CN108960325A - A kind of automobile metal plate work detection Identification of Cracks system based on SVM and Hog - Google Patents
A kind of automobile metal plate work detection Identification of Cracks system based on SVM and Hog Download PDFInfo
- Publication number
- CN108960325A CN108960325A CN201810715062.9A CN201810715062A CN108960325A CN 108960325 A CN108960325 A CN 108960325A CN 201810715062 A CN201810715062 A CN 201810715062A CN 108960325 A CN108960325 A CN 108960325A
- Authority
- CN
- China
- Prior art keywords
- hog
- svm
- image
- hyperplane
- amplitude
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30164—Workpiece; Machine component
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of, and the automobile metal plate work based on SVM and Hog detects Identification of Cracks system, and specific testing process is as follows: extracting picture feature using hog first；Then SVM predicted pictures label is used；Confirm whether detection piece is qualified.Compared to other algorithms, hog extracts feature and has the advantage that being primarily due to HOG is operated on the local pane location of image, so it can keep good invariance to image geometry and optical deformation, both deformation are only appeared on bigger space field；Secondly under the conditions ofs thick airspace sampling, fine direction sampling and the normalization of stronger indicative of local optical etc., as long as entity posture, pedestrian can be allowed there are some subtle limb actions, these subtle movements can be ignored without influencing detection effect.The present invention extracts characteristics of image by Hog, is then classified using SVM, final to determine whether product is qualified, it is ensured that high accuracy rate improves detection efficiency, reduces labor workload, reduces the cost of company.
Description
Technical field
The present invention relates to one kind to detect Identification of Cracks system based on automobile metal plate work, in particular to a kind of to be based on SVM and Hog
Automobile metal plate work detect Identification of Cracks system, belong to technical field of automobile detection.
Background technique
Automobile metal plate work: automobile metal plate work is essential components on automobile, and sheet metal component is the group of automobile body-in-white
At part, each automobile all employs the sheet metal component that at least hundreds of is different, plays a very important role on automobile,
Quality problems directly influence safety, complete vehicle quality and the productive temp of main engine plants of automobile.
Automobile metal plate work detection: be directed to each automobile metal plate work product, need detect be outer profile size, hole count size,
Angle and face crack etc., these can return image to classify, determine if to close by the image of detection sheet metal component
Lattice.These problems can be converted to two classification problem of image.
SVM: full name is support vector machines (Support Vector Machine), is the learning model for having supervision,
Commonly used to carry out pattern-recognition, classification and regression analysis.Its main thought may be summarized to be two o'clock, and it is that linear can
Point situation is analyzed；It is the case where for linearly inseparable, by using non-linear map that the low-dimensional input space is linear
Inseparable sample, which is converted into high-dimensional feature space, makes its linear separability, so that high-dimensional feature space uses linear algorithm pair
The nonlinear characteristic of sample carries out linear analysis and is possibly realized.It is mainly used to determine whether picture is qualified in the system.
Hog: histograms of oriented gradients (Histogram of Oriented Gradient, HOG) is characterized in that one kind exists
It is used to carry out the feature description of object detection in computer vision and image procossing.HOG is by calculating and statistical picture partial zones
The gradient orientation histogram in domain carrys out constitutive characteristic.
, there is the following in the problem of artificial detection: taking long time, inefficiency；It is easy to appear missing inspection and erroneous detection；Professional skill
Energy testing staff's demand is larger；Testing staff's labor intensity is very big；Without detection process data, it can not reflect production process data.
Summary of the invention
The technical problem to be solved by the present invention is to overcome the defects of the prior art, provide a kind of vapour based on SVM and Hog
Vehicle sheet metal component detects Identification of Cracks system.
In order to solve the above-mentioned technical problems, the present invention provides the following technical solutions:
The present invention provides a kind of, and the automobile metal plate work based on SVM and Hog detects Identification of Cracks system, and specific testing process is such as
Under:
1) picture feature is extracted using hog；
2) SVM predicted pictures label is used；Confirm whether detection piece is qualified.
