CN108109137A - The Machine Vision Inspecting System and method of vehicle part - Google Patents
The Machine Vision Inspecting System and method of vehicle part Download PDFInfo
- Publication number
- CN108109137A CN108109137A CN201711329856.3A CN201711329856A CN108109137A CN 108109137 A CN108109137 A CN 108109137A CN 201711329856 A CN201711329856 A CN 201711329856A CN 108109137 A CN108109137 A CN 108109137A
- Authority
- CN
- China
- Prior art keywords
- image
- vehicle
- measured
- component
- vehicle part
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/48—Extraction of image or video features by mapping characteristic values of the pattern into a parameter space, e.g. Hough transformation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/752—Contour matching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Quality & Reliability (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of Machine Vision Inspecting System and method of vehicle part, which includes：Transport mechanism including plummer, for carrying vehicle to be measured, makes plummer at the uniform velocity advance along the production line on transmission mechanism, and sets unique identifier on each plummer；Image collection module is installed on the detection station of transport mechanism both sides, for sensing the image for when vehicle arrival is measured, obtaining the appearance of vehicle all parts to be measured respectively from multiple angles using multiple CCD cameras；Preprocessing module meets the image of default specification for pretreatment image according to the generation of vehicle different parts；Locating module, for utilizing vehicle part generating unit image to be measured in outline or template matches positioning image；Vision-based detection module for being compared image of component with its pre-set image and calculating similarity between the two, detects vehicle part according to similarity and whether there is open defect.The present invention improves detection efficiency and accuracy of detection.
Description
Technical field
The present invention relates to technical field of machine vision, a kind of Machine Vision Inspecting System more particularly to vehicle part and
Method.
Background technology
Machine vision is a branch of artificial intelligence.For popular, exactly human eye is replaced with machine to measure and sentence
It is disconnected.NI Vision Builder for Automated Inspection is will to be ingested mesh by machine vision product (that is, image-pickup device is divided to two kinds of CMOS and CCD)
Mark is converted into picture signal, sends dedicated image processing system to, the shape information of target subject is obtained, according to pixel distribution
With the information such as brightness, color, it is transformed into digitized signal；Picture system carries out various computings to extract target to these signals
Feature, and then control according to the result of differentiation the device action at scene.
Existing machine vision technique is used for the quality testing of product, particularly for automobile, the vehicle part of motorcycle
Or the detection of vehicle.However, when doing the detection of vehicle appearance, vehicle image is often obtained using static mode, outside a vehicle
Sight detect not only the time spend it is longer, meanwhile, influenced the detection efficiency of subsequent production line, if to keep detection effect
Rate, and influence whether accuracy of detection, therefore, how to ensure that a kind of detection efficiency is high and does not reduce the vehicle of original accuracy of detection again
The problem of Machine Vision Inspecting System of component is very urgent.
The content of the invention
In view of the foregoing deficiencies of prior art, it is an object of the invention to provide a kind of machine vision of vehicle part
Detecting system and method, for solve in the prior art Machine Vision Detection vehicle part quality when, how not influence to detect
The problem of detection efficiency is improved while precision again.
In order to achieve the above objects and other related objects, the application's in a first aspect, present invention provides a kind of vehicle part
Machine Vision Inspecting System, it is whether qualified for detecting the vehicle part, the system comprises：
Transport mechanism including plummer, for carrying vehicle to be measured, makes the plummer along the production on transmission mechanism
Line at the uniform velocity advances, and sets unique identifier on each plummer；
Image collection module is installed on the detection station of the transport mechanism both sides or/and top, to be measured for sensing
When being reached to vehicle, the appearance of vehicle all parts to be measured is obtained respectively from multiple angles using multiple CCD cameras
Image；
Preprocessing module meets the image of default specification for pre-processing described image according to the generation of vehicle different parts；
Locating module, for utilizing vehicle part generating unit to be measured in outline or template matches positioning described image
Image；
Vision-based detection module, for being compared the image of component with its pre-set image and calculating phase between the two
Like degree, the vehicle part is detected according to the similarity and whether there is open defect.
