术有限公司 | Method for demarcating camera and device thereof |
| US20130033596A1 (en) * | 2011-08-03 | 2013-02-07 | The Boeing Company | Projection aided feature measurement using uncalibrated camera |
| JP2017045081A (en) * | 2015-08-24 | 2017-03-02 | 株式会社アマダホールディングス | Work position detection device and method, as well as work holding robot control device and control method |
| CN106600600A (en) * | 2016-12-26 | 2017-04-26 | 华南理工大学 | Wafer defect detection method based on characteristic matching |
| CN107942949A (en) * | 2017-03-31 | 2018-04-20 | 沈机（上海）智能系统研发设计有限公司 | A kind of lathe vision positioning method and system, lathe |
| CN107967482A (en) * | 2017-10-24 | 2018-04-27 | 广东中科南海岸车联网技术有限公司 | Icon-based programming method and device |
| CN108256394A (en) * | 2016-12-28 | 2018-07-06 | 中林信达（北京）科技信息有限责任公司 | A kind of method for tracking target based on profile gradients |
| CN110136048A (en) * | 2018-02-09 | 2019-08-16 | 中国科学院上海高等研究院 | A kind of method for registering images and system, storage medium and terminal |
| CN111474184A (en) * | 2020-04-17 | 2020-07-31 | 河海大学常州校区 | AOI character defect detection method and device based on industrial machine vision |
| CN111986178A (en) * | 2020-08-21 | 2020-11-24 | 北京百度网讯科技有限公司 | Product defect detection method and device, electronic equipment and storage medium |
| CN111982921A (en) * | 2020-05-21 | 2020-11-24 | 北京安视中电科技有限公司 | Hole defect detection method and device, conveying platform and storage medium |
| CN112966725A (en) * | 2021-02-08 | 2021-06-15 | 深圳市优必选科技股份有限公司 | Method and device for matching template images and terminal equipment |
| CN113111212A (en) * | 2021-04-01 | 2021-07-13 | 广东拓斯达科技股份有限公司 | Image matching method, device, equipment and storage medium |
-
2021
- 2021-08-23 CN CN202110966633.8A patent/CN113689397A/en active Pending
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101173856A (en) * | 2007-08-30 | 2008-05-07 | 上海交通大学 | Vehicle collision accident reappearance method based on phototopography and exterior profile deformation of car body |
| CN101763632A (en) * | 2008-12-26 | 2010-06-30 | 华为技术有限公司 | Method for demarcating camera and device thereof |
| CN101546350A (en) * | 2009-05-07 | 2009-09-30 | 同济大学 | Matching method of bodywork closure member under line segment aggregate distance measurement |
| CN101698303A (en) * | 2009-09-11 | 2010-04-28 | 大连理工大学 | Automatic calibration method between three-dimensional laser and monocular vision |
| US20130033596A1 (en) * | 2011-08-03 | 2013-02-07 | The Boeing Company | Projection aided feature measurement using uncalibrated camera |
| JP2017045081A (en) * | 2015-08-24 | 2017-03-02 | 株式会社アマダホールディングス | Work position detection device and method, as well as work holding robot control device and control method |
| CN106600600A (en) * | 2016-12-26 | 2017-04-26 | 华南理工大学 | Wafer defect detection method based on characteristic matching |
| CN108256394A (en) * | 2016-12-28 | 2018-07-06 | 中林信达（北京）科技信息有限责任公司 | A kind of method for tracking target based on profile gradients |
| CN107942949A (en) * | 2017-03-31 | 2018-04-20 | 沈机（上海）智能系统研发设计有限公司 | A kind of lathe vision positioning method and system, lathe |
| CN107967482A (en) * | 2017-10-24 | 2018-04-27 | 广东中科南海岸车联网技术有限公司 | Icon-based programming method and device |
| CN110136048A (en) * | 2018-02-09 | 2019-08-16 | 中国科学院上海高等研究院 | A kind of method for registering images and system, storage medium and terminal |
| CN111474184A (en) * | 2020-04-17 | 2020-07-31 | 河海大学常州校区 | AOI character defect detection method and device based on industrial machine vision |
| CN111982921A (en) * | 2020-05-21 | 2020-11-24 | 北京安视中电科技有限公司 | Hole defect detection method and device, conveying platform and storage medium |
| CN111986178A (en) * | 2020-08-21 | 2020-11-24 | 北京百度网讯科技有限公司 | Product defect detection method and device, electronic equipment and storage medium |
| CN112966725A (en) * | 2021-02-08 | 2021-06-15 | 深圳市优必选科技股份有限公司 | Method and device for matching template images and terminal equipment |
| CN113111212A (en) * | 2021-04-01 | 2021-07-13 | 广东拓斯达科技股份有限公司 | Image matching method, device, equipment and storage medium |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114792373A (en) * | 2022-04-24 | 2022-07-26 | 广东天太机器人有限公司 | Visual identification spraying method and system of industrial robot |
| CN114792373B (en) * | 2022-04-24 | 2022-11-25 | 广东天太机器人有限公司 | Visual identification spraying method and system of industrial robot |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111127422B (en) | Image labeling method, device, system and host | |
| CN109949899B (en) | Image three-dimensional measurement method, electronic device, storage medium, and program product | |
| Wang et al. | Digital image correlation in experimental mechanics and image registration in computer vision: Similarities, differences and complements | |
| Wöhler | 3D computer vision: efficient methods and applications | |
| CN107993258B (en) | Image registration method and device | |
| CN109801333B (en) | Volume measurement method, device and system and computing equipment | |
| CN109716348B (en) | Processing multiple regions of interest independently | |
| EP3382645B1 (en) | Method for generation of a 3d model based on structure from motion and photometric stereo of 2d sparse images | |
| CN111028205B (en) | Eye pupil positioning method and device based on binocular distance measurement | |
| CN110223355B (en) | Feature mark point matching method based on dual epipolar constraint | |
| CN113393439A (en) | Forging defect detection method based on deep learning | |
| CN113689397A (en) | Workpiece circular hole feature detection method and workpiece circular hole feature detection device | |
| Li et al. | A camera on-line recalibration framework using SIFT | |
| CN110084736A (en) | A kind of method of detecting watermarks and system based on SURF and pyramid algorith | |
| CN113012279B (en) | Non-contact three-dimensional imaging measurement method and system and computer readable storage medium | |
| Frangione et al. | Multi-step approach for automated scaling of photogrammetric micro-measurements | |
| CN114511631A (en) | Method and device for measuring height of visual object of camera and computer readable storage medium | |
| Li et al. | Structured light based high precision 3D measurement and workpiece pose estimation | |
| Si et al. | A fast and robust template matching method with rotated gradient features and image pyramid | |
| Wöhler et al. | Triangulation-based approaches to three-dimensional scene reconstruction | |
| Correa-Tome et al. | Fast similarity metric for real-time template-matching applications | |
| Vuori et al. | Three-dimensional imaging system with structured lighting and practical constraints | |
| CN116993803B (en) | Landslide deformation monitoring method and device and electronic equipment | |
| Zhang et al. | Normalized weighted shape context and its application in feature-based matching | |
| JP7327484B2 (en) | Template creation device, object recognition processing device, template creation method, object recognition processing method, and program |
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 |