As a preferred technical solution of the present invention, it is as follows that Hog extracts characterization method:
1) color and gamma normalization
In order to reduce the influence of illumination factor, it is necessary first to whole image be standardized (normalization), in the texture of image
In intensity, the specific gravity of local surface layer exposure contribution is larger, so, this compression processing can be effectively reduced image local
Shade and illumination variation；
2) image gradient is calculated
The gradient of image abscissa and ordinate direction is calculated, and calculates the gradient direction value of each location of pixels accordingly；Derivation
Operation can not only capture profile, the shadow and some texture informations, moreover it is possible to the influence that further weakened light shines；
Most common method is: simply using an one-dimensional discrete differential template in one direction or simultaneously in level
Image is handled in vertical both direction, more precisely, this method needs to filter out in image using filter kernel
Color or the violent data of variation；
3) histogram in direction is constructed
Each of cell factory pixel is all some histogram channel ballot based on direction；
Ballot is to take the mode of Nearest Neighbor with Weighted Voting, i.e., each ticket is all with weight, this weight is according to the pixel
Gradient amplitude is calculated, this weight can be indicated using amplitude itself or its function, and actual test shows: being used
Amplitude indicates that weight can obtain optimal effect, it is of course also possible to select the function of amplitude to indicate, such as square of amplitude
Root, amplitude square, the clipped form of amplitude etc.；Cell factory can be rectangle, be also possible to star-shaped, histogram channel
That to be evenly distributed in 0-1800(undirected) or 0-3600(it is oriented) in range, it has been investigated that, it is straight using undirected gradient and 9
Square figure channel can obtain optimal effect in pedestrian detection test；
4) cell factory is combined into big section
The variation of the variation and foreground-background contrast shone due to local light, so that the variation range of gradient intensity is very big,
This just needs to normalize gradient intensity, and normalization can further compress illumination, shade and edge；
The method taken is: each cell factory is combined into big, the coconnected section in space；In this way, HOG descriptor just becomes
At a vector as composed by the histogram component of all cell factories in each section, these sections are mutual overlappings, this
Mean that: the output of each cell factory repeatedly acts on final describer；
There are two main geometries in section --- and rectangle section (R-HOG) and annular section (C-HOG), the section R-HOG are big
It is some rectangular grid on body, it can be characterized there are three parameter: the number of cell factory, each cell in each section
The number of pixel, the histogram number of active lanes of each cell in unit；
5) HOG feature is collected
The HOG feature of extraction is input in SVM classifier, SVM classifies.
As a preferred technical solution of the present invention, svm classifier process is as follows:
SVM is the identification and classification device defined by Optimal Separating Hyperplane, that is to say, that the training sample of given one group of tape label is calculated
Method will export an optimal hyperlane and classify to new samples (test sample)；
How the linearly separable set being made of for one two-dimensional coordinate point finds straight line and separates two class coordinates；
We can intuitively be defined as follows rule: if as soon as the straight line of segmentation is too close from coordinate points, it is not best
, because it cannot can correctly promote noise-sensitive, therefore, our target is to find a cut-off rule, it will be from institute
Some sample points are all remote as far as possible；
SVM algorithm is exactly to look for a hyperplane, and it wants maximum to the distance of closest to him training sample, i.e. optimum segmentation
Hyperplane maximizes training sample boundary；
To linear separability collection, the interface for dividing sample correctly can be found, and has infinite multiple, which is optimal necessary
A kind of optimal boundary criterion is found, the interval for keeping two quasi-modes separated is maximum；
Separating hyperplane: the above-mentioned straight line that Segmentation of Data Set comes is called separating hyperplane；
Hyperplane: if data set is N-dimensional, certain object for just needing N-1 to tie up is split data, the object
Hyperplane is done, that is, the decision boundary classified；
Interval: the distance of a point to divisional plane, referred to as distance of the point relative to divisional plane；
All points are to 2 times of the minimum interval of divisional plane, the referred to as interval of classifier or data set in data set；
Largest interval: SVM classifier is to look for maximum data set interval；
Supporting vector: point those of nearest from segmentation hyperplane.
The beneficial effects obtained by the present invention are as follows being: the present invention extracts characteristics of image by Hog, is then divided using SVM
Class, it is final to determine whether product is qualified, it is ensured that high accuracy rate improves detection efficiency, reduces labor workload, reduces public
The cost of department.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is overall flow figure of the invention；
Fig. 2 is the two-dimensional linear figure of SVM of the present invention；
Fig. 3 is SVM of the present invention linearly boundary figure.
Specific embodiment
Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings, it should be understood that preferred reality described herein
Apply example only for the purpose of illustrating and explaining the present invention and is not intended to limit the present invention.
Embodiment 1
As shown in Figure 1-3, the present invention provides a kind of, the automobile metal plate work based on SVM and Hog detects Identification of Cracks system, specifically
Testing process is as follows:
Picture feature is extracted using hog；
Using SVM predicted pictures label；Confirm whether detection piece is qualified.
