CN109741314A - A kind of visible detection method and system of part - Google Patents
A kind of visible detection method and system of part Download PDFInfo
- Publication number
- CN109741314A CN109741314A CN201811639460.3A CN201811639460A CN109741314A CN 109741314 A CN109741314 A CN 109741314A CN 201811639460 A CN201811639460 A CN 201811639460A CN 109741314 A CN109741314 A CN 109741314A
- Authority
- CN
- China
- Prior art keywords
- image
- circular hole
- detection zone
- radius
- detection
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The invention proposes a kind of visible detection method of part and systems, and the system comprises following steps: obtaining the image of part, and carry out pretreatment and binary conversion treatment to acquired part image；Obtain the edge shape feature of part in part image；The detection zone of part image is obtained, the detection zone is includes round-meshed region in part；There is circular hole by whether there is in hough-circle transform recognition detection region, and if it exists, then obtain the radius size and location information of each existing circular hole；According to the round-meshed radius size of institute, location information identification part category in the edge shape characteristic information and part image of acquired part or judge whether part is qualified.The present invention can with more efficient, accurately complete part category detection, avoid the problem of desk checking leads to error due to visual fatigue in the process, and the position of circular feature can be more accurately understood according to the visual signature information that detection zone is extracted, precision is high and high-efficient.
Description
Technical field
The present invention relates to piece test technical fields, and in particular to a kind of visible detection method and system of part.
Background technique
During production industrially to assembly, the transport in batches of various parts is had, due to various parts
Type and large number of, while some design of part is irregular, what is be difficult to avoid that during manual operation will appear zero
The case where part type malfunctions.In order to avoid this kind of situation, factory is seen before producing finished product by the method for desk checking
Examine whether part malfunctions；However during desk checking, it is easy to produce fatigue phenomenon, when observing excessive part in a period
When, it is easy to appear visual fatigue, causes to be easy to happen mistake when examining, comes into the market so as to cause defective goods, cause enterprise can not
The loss made up.In addition, desk checking is also difficult detected in the case where part has assembled.
Summary of the invention
In view of the deficiencies of the prior art, the present invention proposes the visible detection method and system of a kind of part, avoids artificial
The problem of error is led to due to visual fatigue in checkout procedure.
The technical scheme of the present invention is realized as follows: a kind of visible detection method of part, comprising the following steps:
Step 1, the image of part is obtained, and pretreatment and binary conversion treatment are carried out to acquired part image；
Step 2, the edge shape feature of part in part image is obtained；
Step 3, the detection zone of the part image is obtained, the detection zone is includes round-meshed area in part
Domain；
Step 4, there is circular hole by whether there is in hough-circle transform recognition detection region, and if it exists, present in then obtaining
Each circular hole radius size and location information；If it does not exist, then the frame part image after obtaining prefixed time interval
Afterwards, return step 1；
Step 5, according to the round-meshed radius size of institute in the edge shape characteristic information and part image of acquired part,
Location information identifies the part category or judges whether the part is qualified.
Further, step 1 includes:
Step 101, the image for obtaining part is converted to gray scale by gradation conversion, enhancing to acquired part image
Image；
Step 102, Gaussian Blur processing is carried out to the gray level image, the Gaussian Blur handles function used and is
Step 103, binary conversion treatment is carried out to pretreated part image.
Further, step 103 includes:
The gray value of each pixel in bianry image is compared with preset threshold, if more than the preset threshold,
Then assign the pixel gray value 255；Otherwise, then the pixel gray value 0 is assigned.
Further, step 3 includes:
Step 301, it is converted by morphological image and coarse extraction is carried out to the profile of the circular hole in part image；
Step 302, the detection zone of the part image is obtained using minimum circumscribed rectangle according to the circular hole after coarse extraction,
The detection zone is includes round-meshed region in part.
Further, in step 4, the radius size and location information for obtaining each existing circular hole include:
Edge detection is carried out with Canny algorithm first, obtains the binary map on all circular hole boundaries in detection zone；
The gradient of a certain circular hole in inspection area is calculated with Sobel operator, traverses the non-zero point in edge binary map, along
Gradient direction and opposite direction setting-out section, the starting point and length of line segment are determined that the point for passing through line segment is cumulative by the parameter set
It is counted in device, counts more points and be more likely to become the center of circle, the coordinate for obtaining the center of circle at this time is the location information of the circular hole；
It sorts to all non-zero distances of the point away from the center of circle, is successively counted since minor radius from small to large, differed in some amount
Point to be all approximately considered be the same circle, count all points for belonging to the circle；Gradually amplification radius continues to count, and compares two and half
Line density=points/radius of diameter point, line density is higher, and the confidence level of radius is bigger, in parameter allowed band repeatedly more than
For step until obtaining optimal radius, optimal radius at this time is the radius of the circular hole.
