CN112598632A - Appearance detection method and device for contact element of crimp connector - Google Patents
Appearance detection method and device for contact element of crimp connector Download PDFInfo
- Publication number
- CN112598632A CN112598632A CN202011492715.5A CN202011492715A CN112598632A CN 112598632 A CN112598632 A CN 112598632A CN 202011492715 A CN202011492715 A CN 202011492715A CN 112598632 A CN112598632 A CN 112598632A
- Authority
- CN
- China
- Prior art keywords
- image
- contact
- module
- feature map
- target image
- 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 abstract description 29
- 238000000034 method Methods 0.000 claims abstract description 38
- 238000007781 pre-processing Methods 0.000 claims abstract description 18
- 238000012216 screening Methods 0.000 claims abstract description 13
- 238000005259 measurement Methods 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 18
- 238000001914 filtration Methods 0.000 claims description 15
- 238000007373 indentation Methods 0.000 claims description 15
- 238000010586 diagram Methods 0.000 claims description 14
- 230000006835 compression Effects 0.000 claims description 10
- 238000007906 compression Methods 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 6
- 238000007689 inspection Methods 0.000 claims description 4
- 238000012797 qualification Methods 0.000 abstract description 8
- 241000282414 Homo sapiens Species 0.000 abstract description 7
- 238000002788 crimping Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 8
- 238000013135 deep learning Methods 0.000 description 5
- 238000004519 manufacturing process Methods 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 4
- 238000000605 extraction Methods 0.000 description 4
- 238000006243 chemical reaction Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 210000004556 brain Anatomy 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 238000011179 visual inspection Methods 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
- G06T7/001—Industrial image inspection using an image reference approach
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- 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
- G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Geometry (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method and a device for detecting the appearance of a contact element of a crimp connector, wherein the method comprises the following steps: preprocessing an input contact member image to obtain a target image; identifying a region to be detected in the target image; screening the characteristic map in the area to be detected; and comparing the feature graph with a corresponding template threshold value to determine whether the feature graph is qualified. The appearance detection method for the contact element of the crimp connector disclosed by the invention can improve the qualification judgment efficiency of the contact element and can save a large amount of human resources.
Description
Technical Field
The invention belongs to the technical field of computer vision detection, and particularly relates to a method and a device for detecting the appearance of a contact element of a crimp connector.
Background
The installation of crimp type connectors is now predominant in cable network products.
Generally, the number of crimp terminals required for one model is enormous. At present, however, the qualification judgment of the crimping contact element of the crimping connector basically depends on manual visual inspection, and careless omission and errors are avoided in the process; in addition, manual inspection requires a break time, which reduces the efficiency of cable network product production both in terms of quality of inspection and the time required. The contact used in the conventional press-fit type electrical connector needs to be mounted in the connector after a press-fit process is performed between the contact and a wire. The crimping quality of the contact directly determines the performance of the crimped connector after being butted, and after an unqualified crimping piece is installed in the electric connector, risks of unqualified electrical performance, generation of redundant materials and the like can be caused in the subsequent butting use process and the like, and serious threats can be formed on related equipment.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the traditional appearance detection mode of the contact of the crimp connector is low in efficiency and occupies a large amount of human resources.
In order to solve the technical problem, the invention discloses an appearance detection method for a contact of a crimp connector, wherein the method comprises the following steps:
preprocessing an input contact member image to obtain a target image;
identifying a region to be detected in the target image;
screening the characteristic map in the area to be detected;
and comparing the feature graph with a corresponding template threshold value to determine whether the feature graph is qualified.
Optionally, the step of preprocessing the input contact image to obtain a target image includes:
carrying out mean value filtering processing on an input contact member image to obtain a first image;
and carrying out gray level processing on the first image to obtain a target image.
Optionally, the step of identifying the region to be detected in the target image includes:
dividing the target image into grid images according to a preset proportion;
respectively predicting a plurality of bounding boxes in each prediction unit by taking grids as the prediction units;
for each of the bounding boxes, a confidence is calculated for the contained object bounding box, and a likelihood that the contained object belongs to a preset category is calculated.