It is as follows that Hog extracts characterization method:
Color and gamma normalization
In order to reduce the influence of illumination factor, it is necessary first to whole image be standardized (normalization), in the texture of image
In intensity, the specific gravity of local surface layer exposure contribution is larger, so, this compression processing can be effectively reduced image local
Shade and illumination variation；
Calculate image gradient
The gradient of image abscissa and ordinate direction is calculated, and calculates the gradient direction value of each location of pixels accordingly；Derivation
Operation can not only capture profile, the shadow and some texture informations, moreover it is possible to the influence that further weakened light shines；
Most common method is: simply using an one-dimensional discrete differential template in one direction or simultaneously in level
Image is handled in vertical both direction, more precisely, this method needs to filter out in image using filter kernel
Color or the violent data of variation；
Construct the histogram in direction
Each of cell factory pixel is all some histogram channel ballot based on direction；
Ballot is to take the mode of Nearest Neighbor with Weighted Voting, i.e., each ticket is all with weight, this weight is according to the pixel
Gradient amplitude is calculated, this weight can be indicated using amplitude itself or its function, and actual test shows: being used
Amplitude indicates that weight can obtain optimal effect, it is of course also possible to select the function of amplitude to indicate, such as square of amplitude
Root, amplitude square, the clipped form of amplitude etc.；Cell factory can be rectangle, be also possible to star-shaped, histogram channel
That to be evenly distributed in 0-1800(undirected) or 0-3600(it is oriented) in range, it has been investigated that, it is straight using undirected gradient and 9
Square figure channel can obtain optimal effect in pedestrian detection test；
Cell factory is combined into big section
The variation of the variation and foreground-background contrast shone due to local light, so that the variation range of gradient intensity is very big,
This just needs to normalize gradient intensity, and normalization can further compress illumination, shade and edge；
The method taken is: each cell factory is combined into big, the coconnected section in space；In this way, HOG descriptor just becomes
At a vector as composed by the histogram component of all cell factories in each section, these sections are mutual overlappings, this
Mean that: the output of each cell factory repeatedly acts on final describer；
There are two main geometries in section --- and rectangle section (R-HOG) and annular section (C-HOG), the section R-HOG are big
It is some rectangular grid on body, it can be characterized there are three parameter: the number of cell factory, each cell in each section
The number of pixel, the histogram number of active lanes of each cell in unit；
Collect HOG feature
The HOG feature of extraction is input in SVM classifier, SVM classifies.
Compared to other algorithms, hog extracts feature and has the advantage that
Firstly, since HOG be operated on the local pane location of image, so it to image geometry and optical deformation all
It is able to maintain good invariance, both deformation only appear on bigger space field；
Secondly, under the conditions ofs thick airspace sampling, fine direction sampling and the normalization of stronger indicative of local optical etc., as long as real
Body posture can allow pedestrian to have some subtle limb actions, these subtle movements can be ignored without influencing to detect
Effect；
Therefore HOG feature is detected particularly suitable for doing the entity in image.
SVM is the identification and classification device defined by Optimal Separating Hyperplane, that is to say, that the training sample of given one group of tape label
This, algorithm will export an optimal hyperlane and classify to new samples (test sample)；
How the linearly separable set being made of for one two-dimensional coordinate point finds straight line and separates two class coordinates；
We can intuitively be defined as follows rule: if as soon as the straight line of segmentation is too close from coordinate points, it is not best
, because it cannot can correctly promote noise-sensitive, therefore, our target is to find a cut-off rule, it will be from institute
Some sample points are all remote as far as possible；
SVM algorithm is exactly to look for a hyperplane, and it wants maximum to the distance of closest to him training sample, i.e. optimum segmentation
Hyperplane maximizes training sample boundary；
To linear separability collection, the interface for dividing sample correctly can be found, and has infinite multiple, which is optimal necessary
A kind of optimal boundary criterion is found, the interval for keeping two quasi-modes separated is maximum；
Separating hyperplane: the above-mentioned straight line that Segmentation of Data Set comes is called separating hyperplane；
Hyperplane: if data set is N-dimensional, certain object for just needing N-1 to tie up is split data, the object
Hyperplane is done, that is, the decision boundary classified；
Interval: the distance of a point to divisional plane, referred to as distance of the point relative to divisional plane；
All points are to 2 times of the minimum interval of divisional plane, the referred to as interval of classifier or data set in data set；
Largest interval: SVM classifier is to look for maximum data set interval；
Supporting vector: point those of nearest from segmentation hyperplane.
The beneficial effects obtained by the present invention are as follows being: the present invention extracts characteristics of image by Hog, is then divided using SVM
Class, it is final to determine whether product is qualified, it is ensured that high accuracy rate improves detection efficiency, reduces labor workload, reduces public
The cost of department.