A kind of visible detection method of part, including image collection module, image pre-processing module, binarization block, zero
Part edge obtains module, detection zone obtains module, circular hole feature obtains module and judgment module, wherein
Image collection module, for obtaining the image of part；
Image pre-processing module, for being pre-processed to acquired part image；
Binarization block, for carrying out binaryzation to pretreated part image；
Part edge obtains module, for obtaining the edge shape feature of part in part image；
Detection zone obtains module, and for obtaining the detection zone of the part image, the detection zone is in part
Include the round-meshed region of institute；
Circular hole feature obtains module, for having circular hole by whether there is in hough-circle transform recognition detection region, if depositing
Then obtaining the radius size and location information of each existing circular hole；If it does not exist, then scheming after prefixed time interval
After obtaining a frame part image in acquisition module；
Judgment module is big according to the round-meshed radius of institute in the edge shape characteristic information and part image of acquired part
Small, location information identifies the part category or judges whether the part is qualified.
Further, described image preprocessing module includes greyscale image transitions unit and Gaussian Blur processing unit,
In
Greyscale image transitions unit, for being converted to grayscale image by gradation conversion, enhancing to acquired part image
Picture；
Gaussian Blur processing unit, for carrying out Gaussian Blur processing, the Gaussian Blur processing to the gray level image
Function used is
Further, the binary processing module includes threshold setting unit and assignment unit, wherein
Threshold setting unit, for setting preset threshold；
Assignment unit, for the gray value of each pixel in bianry image to be compared with preset threshold, if more than
The preset threshold then assigns the pixel gray value 255；Otherwise, then the pixel gray value 0 is assigned.
Further, it includes that circular hole wheel contours extract unit and detection zone obtain list that the detection zone, which obtains module,
Member, wherein
Circular hole wheel contours extract unit, for being carried out by profile of the morphological image transformation to the circular hole in part image
Coarse extraction；
Detection zone acquiring unit, for obtaining the part drawing using minimum circumscribed rectangle according to the circular hole after coarse extraction
The detection zone of picture, the detection zone is includes round-meshed region in part.
Further, the circular hole feature obtain module include bore edges detection unit, location information acquiring unit and
Radius size acquiring unit, wherein
Bore edges detection unit carries out edge detection with Canny algorithm, obtains two of the boundary of circular hole in detection zone
Value figure；
Location information acquiring unit, for obtaining the location information of circular hole；
Radius size acquiring unit, for obtaining the radius size of circular hole.
Compared with prior art, the invention has the following advantages that the present invention is carried out according to acquired part image to be measured
The edge shape of edge detection acquisition part；Then edge detection is carried out by the circular hole in part image again, and extracts and includes
The round-meshed detection zone image of institute is split extraction with background interference region to will test region；Pass through Hough ladder again
Degree method obtains the round-meshed radius size of institute and location information in all detection zones；Finally according to the edge shape of acquired part
The round-meshed radius size of institute, location information identify the part category or judge described zero in shape characteristic information and part image
Whether part is qualified.The present invention can with more efficient, accurately complete part category detection, and can according to detection zone extract view
Feel that characteristic information more accurately understands the position of circular feature, has many advantages, such as that precision is high, non-contact, high-efficient, at low cost.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention without any creative labor, may be used also for those of ordinary skill in the art
To obtain other drawings based on these drawings.
Fig. 1 is the flow chart of one embodiment of visible detection method of part of the present invention；
Fig. 2 is the structural block diagram of one embodiment of vision detection system of part of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
Refering to fig. 1, a kind of visible detection method of part disclosed in embodiment of the present invention, comprising the following steps:
Step 1, the image of part is obtained, and pretreatment and binary conversion treatment are carried out to acquired part image；
In embodiment of the present invention, shooting or recorded video, acquired part can be carried out to part by camera
Image is a wherein frame video pictures for photo captured by camera or video stream data.
Step 2, the edge shape feature of part in part image is obtained；
In embodiment of the present invention, it can be separated by multichannel and edge gradient feature extracting method obtains in part image
The edge shape feature of part.
Step 3, the detection zone of the part image is obtained, the detection zone is includes round-meshed area in part
Domain；
In embodiment of the present invention, it will be extracted comprising round-meshed Minimum Area with minimum circumscribed rectangle in part image
Out, so that background interference be avoided to impact the testing result of circular hole.