Optionally, the step of comparing the feature map with a corresponding template threshold to determine whether the feature map is qualified includes:
under the condition that the feature map is the feature map in the observation hole, adopting an image histogram to represent the pixel distribution of the feature map;
determining the matching degree of the image histogram and a corresponding template threshold value;
determining that a wire core exists in the observation hole under the condition that the matching degree is higher than a preset matching degree;
determining a pixel value of each measurement ratio according to the resolution magnification factor of the contact image and a preset contact reference size;
determining the indentation position and the wire core length value according to the pixel value of each measurement ratio;
and judging whether the characteristic diagram is qualified or not according to the indentation position and the length value of the wire core.
In order to solve the above technical problem, the present invention also discloses an appearance detection device for a contact of a crimp connector, comprising:
the preprocessing module is used for preprocessing an input contact image to obtain a target image;
the identification module is used for identifying the area to be detected in the target image;
the screening module is used for screening the characteristic diagram in the area to be detected;
and the comparison module is used for comparing the feature map with the corresponding template threshold value so as to determine whether the feature map is qualified.
Optionally, the preprocessing module comprises:
the first submodule is used for carrying out mean value filtering processing on an input contact piece image to obtain a first image;
and the second sub-module is used for carrying out gray processing on the first image to obtain a target image.
Optionally, the identification module comprises:
the third sub-module is used for dividing the target image into grid images according to a preset proportion;
a fourth sub-module, configured to respectively predict a plurality of bounding boxes in each prediction unit by using a mesh as a prediction unit;
and the fifth submodule is used for calculating the confidence of the contained object bounding box and calculating the possibility that the contained target belongs to a preset class for each bounding box.
Optionally, the alignment module comprises:
the sixth sub-module is used for representing the pixel distribution of the feature map by adopting an image histogram under the condition that the feature map is the feature map in the observation hole;
the seventh sub-module is used for determining the matching degree of the image histogram and a corresponding template threshold value;
the eighth submodule is used for determining that a wire core exists in the observation hole under the condition that the matching degree is higher than the preset matching degree;
the ninth submodule is used for determining the pixel value of each measurement ratio according to the resolution magnification of the contact image and the preset contact reference size;
the tenth submodule is used for determining the indentation position and the wire core length value according to the pixel value of each measurement ratio;
and the eleventh submodule is used for judging whether the characteristic diagram is qualified or not according to the indentation position and the length value of the wire core.
The invention has the following advantages:
the embodiment of the invention discloses a method and a device for detecting the appearance of a contact element of a crimp connector, wherein an input contact element image is preprocessed to obtain a target image; identifying a region to be detected in a target image; screening a characteristic map in a region to be detected; the characteristic graph is compared with the corresponding template threshold value to determine whether the characteristic graph is qualified or not, the scheme can carry out non-contact judgment on the qualification of the crimping contact element of the electric connector, the efficiency of judging the qualification of the contact element can be improved by a computer vision method, and meanwhile, a large amount of manpower resources can be saved.
Drawings
Fig. 1 is a flowchart illustrating steps of a method for detecting appearance of a contact of a compression connector according to an embodiment of the present invention;
FIG. 2 is a schematic view of the external appearance of a contact of a crimp connector;
FIG. 3 is a schematic diagram of a mean filtering principle;
FIG. 4 is a schematic diagram of a process for training characteristics of a crimp contact through a convolutional neural network;
fig. 5 is a schematic structural diagram of an appearance inspection device for a contact of a compression connector according to an embodiment of the invention.
Detailed Description
The present invention will be described in detail below with reference to specific embodiments and with reference to the attached drawings. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The computer vision detection control technology is the key of the technical problem of environment perception and autonomous control of intelligent manufacturing equipment. The computer vision technology aims at providing an intelligent measurement and control technology for detection, measurement, analysis, judgment and decision control for human eyes, human brains and human hands, is an important means for realizing automatic measurement and control by simulating the self vision perception capability of human beings compared with other detection and control technologies, has the advantages of non-contact, can greatly ensure that the product properties are not changed, and is very suitable for a system which has strict requirements on the product quality in the aerospace industry.