Finally, it should be noted that the foregoing is only a preferred embodiment of the present invention, it is not intended to restrict the invention,
Although the present invention is described in detail referring to the foregoing embodiments, for those skilled in the art, still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features.
All within the spirits and principles of the present invention, any modification, equivalent replacement, improvement and so on should be included in of the invention
Within protection scope.
Claims (3)
1. a kind of automobile metal plate work based on SVM and Hog detects Identification of Cracks system, which is characterized in that specific testing process is such as
Under:
1) picture feature is extracted using hog；
2) SVM predicted pictures label is used；Confirm whether detection piece is qualified.
2. a kind of automobile metal plate work based on SVM and Hog according to claim 1 detects Identification of Cracks system, feature
It is, it is as follows that Hog extracts characterization method:
1) color and gamma normalization
In order to reduce the influence of illumination factor, it is necessary first to whole image be standardized (normalization), in the texture of image
In intensity, the specific gravity of local surface layer exposure contribution is larger, so, this compression processing can be effectively reduced image local
Shade and illumination variation；
2) image gradient is calculated
The gradient of image abscissa and ordinate direction is calculated, and calculates the gradient direction value of each location of pixels accordingly；Derivation
Operation can not only capture profile, the shadow and some texture informations, moreover it is possible to the influence that further weakened light shines；
Most common method is: simply using an one-dimensional discrete differential template in one direction or simultaneously in level
Image is handled in vertical both direction, more precisely, this method needs to filter out in image using filter kernel
Color or the violent data of variation；
3) histogram in direction is constructed
Each of cell factory pixel is all some histogram channel ballot based on direction；
Ballot is to take the mode of Nearest Neighbor with Weighted Voting, i.e., each ticket is all with weight, this weight is according to the pixel
Gradient amplitude is calculated, this weight can be indicated using amplitude itself or its function, and actual test shows: being used
Amplitude indicates that weight can obtain optimal effect, it is of course also possible to select the function of amplitude to indicate, such as square of amplitude
Root, amplitude square, the clipped form of amplitude etc.；Cell factory can be rectangle, be also possible to star-shaped, histogram channel
That to be evenly distributed in 0-1800(undirected) or 0-3600(it is oriented) in range, it has been investigated that, it is straight using undirected gradient and 9
Square figure channel can obtain optimal effect in pedestrian detection test；
4) cell factory is combined into big section
The variation of the variation and foreground-background contrast shone due to local light, so that the variation range of gradient intensity is very big,
This just needs to normalize gradient intensity, and normalization can further compress illumination, shade and edge；
The method taken is: each cell factory is combined into big, the coconnected section in space；In this way, HOG descriptor just becomes
At a vector as composed by the histogram component of all cell factories in each section, these sections are mutual overlappings, this
Mean that: the output of each cell factory repeatedly acts on final describer；
There are two main geometries in section --- and rectangle section (R-HOG) and annular section (C-HOG), the section R-HOG are big
It is some rectangular grid on body, it can be characterized there are three parameter: the number of cell factory, each cell in each section
The number of pixel, the histogram number of active lanes of each cell in unit；
5) HOG feature is collected
The HOG feature of extraction is input in SVM classifier, SVM classifies.