Step 4, there is circular hole by whether there is in hough-circle transform recognition detection region, and if it exists, present in then obtaining
Each circular hole radius size and location information；If it does not exist, then the frame part image after obtaining prefixed time interval
Afterwards, return step 1；
In embodiment of the present invention, hough-circle transform can identify the circular hole in detection zone, and may recognize that
The radius size and location information of each circular hole in detection zone.
Step 5, according to the round-meshed radius size of institute in the edge shape characteristic information and part image of acquired part,
Location information identifies the part category or judges whether the part is qualified.
As shown in Fig. 2, the present invention also proposes a kind of visible detection method of part, including image collection module, image are pre-
Processing module, binarization block, part edge obtain module, detection zone obtains module, circular hole feature obtains module and judgement
Module.
Wherein, modules are described below:
Image collection module, for obtaining the image of part；
Image pre-processing module, for being pre-processed to acquired part image；
Binarization block, for carrying out binaryzation to pretreated part image；
Part edge obtains module, for obtaining the edge shape feature of part in part image；
Detection zone obtains module, and for obtaining the detection zone of the part image, the detection zone is in part
Include the round-meshed region of institute；
Circular hole feature obtains module, for having circular hole by whether there is in hough-circle transform recognition detection region, if depositing
Then obtaining the radius size and location information of each existing circular hole；If it does not exist, then scheming after prefixed time interval
After obtaining a frame part image in acquisition module；
Judgment module is big according to the round-meshed radius of institute in the edge shape characteristic information and part image of acquired part
Small, location information identifies the part category or judges whether the part is qualified.
In this embodiment, the visible detection method of part is holding using the vision detection system of part as step
Row object, can also execution object using the module in system as step.Specifically, step 1 is by image collection module, image
The execution object of preprocessing module, binarization block as step, step 2 obtain execution of the module as step by part edge
Object, step 3 obtain execution object of the module as step by detection zone, and step 4 obtains module as step by circular hole feature
Rapid execution object, execution object of the step 5 by judgment module as step.
When classifying to part: by the way that the edge shape characteristic information of part detected, the radius of circular hole is big
The radius of the edge shape characteristic information of small, circular hole location information and storage various types of part in a computer, circular hole
Size, circular hole location information compared one by one respectively, if each Testing index (edge shape characteristic information, circular hole of part
Radius size, circular hole position) in the allowable range of error of A class part, then part to be detected belongs to A class part, and
It will test result to show by display screen or prompt by warning device.
It is same when to part, whether qualification is detected: by believing the edge shape feature of part detected
The edge of the standardized element of breath, the location information of the radius size of circular hole, circular hole and storage part to be detected in a computer
Shape feature information, the radius size of circular hole, circular hole location information compared one by one respectively, if each Testing index (part
Edge shape characteristic information, the radius size of circular hole, circular hole position) in the allowable range of error of standardized element, then to
Detecting part is qualification, otherwise to be unqualified, and will test result and is shown by display screen or mentioned by warning device
Show.
The present invention mainly detects quarter round-meshed part, according to the radius of circular hole on the profile and part of part
Size, center location distinguish the type of part, or determine whether the part to be tested is qualified.The present invention is in real-time detection
Historical part type data can be retained during part, can with more efficient, accurately complete part category detection, and can root
The position that circular feature is more accurately understood according to the visual signature information that detection zone is extracted, high, the non-contact, efficiency with precision
The advantages that high, at low cost.
In embodiment of the present invention, specifically, step 1 includes:
Step 101, the image for obtaining part is converted to gray level image to acquired part image；Specifically, passing through ash
Degree conversion, enhancing are converted to gray level image；
Specifically, gradation conversion process is that the part image (color image of RGB triple channel) that will acquire carries out image
Tri- components of RGB be weighted and averaged to obtain the gray level image that gray value is Gray, the gray processing function used are as follows:
Gray=(B+G+R)/3；
Or
Gray=0.072169B+0.715160G+0.212671R.
Step 102, Gaussian Blur processing is carried out to the gray level image, the Gaussian Blur handles function used and is
In embodiment of the present invention, Gaussian function is utilizedAs filter function, a volume is set
Each of product scanning part image pixel goes alternate template with the weighted average gray value of pixel in the determining field of template
The value of central pixel point.Embodiment of the present invention is handled by Gaussian Blur to reduce image noise.
Step 103, binary conversion treatment is carried out to pretreated part image.