The use of computer vision technology for object detection in industrial production is a well established field of application. The intelligent assistance even replaces the manual judgment of the state of the assembly part, so that the processing and manufacturing efficiency can be improved to a great extent, and the problem of high cost caused by manual low efficiency is solved.
Industrial target detection comes from the target detection branch direction of computer vision technology. The target detection task is to use a rectangular frame to define the area where the target appears in the image and identify the target category, including two parts of positioning and classification. The target detection method is to use a sliding window mode to realize the region-by-region detection of the image, and the feature extraction mainly depends on manual design. With the rapid development of the deep learning technology, the target detection greatly improves the detection precision and accuracy rate under the support of the deep learning.
When computer vision technology is selected to detect and judge the appearance of the contact of the crimp connector, the current industrial target detection method cannot be directly applied. Direct application will have the following problems:
the first problem is that: uncertainty of lighting condition
The second problem is that: uncertainty of angle of measured pressure connecting piece
The third problem is that: crimp contact feature extraction
The fourth problem is that: problem of size measurement
Aiming at the characteristics of the appearance of the contact piece of the crimping type connector, the following technical scheme is adopted in the design to solve the problems.
(1) Uncertainty of lighting condition
For the problem, because the number of interference light sources in the production environment is large, the problem of inconsistent illumination brightness cannot be avoided in the production field of a cable network workshop, and the image contains noise interference, the subsequent process can be ensured only by pre-processing the image. The method for processing the image comprises the steps of preprocessing an input image by adopting a mean filtering method. The mean filtering is a spatial domain filtering method, namely, pixels of an image are directly processed without calculation of methods such as Fourier change and the like. The method can well smooth some pixel points with large pixel value fluctuation to be similar to surrounding pixel information, and achieves a filtering effect. Further, the input image is converted into a single-channel gray-scale image by converting the number of channels to an image containing RGB (color of three channels, red, green, and blue) information. Through the conversion of the gray information, the influence of the to-be-processed image of the light source can be greatly reduced.
(2) Uncertainty of angle of measured pressure connecting piece
Since there is a small change in the angle of view when the crimp contact to be tested is photographed, although this change does not greatly interfere with the judgment of the human eye, the characteristics have changed significantly with respect to machine vision. The method and the device aim to position the image processing area in a deep learning mode. The traditional mode identification needs to detect the positioning of the crimping connector in a sliding mode on a window with a fixed size on an image, and when the window slides to a target area, the characteristics are extracted and compared, so that a detection result is obtained. The conventional method is computationally intensive, and for the size of the window, it is necessary to assume that the features such as the width and height of the target are not changed, which adds much unnecessary computation cost and relatively low accuracy for the detection of the contact. By carrying out feature training on the neural network, the image to be detected can be accurately searched and positioned to the detected area after being preprocessed, and the operation cost is reduced. The application is mainly directed to the crimp connector contact shown in fig. 2, and after training of a neural network, corresponding regions of three features in the graph can be extracted from an input image (where 0< a <1 mm).
(3) Crimp contact feature extraction
Combining with the output result of the neural network, respectively detecting and judging the characteristic regions which need to be detected and are screened in the application; the indentation position and the size of a in fig. 2 belong to the same detection problem, and will be described in detail in (4). For the problem whether the wire core in the observation hole is visible or not, the image segment after feature screening is further segmented, then the image segment is matched with a preset template according to the space histogram feature in the observation hole, and when the similarity exceeds the preset matching degree, the image segment can be judged to be qualified. Therefore, the method and the device select the edge feature extraction algorithm of the crimping contact piece in a targeted manner, accurately segment the image and input the segmented image into the histogram comparison calculation module.
(4) Problem of size measurement
For the qualified judgment problems of the indentation position of the crimping contact element and the wire core length, the calibration and the resolution of a camera need to be tested on the basis of a hardware shooting platform. Through the confirmation of hardware parameters, the size in the picture can be converted with the real size in a fixed ratio, the mode is also the main mode of the size measurement of the two-dimensional picture at present, and under the condition that the positioning conversion is accurate, the error can be controlled within 0.1 mm.