3. a kind of automobile metal plate work based on SVM and Hog according to claim 1 detects Identification of Cracks system, feature
It is, svm classifier process is as follows:
SVM is the identification and classification device defined by Optimal Separating Hyperplane, that is to say, that the training sample of given one group of tape label is calculated
Method will export an optimal hyperlane and classify to new samples (test sample)；
How the linearly separable set being made of for one two-dimensional coordinate point finds straight line and separates two class coordinates；
We can intuitively be defined as follows rule: if as soon as the straight line of segmentation is too close from coordinate points, it is not best
, because it cannot can correctly promote noise-sensitive, therefore, our target is to find a cut-off rule, it will be from institute
Some sample points are all remote as far as possible；
SVM algorithm is exactly to look for a hyperplane, and it wants maximum to the distance of closest to him training sample, i.e. optimum segmentation
Hyperplane maximizes training sample boundary；
To linear separability collection, the interface for dividing sample correctly can be found, and has infinite multiple, which is optimal necessary
A kind of optimal boundary criterion is found, the interval for keeping two quasi-modes separated is maximum；
Separating hyperplane: the above-mentioned straight line that Segmentation of Data Set comes is called separating hyperplane；
Hyperplane: if data set is N-dimensional, certain object for just needing N-1 to tie up is split data, the object
Hyperplane is done, that is, the decision boundary classified；
Interval: the distance of a point to divisional plane, referred to as distance of the point relative to divisional plane；
All points are to 2 times of the minimum interval of divisional plane, the referred to as interval of classifier or data set in data set；
Largest interval: SVM classifier is to look for maximum data set interval；
Supporting vector: point those of nearest from segmentation hyperplane.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810715062.9A CN108960325A (en) | 2018-07-03 | 2018-07-03 | A kind of automobile metal plate work detection Identification of Cracks system based on SVM and Hog |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810715062.9A CN108960325A (en) | 2018-07-03 | 2018-07-03 | A kind of automobile metal plate work detection Identification of Cracks system based on SVM and Hog |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108960325A true CN108960325A (en) | 2018-12-07 |
Family
ID=64485269
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810715062.9A Pending CN108960325A (en) | 2018-07-03 | 2018-07-03 | A kind of automobile metal plate work detection Identification of Cracks system based on SVM and Hog |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108960325A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114839054A (en) * | 2022-03-16 | 2022-08-02 | 徐州大成环境科技有限公司 | Building material strength detection method and system based on artificial intelligence |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106846313A (en) * | 2017-01-23 | 2017-06-13 | 广东工业大学 | Surface Flaw Detection method and apparatus |
| CN107169953A (en) * | 2017-04-07 | 2017-09-15 | 西安电子科技大学 | Bridge concrete surface crack detection method based on HOG features |
-
2018
- 2018-07-03 CN CN201810715062.9A patent/CN108960325A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106846313A (en) * | 2017-01-23 | 2017-06-13 | 广东工业大学 | Surface Flaw Detection method and apparatus |
| CN107169953A (en) * | 2017-04-07 | 2017-09-15 | 西安电子科技大学 | Bridge concrete surface crack detection method based on HOG features |
Non-Patent Citations (1)
| Title |
|---|
| NAVNEET DALAL等: "Histograms of Oriented Gradients for Human Detection", 《PROCEEDINGS OF THE 2005 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114839054A (en) * | 2022-03-16 | 2022-08-02 | 徐州大成环境科技有限公司 | Building material strength detection method and system based on artificial intelligence |
| CN114839054B (en) * | 2022-03-16 | 2023-02-28 | 徐州大成环境科技有限公司 | Building material strength detection method and system based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2017190574A1 (en) | Fast pedestrian detection method based on aggregation channel features | |
| CN103034838B (en) | A kind of special vehicle instrument type identification based on characteristics of image and scaling method | |
| CN109190475B (en) | Face recognition network and pedestrian re-recognition network collaborative training method | |
| CN110728252B (en) | Face detection method applied to regional personnel motion trail monitoring | |
| CN109684906B (en) | Method for detecting red fat bark beetles based on deep learning | |
| Malik et al. | Detection and recognition of traffic signs from road scene images | |
| CN111402224B (en) | Target identification method for power equipment | |
| CN110287787B (en) | Image recognition method, image recognition device and computer-readable storage medium | |
| CN110335233B (en) | Highway guardrail plate defect detection system and method based on image processing technology | |
| CN111080691A (en) | Infrared hot spot detection method and device for photovoltaic module | |
| Sugiharto et al. | Traffic sign detection based on HOG and PHOG using binary SVM and k-NN | |
| CN108073940B (en) | Method for detecting 3D target example object in unstructured environment | |
| CN115731493A (en) | Rainfall micro physical characteristic parameter extraction and analysis method based on video image recognition | |
| CN113591973B (en) | Intelligent comparison method for appearance state change of track plate | |
| Budianto | Automatic License Plate Recognition: A Review with Indonesian Case Study | |
| CN108960325A (en) | A kind of automobile metal plate work detection Identification of Cracks system based on SVM and Hog | |
| CN113989196A (en) | Vision-based earphone silica gel gasket appearance defect detection method | |
| CN117475353A (en) | Video-based abnormal smoke identification method and system | |
| Pant et al. | Automatic Nepali number plate recognition with support vector machines | |
| CN110516547B (en) | Fake-licensed vehicle detection method based on weighted non-negative matrix factorization | |
| CN116934674A (en) | Large-granularity silicon material detection and removal system based on YOLOv8 network | |
| Guo et al. | Fault diagnosis of power equipment based on infrared image analysis | |
| CN115810161A (en) | Transformer substation fire identification method and system | |
| CN113570587A (en) | Photovoltaic cell broken grid detection method and system based on computer vision | |
| Evstafev et al. | Surface Defect Detection and Recognition Based on CNN |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181207 |