Specifically, preset threshold is set first, then by the gray value and preset threshold of each pixel in bianry image
It compares, if more than the preset threshold, then assigns the pixel gray value 255；Otherwise, then the pixel gray value is assigned
0。
Correspondingly, image pre-processing module includes greyscale image transitions unit and Gaussian mode in the vision detection system of part
Paste processing unit, in which:
Greyscale image transitions unit, for being converted to grayscale image by gradation conversion, enhancing to acquired part image
Picture；
Gaussian Blur processing unit, for carrying out Gaussian Blur processing, the Gaussian Blur processing to the gray level image
Function used is
Specifically, binary processing module includes threshold setting unit and assignment unit in the vision detection system of part,
Wherein
Threshold setting unit, for setting preset threshold；
Assignment unit, for the gray value of each pixel in bianry image to be compared with preset threshold, if more than
The preset threshold then assigns the pixel gray value 255；Otherwise, then the pixel gray value 0 is assigned.
Wherein, step 101 can execution object by greyscale image transitions unit as step, step 102 can be by height mould
Paste execution object of the processing unit as step, step 103 can execution unit by binary processing module as step.
Specifically, obtaining part in part image by multichannel separation and edge gradient feature extracting method in step 2
Edge shape feature, specifically may include following sub-step:
Step 201,
Specifically, obtaining the edge shape of part in part image by multichannel separation and edge gradient feature extracting method
Shape feature:
Edge gradient feature extraction is horizontal, the vertical and diagonal line detected respectively in part image by four filters
Edge.According to detection as a result, the value of edge detector returns to a horizontal direction component Gx and vertical direction component Gy, thus
Edge gradient and direction can be determined by lower section formula:
Therefore, the angle at all edges is all in above-mentioned selected (0 °, 45 °, 90 °, 135 °) surrounding of four direction.
After the similar edge gradient for finding out part around aforementioned four direction according to the above method, two thresholdings are used
Height boundary, i.e. given threshold are defined respectively.Assuming that all edges should not be affected by noise and be continuous curves, because
High threshold is arranged for determining that determination is the curve at edge in this, then based on this, utilization orientation information trace those can chase after
The image border of track；When tracking the edge, using low threshold can allow tracking those contain edge region until finding down
The starting point of one curve；The edge shape of part in part image can be finally obtained after searching process.
Specifically, step 3 includes:
Step 301, it is converted by morphological image and coarse extraction is carried out to the profile of the circular hole in part image；
Step 302, the detection zone of the part image is obtained using minimum circumscribed rectangle according to the circular hole after coarse extraction,
The detection zone is includes round-meshed region in part.
Correspondingly, it includes circular hole wheel contours extract unit and inspection that detection zone, which obtains module, in the vision detection system of part
Area acquisition unit is surveyed, wherein
Circular hole wheel contours extract unit, for being carried out by profile of the morphological image transformation to the circular hole in part image
Coarse extraction；
Detection zone acquiring unit, for obtaining the part drawing using minimum circumscribed rectangle according to the circular hole after coarse extraction
The detection zone of picture, the detection zone is includes round-meshed region in part.
Wherein, step 301 can execution object by circular hole wheel contours extract unit as step, step 302 can be by detecting
Execution object of the area acquisition unit as step.
In embodiment of the present invention, slightly mentioned by the profile that morphological image converts the circular hole in acquisition part image
It takes, then obtains farthest avoiding uncorrelated comprising a round-meshed minimum detection region using minimum circumscribed rectangle
The interference in region (region for not including circular hole)；It obtains finally by Hough gradient method to each circular hole in minimum detection region
Radius size and center location.
Hough gradient method specifically: indicate a round C=(X with three parametersc, Yc, r), round expression formula is (Xc-a)2+
(Yc-b)2=r2, therefore problem can be converted into solving (a, b, r) parameter pair most by pixel.Therefore, by obtaining
Take round central coordinate of circle (a, b) that the coordinate of circular hole, the i.e. location information of circular hole can be obtained；By the radius r for obtaining circle
Obtain the radius size of circular hole.