Fig. 1 is a flowchart illustrating steps of a method for detecting an appearance of a contact of a compression connector according to an embodiment of the present invention.
The appearance detection method for the contact piece of the compression connector provided by the embodiment of the invention comprises the following steps:
step 101: and preprocessing the input contact member image to obtain a target image.
In an alternative embodiment, the input contact image is preprocessed to obtain the target image by: firstly, carrying out mean value filtering processing on an input contact member image to obtain a first image; and secondly, carrying out gray level processing on the first image to obtain a target image. The contact image is the image of the crimp connector contact to be tested.
In the actual implementation process, the contact image, i.e., the image to be detected, can be read into the processing host in a matrix storage manner, so that the subsequent module can traverse to each pixel point in the image. And then, smoothing the image to be detected by using a mean filtering method. The mean filtering can filter out some unnecessary noise from the high-frequency noise input into the image to be detected. In the embodiment of the application, a basic implementation principle is that a window with a fixed size is used for sliding on an image to be measured, and the value of an original central pixel is replaced by the mean value of pixels in the window. Through analysis of the sample image, in the embodiment of the present application, a window with a size of 5 × 5 is selected for performing the traversal operation, as shown in the schematic diagram of the principle of mean value filtering in fig. 3, so that the edge characteristics of the crimp contact can be retained to the maximum extent.
And obtaining an image after the average filtering, and carrying out gray processing on the image. The number of channels of the gray-scale image, namely the image, is changed into 1, and the dynamic value range of each pixel point is in the value range of 0-255. The conversion can simplify the texture information in the picture, and the contact feature to be extracted is irrelevant to the texture information when the appearance of the contact of the compression connector is detected, so the later operation cost can be reduced after the simplification.
Step 102: and identifying the region to be detected in the target image.
One way to optionally identify the region to be detected in the target image is as follows:
firstly, dividing a target image into grid images according to a preset proportion; secondly, respectively predicting a plurality of bounding boxes in each prediction unit by taking the grids as the prediction units; finally, for each bounding box, the confidence of the contained object bounding box is calculated, and the likelihood that the contained object belongs to a preset category is calculated.
In the practical implementation process, after the image is preprocessed, the image needs to be positioned in the area where the contact needs to be detected, and the related algorithm in deep learning is cited in the embodiment of the application to realize the positioning of the detection area. Early object detection algorithms based on deep learning employed a selective search to narrow the number of bounding boxes that had to be tested (bounding boxes in the present embodiment refer to a rectangle that frames an object on a picture after a suspected identified object is predicted). The method selected in the embodiment of the application is to perform one-time forward processing on the image by using a neural network. As shown in fig. 4, the specific process is to first divide the original image into one picture with 13 × 13 meshes by scale average. The 169 cells are changed according to the size of the original image. For example, for a 416x416 pixel picture, the size of each picture element is 32x32 pixels. When a picture is processed, a plurality of bounding boxes in a unit are predicted in a unit of picture unit. For each bounding box, the network calculates the confidence of the bounding box of the contained object, and also calculates the likelihood that the contained object belongs to a particular class. Non-maximal suppression may eliminate low confidence bounding boxes and multiple high confidence bounding boxes that simultaneously surround a single object to the point where only one remains.
Step 103: and screening the characteristic map in the area to be detected.
And further segmenting the region to be detected to determine a characteristic map. The method adopted in the embodiment of the application belongs to a region growing method, and the characteristic region of the compression joint contact, namely the edge of the region to be detected is kept obvious through the positioning of the steps, and specifically comprises the following steps: and (3) realizing image segmentation by the uniformity of pixel characteristics in the observation holes of the compression contact, and taking the observation hole part with the largest area in the contact characteristic diagram as a detection area by using an 8-neighborhood growth method and a threshold termination rule. And setting a random seed pixel point of the detection area as X, and if X belongs to the T and the neighborhood image block is intersected with the multi-class, then X belongs to Am. The specific expression is shown as the following formula:
wherein x is a random seed pixel point, y is a pixel point in the region to be detected, Am is a set to be detected, and g is a gray value.