Therefore, specifically, in step 4, the radius size and location information for obtaining each existing circular hole include:
Step 401, edge detection is carried out with Canny algorithm first, obtains the two-value on all circular hole boundaries in detection zone
Figure；
Step 402, the gradient of a certain circular hole in inspection area is calculated with Sobel operator, is traversed non-zero in edge binary map
Point, along gradient direction and opposite direction setting-out section, the starting point and length of line segment are determined by the parameter set, the point that line segment is passed through
It in accumulator number, counts more points and is more likely to become the center of circle, the coordinate for obtaining the center of circle at this time is the position of the circular hole
Information；
Step 403, it sorts to all non-zero distances of the point away from the center of circle, is successively counted since minor radius from small to large, differed
All being approximately considered in the point of some amount is the same circle, counts all points for belonging to the circle；Gradually amplification radius continues to count, than
Compared with line density=points/radius of two radius points, line density is higher, and the confidence level of radius is bigger, in parameter allowed band
For above step until obtaining optimal radius, optimal radius at this time is the radius of the circular hole repeatedly.
Equally 401~step 403 obtains the radius size and location information of other circular holes through the above steps.
Correspondingly, it includes bore edges detection unit, position that circular hole feature, which obtains module, in the visible detection method of part
Information acquisition unit and radius size acquiring unit, wherein
Bore edges detection unit carries out edge detection with Canny algorithm, obtains two of the boundary of circular hole in detection zone
Value figure；
Location information acquiring unit calculates the gradient of a certain circular hole in inspection area with Sobel operator, traverses edge two-value
Non-zero point in figure, along gradient direction and opposite direction setting-out section, the starting point and length of line segment are determined by the parameter set, by line
The point that section is passed through counts more points and is more likely to become the center of circle in accumulator number, obtain the center of circle at this time coordinate be should
The location information of circular hole；
Radius size acquiring unit, for sorting from small to large to all non-zero distances of the point away from the center of circle, since minor radius
It successively counts, it is the same circle that difference is all approximately considered in the point of some amount, counts all points for belonging to the circle；Gradually amplification half
Diameter continues to count, and compares the line density=points/radius of two radius points, and line density is higher, and the confidence level of radius is bigger, is joining
For above step until obtaining optimal radius, optimal radius at this time is the radius of the circular hole repeatedly in number allowed band；
Wherein, step 401 can execution object by bore edges detection unit as step, step 402 can believe by position
Cease execution object of the acquiring unit as step, step 403 can execution object by radius size acquiring unit as step.
The present invention is imaged part to be measured to obtain a frame target retrieval image, first passes through multi-channel feature decomposition, and combine
The edge shape of part in edge gradient feature, local shape factor part image；Then pass through edge detection, extraction target area
Domain shape feature is split extraction to detection zone and background interference region, the image for the detection zone isolated；Again
The round-meshed location information of institute and radius size are obtained according to the image of detection zone；Finally according to the edge shape of acquired part
The round-meshed radius size of institute, location information identify the part category or judge described zero in shape characteristic information and part image
Whether part is qualified.The present invention is in real-time detection part category and retains historical part type data, can with more efficient, accurately
Part category detection is completed, and can more accurately understand the position of circular feature according to the visual signature information that detection zone is extracted
It sets, has many advantages, such as that precision is high, non-contact, high-efficient, at low cost.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Within mind and principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of visible detection method of part, which comprises the following steps:
Step 1, the image of part is obtained, and pretreatment and binary conversion treatment are carried out to acquired part image；
Step 2, the edge shape feature of part in part image is obtained；
Step 3, the detection zone of the part image is obtained, the detection zone is includes round-meshed region in part；
Step 4, there is circular hole by whether there is in hough-circle transform recognition detection region, and if it exists, every present in then obtaining
The radius size and location information of one circular hole；If it does not exist, then it after the frame part image after obtaining prefixed time interval, returns
Return step 1；
Step 5, according to the round-meshed radius size of institute, position in the edge shape characteristic information and part image of acquired part
Information identifies the part category or judges whether the part is qualified.
2. the visible detection method of part as described in claim 1, which is characterized in that step 1 includes:
Step 101, the image for obtaining part is converted to gray level image by gradation conversion, enhancing to acquired part image；
Step 102, Gaussian Blur processing is carried out to the gray level image, the Gaussian Blur handles function used and is
Step 103, binary conversion treatment is carried out to pretreated part image.
3. the visible detection method of part as claimed in claim 2, which is characterized in that step 103 includes:
The gray value of each pixel in bianry image is compared with preset threshold, if more than the preset threshold, is then assigned
Give the pixel gray value 255；Otherwise, then the pixel gray value 0 is assigned.
4. the visible detection method of part as claimed in claim 3, which is characterized in that step 3 includes:
Step 301, it is converted by morphological image and coarse extraction is carried out to the profile of the circular hole in part image；
Step 302, the detection zone of the part image is obtained using minimum circumscribed rectangle according to the circular hole after coarse extraction, it is described
Detection zone is includes round-meshed region in part.
5. the visible detection method of part as claimed in claim 4, which is characterized in that in step 4, obtain each existing
The radius size and location information of circular hole include:
Edge detection is carried out with Canny algorithm first, obtains the binary map on all circular hole boundaries in detection zone；
The gradient of a certain circular hole in inspection area is calculated with Sobel operator, the non-zero point in edge binary map is traversed, along gradient
Direction and opposite direction setting-out section, the starting point and length of line segment are determined that the point for passing through line segment is in accumulator by the parameter set
It counts, counts more points and be more likely to become the center of circle, the coordinate for obtaining the center of circle at this time is the location information of the circular hole；
It sorts to all non-zero distances of the point away from the center of circle, is successively counted since minor radius from small to large, differ the point in some amount
All being approximately considered is the same circle, counts all points for belonging to the circle；Gradually amplification radius continues to count, and compares two radius points
Line density=points/radius, line density is higher, and the confidence level of radius is bigger, the above step repeatedly in parameter allowed band
Until obtaining optimal radius, optimal radius at this time is the radius of the circular hole.
6. a kind of visible detection method of part, which is characterized in that including image collection module, image pre-processing module, two-value
Change module, part edge acquisition module, detection zone and obtains module, circular hole feature acquisition module and judgment module, wherein
Image collection module, for obtaining the image of part；
Image pre-processing module, for being pre-processed to acquired part image；
Binarization block, for carrying out binaryzation to pretreated part image；
Part edge obtains module, for obtaining the edge shape feature of part in part image；
Detection zone obtains module, and for obtaining the detection zone of the part image, the detection zone, which is in part, includes
The round-meshed region of institute；
Circular hole feature obtains module, for having circular hole by whether there is in hough-circle transform recognition detection region, and if it exists, then
Obtain the radius size and location information of each existing circular hole；If it does not exist, then it is obtained after prefixed time interval in image
After obtaining a frame part image in modulus block；
Judgment module, according to the round-meshed radius size of institute in the edge shape characteristic information and part image of acquired part,
Location information identifies the part category or judges whether the part is qualified.
7. the visible detection method of part as claimed in claim 6, which is characterized in that described image preprocessing module includes gray scale
Image conversion unit and Gaussian Blur processing unit, wherein
Greyscale image transitions unit, for being converted to gray level image by gradation conversion, enhancing to acquired part image；
Gaussian Blur processing unit, used in gray level image progress Gaussian Blur processing, the Gaussian Blur is handled
Function is
8. the visible detection method of part as claimed in claim 7, which is characterized in that the binary processing module includes threshold value
Setup unit and assignment unit, wherein
Threshold setting unit, for setting preset threshold；
Assignment unit, for comparing the gray value of each pixel in bianry image with preset threshold, if more than described
Preset threshold then assigns the pixel gray value 255；Otherwise, then the pixel gray value 0 is assigned.
9. the visible detection method of part as claimed in claim 8, which is characterized in that it includes circle that the detection zone, which obtains module,
Contours extract unit and detection zone acquiring unit are taken turns in hole, wherein
Circular hole wheel contours extract unit, for slightly being mentioned by morphological image transformation to the profile of the circular hole in part image
It takes；
Detection zone acquiring unit, for obtaining the part image using minimum circumscribed rectangle according to the circular hole after coarse extraction
Detection zone, the detection zone is includes round-meshed region in part.