Step 104: and comparing the feature map with the corresponding template threshold value to determine whether the feature map is qualified.
In an optional embodiment, the manner of comparing the feature map with the corresponding template threshold to determine whether the feature map is qualified comprises the following sub-steps:
the first substep: under the condition that the feature map is the feature map in the observation hole, adopting an image histogram to represent the pixel distribution of the feature map; secondly, determining the matching degree of the image histogram and a corresponding template threshold;
in the second step: determining that a wire core exists in the observation hole under the condition that the matching degree is higher than the preset matching degree;
and a third substep: determining the pixel value of each measurement ratio according to the resolution magnification of the contact image and the preset contact reference size;
and a fourth substep: determining the indentation position and the wire core length value according to the pixel value of each measurement ratio;
and a fifth substep: and judging whether the characteristic diagram is qualified or not according to the indentation position and the wire core length value.
In the actual implementation process, the segmented characteristic diagram in the observation hole reflects the pixel subsection statistics by using an image histogram, and is compared with a preset template threshold, and when the similarity reaches a preset matching degree such as more than 80%, the existence of a wire core in the observation hole of the contact element can be considered. In addition, the length of the indentation and the wire core needs to be obtained by converting the pixel value of each measurement ratio by taking the resolution magnification of an image input by hardware equipment as a standard and taking a contact reference measured in advance as a standard, and judging whether the value meets the specification or not.
The embodiment of the application provides a method and a device for detecting the appearance of a contact element of a crimp connector, and the method and the device are used for preprocessing an input contact element image to obtain a target image; identifying a region to be detected in a target image; screening a characteristic map in a region to be detected; the characteristic graph is compared with the corresponding template threshold value to determine whether the characteristic graph is qualified or not, the scheme can carry out non-contact judgment on the qualification of the crimping contact element of the electric connector, the efficiency of judging the qualification of the contact element can be improved by a computer vision method, and meanwhile, a large amount of manpower resources can be saved.
Fig. 5 is a flowchart illustrating steps of a method for detecting an appearance of a contact of a compression connector according to an embodiment of the present invention.
The appearance detection device for the contact piece of the compression connector provided by the embodiment of the invention comprises the following modules:
the preprocessing module 501 is configured to preprocess an input contact image to obtain a target image;
an identifying module 502, configured to identify a region to be detected in the target image;
a screening module 503, configured to screen a feature map in the area to be detected;
a comparison module 504, configured to compare the feature map with a corresponding template threshold to determine whether the feature map is qualified.
Optionally, the preprocessing module comprises:
the first submodule is used for carrying out mean value filtering processing on an input contact piece image to obtain a first image;
and the second sub-module is used for carrying out gray processing on the first image to obtain a target image.
Optionally, the identification module comprises:
the third sub-module is used for dividing the target image into grid images according to a preset proportion;
a fourth sub-module, configured to respectively predict a plurality of bounding boxes in each prediction unit by using a mesh as a prediction unit;
and the fifth submodule is used for calculating the confidence of the contained object bounding box and calculating the possibility that the contained target belongs to a preset class for each bounding box.
Optionally, the alignment module comprises:
the sixth sub-module is used for representing the pixel distribution of the feature map by adopting an image histogram under the condition that the feature map is the feature map in the observation hole;
the seventh sub-module is used for determining the matching degree of the image histogram and a corresponding template threshold value;
the eighth submodule is used for determining that a wire core exists in the observation hole under the condition that the matching degree is higher than the preset matching degree;
the ninth submodule is used for determining the pixel value of each measurement ratio according to the resolution magnification of the contact image and the preset contact reference size;
the tenth submodule is used for determining the indentation position and the wire core length value according to the pixel value of each measurement ratio;
and the eleventh submodule is used for judging whether the characteristic diagram is qualified or not according to the indentation position and the length value of the wire core.
The embodiment of the invention discloses an appearance detection device for a contact element of a crimp connector, which is used for preprocessing an input contact element image to obtain a target image; identifying a region to be detected in a target image; screening a characteristic map in a region to be detected; the device can judge the qualification of the crimping contact element of the electric connector in a non-contact way, can improve the efficiency of judging the qualification of the contact element by a computer vision method, and can save a large amount of manpower resources at the same time.
It should be noted that the above description is only a preferred embodiment of the present invention, and it should be understood that various changes and modifications can be made by those skilled in the art without departing from the technical idea of the present invention, and these changes and modifications are included in the protection scope of the present invention.
Those skilled in the art will appreciate that the details of the invention not described in detail in this specification are well within the skill of those in the art.
Claims (8)
1. A method for detecting the appearance of a contact of a compression connector is characterized by comprising the following steps:
preprocessing an input contact member image to obtain a target image;
identifying a region to be detected in the target image;
screening the characteristic map in the area to be detected;
and comparing the feature graph with a corresponding template threshold value to determine whether the feature graph is qualified.
2. The method of claim 1, wherein the step of preprocessing the input contact image to obtain the target image comprises:
carrying out mean value filtering processing on an input contact member image to obtain a first image;
and carrying out gray level processing on the first image to obtain a target image.
3. The method of claim 1, wherein the step of identifying the region to be detected in the target image comprises:
dividing the target image into grid images according to a preset proportion;
respectively predicting a plurality of bounding boxes in each prediction unit by taking grids as the prediction units;
for each of the bounding boxes, a confidence is calculated for the contained object bounding box, and a likelihood that the contained object belongs to a preset category is calculated.
4. The method of claim 1, wherein the step of comparing the feature map to a corresponding template threshold to determine whether the feature map is qualified comprises:
under the condition that the feature map is the feature map in the observation hole, adopting an image histogram to represent the pixel distribution of the feature map;
determining the matching degree of the image histogram and a corresponding template threshold value;
determining that a wire core exists in the observation hole under the condition that the matching degree is higher than a preset matching degree;
determining a pixel value of each measurement ratio according to the resolution magnification factor of the contact image and a preset contact reference size;
determining the indentation position and the wire core length value according to the pixel value of each measurement ratio;
and judging whether the characteristic diagram is qualified or not according to the indentation position and the length value of the wire core.
5. An appearance inspection device for a contact of a crimp connector, comprising:
the preprocessing module is used for preprocessing an input contact image to obtain a target image;
the identification module is used for identifying the area to be detected in the target image;
the screening module is used for screening the characteristic diagram in the area to be detected;
and the comparison module is used for comparing the feature map with the corresponding template threshold value so as to determine whether the feature map is qualified.
6. The apparatus of claim 5, wherein the pre-processing module comprises:
the first submodule is used for carrying out mean value filtering processing on an input contact piece image to obtain a first image;
and the second sub-module is used for carrying out gray processing on the first image to obtain a target image.
7. The apparatus of claim 5, wherein the identification module comprises:
the third sub-module is used for dividing the target image into grid images according to a preset proportion;
a fourth sub-module, configured to respectively predict a plurality of bounding boxes in each prediction unit by using a mesh as a prediction unit;
and the fifth submodule is used for calculating the confidence of the contained object bounding box and calculating the possibility that the contained target belongs to a preset class for each bounding box.