10. the visible detection method of part as claimed in claim 9, which is characterized in that the circular hole feature obtains module and includes
Bore edges detection unit, location information acquiring unit and radius size acquiring unit, wherein
Bore edges detection unit carries out edge detection with Canny algorithm, obtains the two-value on the boundary of circular hole in detection zone
Figure；
Location information acquiring unit, for obtaining the location information of circular hole；
Radius size acquiring unit, for obtaining the radius size of circular hole.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811639460.3A CN109741314A (en) | 2018-12-29 | 2018-12-29 | A kind of visible detection method and system of part |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811639460.3A CN109741314A (en) | 2018-12-29 | 2018-12-29 | A kind of visible detection method and system of part |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109741314A true CN109741314A (en) | 2019-05-10 |
Family
ID=66362436
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811639460.3A Pending CN109741314A (en) | 2018-12-29 | 2018-12-29 | A kind of visible detection method and system of part |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109741314A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110969357A (en) * | 2019-12-03 | 2020-04-07 | 西安交通大学 | Visual detection method for holes of aluminum alloy machined part |
| CN111531203A (en) * | 2020-05-12 | 2020-08-14 | 西北工业大学 | Online visual detection method and system for depth of aircraft panel countersink |
| CN112084964A (en) * | 2020-09-11 | 2020-12-15 | 浙江水晶光电科技股份有限公司 | Product identification apparatus, method and storage medium |
| CN112232222A (en) * | 2020-10-19 | 2021-01-15 | 哈尔滨市科佳通用机电股份有限公司 | Bullet train axle box end cover bolt loss fault detection method based on image processing |
| CN112734689A (en) * | 2020-12-16 | 2021-04-30 | 广州大学 | Gasket quality detection method, system, device and storage medium |
| CN113049600A (en) * | 2021-03-31 | 2021-06-29 | 上汽通用五菱汽车股份有限公司 | Engine part wrong and neglected loading detection method and system based on visual detection |
| CN113269767A (en) * | 2021-06-07 | 2021-08-17 | 中电科机器人有限公司 | Batch part feature detection method, system, medium and equipment based on machine vision |
| CN113920146A (en) * | 2021-12-10 | 2022-01-11 | 杭州安脉盛智能技术有限公司 | Complex circle positioning method based on gray level and edge information fusion |
| CN114663427A (en) * | 2022-04-25 | 2022-06-24 | 北京与子成科技有限公司 | Boiler part size detection method based on image processing |
| CN114764790A (en) * | 2022-03-31 | 2022-07-19 | 南通市立新机械制造有限公司 | Gear broken tooth detection method based on Hough circle detection |
| CN114998311A (en) * | 2022-07-11 | 2022-09-02 | 江苏名欧精密机械有限公司 | Part precision detection method based on homomorphic filtering |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060210116A1 (en) * | 2005-03-18 | 2006-09-21 | Honda Elesys Co., Ltd. | Lane recognition apparatus |
| WO2013151266A1 (en) * | 2012-04-04 | 2013-10-10 | Movon Corporation | Method and system for lane departure warning based on image recognition |
| CN103793712A (en) * | 2014-02-19 | 2014-05-14 | 华中科技大学 | Image recognition method and system based on edge geometric features |
| CN104613869A (en) * | 2015-01-28 | 2015-05-13 | 南京信息工程大学 | Method and system for detecting elliptical hole group based on maximum inscribed circle |
| CN105139386A (en) * | 2015-08-12 | 2015-12-09 | 南京航空航天大学 | Image processing method for quickly and automatically detecting soldered dot unsatisfied products of electric connectors |
| CN106680287A (en) * | 2016-12-28 | 2017-05-17 | 无锡浩远视觉科技有限公司 | Visual inspection method for step defects of bearing rivets |
| CN107818583A (en) * | 2017-11-07 | 2018-03-20 | 歌尔股份有限公司 | Cross searching detection method and device |
| CN109003258A (en) * | 2018-06-15 | 2018-12-14 | 广东工业大学 | A kind of high-precision sub-pix circular pieces measurement method |
| CN109063634A (en) * | 2018-07-29 | 2018-12-21 | 国网上海市电力公司 | Using the method for hough-circle transform identification target object region electrical symbol in power monitoring |
-
2018
- 2018-12-29 CN CN201811639460.3A patent/CN109741314A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060210116A1 (en) * | 2005-03-18 | 2006-09-21 | Honda Elesys Co., Ltd. | Lane recognition apparatus |
| WO2013151266A1 (en) * | 2012-04-04 | 2013-10-10 | Movon Corporation | Method and system for lane departure warning based on image recognition |
| CN103793712A (en) * | 2014-02-19 | 2014-05-14 | 华中科技大学 | Image recognition method and system based on edge geometric features |
| CN104613869A (en) * | 2015-01-28 | 2015-05-13 | 南京信息工程大学 | Method and system for detecting elliptical hole group based on maximum inscribed circle |
| CN105139386A (en) * | 2015-08-12 | 2015-12-09 | 南京航空航天大学 | Image processing method for quickly and automatically detecting soldered dot unsatisfied products of electric connectors |