8. The apparatus of claim 5, wherein the alignment module comprises:
the sixth sub-module is used for representing the pixel distribution of the feature map by adopting an image histogram under the condition that the feature map is the feature map in the observation hole;
the seventh sub-module is used for determining the matching degree of the image histogram and a corresponding template threshold value;
the eighth submodule is used for determining that a wire core exists in the observation hole under the condition that the matching degree is higher than the preset matching degree;
the ninth submodule is used for determining the pixel value of each measurement ratio according to the resolution magnification of the contact image and the preset contact reference size;
the tenth submodule is used for determining the indentation position and the wire core length value according to the pixel value of each measurement ratio;
and the eleventh submodule is used for judging whether the characteristic diagram is qualified or not according to the indentation position and the length value of the wire core.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011492715.5A CN112598632A (en) | 2020-12-16 | 2020-12-16 | Appearance detection method and device for contact element of crimp connector |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011492715.5A CN112598632A (en) | 2020-12-16 | 2020-12-16 | Appearance detection method and device for contact element of crimp connector |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112598632A true CN112598632A (en) | 2021-04-02 |
Family
ID=75196636
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011492715.5A Pending CN112598632A (en) | 2020-12-16 | 2020-12-16 | Appearance detection method and device for contact element of crimp connector |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112598632A (en) |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103106663A (en) * | 2013-02-19 | 2013-05-15 | 公安部第三研究所 | Method for detecting defect of subscriber identity module (SIM) card based on image processing in computer system |
| US20130195365A1 (en) * | 2012-02-01 | 2013-08-01 | Sharp Laboratories Of America, Inc. | Edge based template matching |
| CN104050745A (en) * | 2013-03-13 | 2014-09-17 | 苏州日宝科技有限责任公司 | High-speed coin sorting technology based on image identification |
| CN106291494A (en) * | 2016-07-21 | 2017-01-04 | 深圳大学 | The SAR cheating interference target identification method and system strengthened based on differential characteristics |
| CN106600600A (en) * | 2016-12-26 | 2017-04-26 | 华南理工大学 | Wafer defect detection method based on characteristic matching |
| CN109429473A (en) * | 2017-08-28 | 2019-03-05 | 株洲中车时代电气股份有限公司 | Automatic check method and device with polarity electronic component in circuit board |
| CN109948501A (en) * | 2019-03-13 | 2019-06-28 | 东华大学 | The detection method of personnel and safety cap in a kind of monitor video |
| CN109974582A (en) * | 2019-04-04 | 2019-07-05 | 江南大学 | A kind of the conductor diameters non-contact vision detection device and method of automotive wire bundle |
| CN110321853A (en) * | 2019-07-05 | 2019-10-11 | 杭州巨骐信息科技股份有限公司 | Distribution cable external force damage prevention system based on video intelligent detection |
| CN110793966A (en) * | 2019-11-07 | 2020-02-14 | 哈尔滨工业大学 | Automatic detection device and detection method for aviation electric connector contact piece shrinkage needle |
| CN111242123A (en) * | 2020-01-07 | 2020-06-05 | 西安交通大学 | Power equipment fault diagnosis method based on infrared image |
| CN111665267A (en) * | 2020-06-16 | 2020-09-15 | 沈阳兴华航空电器有限责任公司 | Visual detection method for compression joint quality of pit-pressing type contact |
| CN111768378A (en) * | 2020-06-29 | 2020-10-13 | 广东电网有限责任公司 | Crimping hardware tool live-line detection method and device |
| CN111783524A (en) * | 2020-05-19 | 2020-10-16 | 普联国际有限公司 | Scene change detection method and device, storage medium and terminal equipment |
-
2020
- 2020-12-16 CN CN202011492715.5A patent/CN112598632A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130195365A1 (en) * | 2012-02-01 | 2013-08-01 | Sharp Laboratories Of America, Inc. | Edge based template matching |
| CN103106663A (en) * | 2013-02-19 | 2013-05-15 | 公安部第三研究所 | Method for detecting defect of subscriber identity module (SIM) card based on image processing in computer system |
| CN104050745A (en) * | 2013-03-13 | 2014-09-17 | 苏州日宝科技有限责任公司 | High-speed coin sorting technology based on image identification |
| CN106291494A (en) * | 2016-07-21 | 2017-01-04 | 深圳大学 | The SAR cheating interference target identification method and system strengthened based on differential characteristics |