| CN106680287A (en) * | 2016-12-28 | 2017-05-17 | 无锡浩远视觉科技有限公司 | Visual inspection method for step defects of bearing rivets |
| CN107818583A (en) * | 2017-11-07 | 2018-03-20 | 歌尔股份有限公司 | Cross searching detection method and device |
| CN109003258A (en) * | 2018-06-15 | 2018-12-14 | 广东工业大学 | A kind of high-precision sub-pix circular pieces measurement method |
| CN109063634A (en) * | 2018-07-29 | 2018-12-21 | 国网上海市电力公司 | Using the method for hough-circle transform identification target object region electrical symbol in power monitoring |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110969357A (en) * | 2019-12-03 | 2020-04-07 | 西安交通大学 | Visual detection method for holes of aluminum alloy machined part |
| CN111531203A (en) * | 2020-05-12 | 2020-08-14 | 西北工业大学 | Online visual detection method and system for depth of aircraft panel countersink |
| CN112084964A (en) * | 2020-09-11 | 2020-12-15 | 浙江水晶光电科技股份有限公司 | Product identification apparatus, method and storage medium |
| CN112232222A (en) * | 2020-10-19 | 2021-01-15 | 哈尔滨市科佳通用机电股份有限公司 | Bullet train axle box end cover bolt loss fault detection method based on image processing |
| CN112232222B (en) * | 2020-10-19 | 2021-05-07 | 哈尔滨市科佳通用机电股份有限公司 | Bullet train axle box end cover bolt loss fault detection method based on image processing |
| CN112734689A (en) * | 2020-12-16 | 2021-04-30 | 广州大学 | Gasket quality detection method, system, device and storage medium |
| CN113049600A (en) * | 2021-03-31 | 2021-06-29 | 上汽通用五菱汽车股份有限公司 | Engine part wrong and neglected loading detection method and system based on visual detection |
| CN113269767A (en) * | 2021-06-07 | 2021-08-17 | 中电科机器人有限公司 | Batch part feature detection method, system, medium and equipment based on machine vision |
| CN113920146A (en) * | 2021-12-10 | 2022-01-11 | 杭州安脉盛智能技术有限公司 | Complex circle positioning method based on gray level and edge information fusion |
| CN114764790A (en) * | 2022-03-31 | 2022-07-19 | 南通市立新机械制造有限公司 | Gear broken tooth detection method based on Hough circle detection |
| CN114663427A (en) * | 2022-04-25 | 2022-06-24 | 北京与子成科技有限公司 | Boiler part size detection method based on image processing |
| CN114663427B (en) * | 2022-04-25 | 2023-06-16 | 北京与子成科技有限公司 | Boiler part size detection method based on image processing |
| CN114998311A (en) * | 2022-07-11 | 2022-09-02 | 江苏名欧精密机械有限公司 | Part precision detection method based on homomorphic filtering |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109741314A (en) | A kind of visible detection method and system of part | |
| CN109816644B (en) | Bearing defect automatic detection system based on multi-angle light source image | |
| US8340420B2 (en) | Method for recognizing objects in images | |
| CN109870461B (en) | Electronic components quality detection system | |
| CN109978822B (en) | Banana maturity judging modeling method and judging method based on machine vision | |
| CN106920219B (en) | Article defect detection method, image processing system and computer readable recording medium | |
| CN110220917A (en) | A kind of crown plug surface defect online test method based on image procossing | |
| CN101995223B (en) | Chip appearance detection method and system | |
| CN104990925B (en) | One kind is based on gradient multi thresholds optimization defect inspection method | |
| CN107389701A (en) | A kind of PCB visual defects automatic checkout system and method based on image | |
| CN106529559A (en) | Pointer-type circular multi-dashboard real-time reading identification method | |
| CN104537651B (en) | Proportion detecting method and system for cracks in road surface image | |
| CN104809433B (en) | A kind of zebra line detecting method based on maximum stable region and stochastical sampling | |
| CN111382704A (en) | Vehicle line-pressing violation judgment method and device based on deep learning and storage medium | |
| CN104008550A (en) | Wafer surface defect feature analysis method and system and wafer surface detect feature classification method and system | |
| CN103971126A (en) | Method and device for identifying traffic signs | |
| CN103034838A (en) | Special vehicle instrument type identification and calibration method based on image characteristics | |
| CN106407924A (en) | Binocular road identifying and detecting method based on pavement characteristics | |
| CN106251333B (en) | Element reverse detection method and system | |
| CN110047063B (en) | Material drop detection method, device, equipment and storage medium | |
| CN110189375A (en) | A kind of images steganalysis method based on monocular vision measurement | |
| CN111695373A (en) | Zebra crossing positioning method, system, medium and device | |
| CN111476804A (en) | Method, device and equipment for efficiently segmenting carrier roller image and storage medium | |
| CN109387524A (en) | Thread defect detection method and device based on linearly polarized photon | |
| CN117237747B (en) | Hardware defect classification and identification method based on artificial intelligence |
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: 20190510 |