| CN106600600A (en) * | 2016-12-26 | 2017-04-26 | 华南理工大学 | Wafer defect detection method based on characteristic matching |
| CN109429473A (en) * | 2017-08-28 | 2019-03-05 | 株洲中车时代电气股份有限公司 | Automatic check method and device with polarity electronic component in circuit board |
| CN109948501A (en) * | 2019-03-13 | 2019-06-28 | 东华大学 | The detection method of personnel and safety cap in a kind of monitor video |
| CN109974582A (en) * | 2019-04-04 | 2019-07-05 | 江南大学 | A kind of the conductor diameters non-contact vision detection device and method of automotive wire bundle |
| CN110321853A (en) * | 2019-07-05 | 2019-10-11 | 杭州巨骐信息科技股份有限公司 | Distribution cable external force damage prevention system based on video intelligent detection |
| CN110793966A (en) * | 2019-11-07 | 2020-02-14 | 哈尔滨工业大学 | Automatic detection device and detection method for aviation electric connector contact piece shrinkage needle |
| CN111242123A (en) * | 2020-01-07 | 2020-06-05 | 西安交通大学 | Power equipment fault diagnosis method based on infrared image |
| CN111783524A (en) * | 2020-05-19 | 2020-10-16 | 普联国际有限公司 | Scene change detection method and device, storage medium and terminal equipment |
| CN111665267A (en) * | 2020-06-16 | 2020-09-15 | 沈阳兴华航空电器有限责任公司 | Visual detection method for compression joint quality of pit-pressing type contact |
| CN111768378A (en) * | 2020-06-29 | 2020-10-13 | 广东电网有限责任公司 | Crimping hardware tool live-line detection method and device |
Non-Patent Citations (4)
| Title |
|---|
| STEFANO GHIDONI 等: "Self-learning visual inspection system for cable crimping machines", 《2011 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION》, 18 August 2011 (2011-08-18), pages 1 - 4 * |
| 卢山: "EWIS的端接技术", 《飞机设计》, vol. 40, no. 1, 29 February 2020 (2020-02-29), pages 66 - 69 * |
| 李天建;: "基于机器人视觉的汽车零配件表面缺陷检测算法研究与实现", 《沈阳大学学报(自然科学版)》, vol. 25, no. 06, 15 December 2013 (2013-12-15), pages 476 - 480 * |
| 李瑞: "压接质量及导线的X射线数字成像缺陷检测", 《中国优秀硕士学位论文全文数据库基础科学辑》, 15 April 2019 (2019-04-15), pages 005 - 71 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110544258B (en) | Image segmentation method and device, electronic equipment and storage medium | |
| WO2021000524A1 (en) | Hole protection cap detection method and apparatus, computer device and storage medium | |
| CN107944396B (en) | Knife switch state identification method based on improved deep learning | |
| CN106875373B (en) | Mobile phone screen MURA defect detection method based on convolutional neural network pruning algorithm | |
| CN113808138B (en) | Artificial intelligence-based wire and cable surface defect detection method | |
| CN111784633B (en) | Insulator defect automatic detection algorithm for electric power inspection video | |
| CN109671058B (en) | Defect detection method and system for large-resolution image | |
| CN111612737B (en) | Artificial board surface flaw detection device and detection method | |
| CN111242026B (en) | Remote sensing image target detection method based on spatial hierarchy perception module and metric learning | |
| CN116721107B (en) | Intelligent monitoring system for cable production quality | |
| JP2021515885A (en) | Methods, devices, systems and programs for setting lighting conditions and storage media | |
| CN115184359A (en) | Surface defect detection system and method capable of automatically adjusting parameters | |
| Huang et al. | Obstacle distance measurement under varying illumination conditions based on monocular vision using a cable inspection robot | |
| CN116228780B (en) | Silicon wafer defect detection method and system based on computer vision | |
| CN115917595A (en) | Surface defect detection system | |
| CN116071315A (en) | Product visual defect detection method and system based on machine vision | |
| CN111598913A (en) | Image segmentation method and system based on robot vision | |
| CN115620061A (en) | Hardware part defect detection method and system based on image recognition technology | |
| CN112862744A (en) | Intelligent detection method for internal defects of capacitor based on ultrasonic image | |
| CN116543247A (en) | Data set manufacturing method and verification system based on photometric stereo surface reconstruction | |
| CN112926483A (en) | Standard cabinet state indicator lamp identification monitoring method, device and system | |
| CN112750113B (en) | Glass bottle defect detection method and device based on deep learning and linear detection | |
| CN116363136B (en) | On-line screening method and system for automatic production of motor vehicle parts | |
| CN114140373A (en) | Switch defect detection method based on LabVIEW deep learning | |
| CN114332069B (en) | Connector detection method and device based on machine vision |
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 |