In the certain embodiments of first aspect, each CCD camera is correspondingly arranged in described image acquisition module
There are the sensor and the reader and antenna for identifying the plummer identifier that triggering CDD video cameras are taken pictures.
Each CCD camera, which is provided with, in the certain embodiments of first aspect, in described image acquisition module is used for
Compensate the lighting device for light luminance of taking pictures.
In the certain embodiments of first aspect, identification module is handled, for when certain component for detecting the vehicle
During appearance existing defects, processing is split according to described image threshold value, it is special the defects of to obtain region corresponding to different threshold values
Sign is characterized in using the defects of neural network algorithm identification component in the damage that jumps, dislocation, end face wound, crackle, few tooth, gas
It is one or more of in hole or trachoma.
In the certain embodiments of first aspect, management module, for according to the corresponding identifier of the plummer with
The all parts testing result of vehicle to be measured is associated binding, and classifies by the testing result, by underproof portion
Part according to its defect type is counted and reported respectively.
In the certain embodiments of first aspect, subsystem is sorted：For identifying each plummer using radio-frequency technique
Underproof vehicle to be measured is transferred to reprocessing region by upper vehicle to be measured according to the testing result difference of the vehicle.
The second aspect of the application, the present invention provides a kind of machine vision detection method of vehicle part, for detecting
Whether qualified state vehicle part, including：
When sensing vehicle arrival to be measured, each portion of vehicle to be measured is obtained respectively from multiple angles using multiple CCD cameras
The image of the appearance of part；
When sensing vehicle to be measured and reaching, identify the identifier of the vehicle to be measured, by the corresponding image of the component with
The identifier of vehicle associates one by one；
Pretreatment described image meets the image of default specification according to the generation of vehicle different parts；
Utilize vehicle part generating unit image to be measured in outline or template matches positioning described image；
The image of component with its pre-set image is compared and calculates similarity between the two, according to described similar
Degree detects the vehicle part and whether there is open defect.
It is described to be compared simultaneously with its pre-set image by by the image of component in the certain embodiments of second aspect
The step of calculating similarity between the two, including：
Detect whether the image of component is the single component of pattern；
When the vehicle part to be measured is the single component of pattern, using template matches, outline or color rgb value
Calculate the similarity between the image of component and its pre-set image；
When the vehicle part to be measured is not the single component of pattern, the portion is calculated using convolutional neural networks mode
Similarity between part image and its pre-set image.
In the certain embodiments of second aspect, including：When certain the component appearance existing defects for detecting the vehicle
When, processing is split according to described image threshold value, feature the defects of to obtain region corresponding to different threshold values, using nerve net
It is a kind of in the damage that jumps, dislocation, end face wound, crackle, few tooth, stomata or trachoma that network algorithm identifies that the defects of component is characterized in
It is or several.
In the certain embodiments of second aspect, vehicle to be measured on each plummer, root are identified using radio-frequency technique
Underproof vehicle to be measured is transferred to reprocessing region according to the testing result difference of the vehicle.
As described above, the Machine Vision Inspecting System and method of the vehicle part of the present invention, have the advantages that：
The present invention is detected vehicle part to be measured on production line successively, using hierarchical detection mode, first, determines to treat
Whether the appearance for surveying vehicle part is defective；Secondly, if zero defect, no longer detect；If on the contrary, defective, then into one
Step detection determines the type of the defect.Inherently reduce the workload of detection, improve detection efficiency, meanwhile, subdivision
Detection improves accuracy of detection.
Description of the drawings
Fig. 1 is shown as a kind of Machine Vision Inspecting System structure diagram of vehicle part provided by the invention；
Fig. 2 is shown as a kind of Machine Vision Inspecting System complete structure block diagram of vehicle part provided by the invention；
Fig. 3 is shown as a kind of machine vision detection method flow chart of vehicle part provided by the invention；
Fig. 4 is shown as the present invention and provides a kind of machine vision detection method step S5 flow charts of vehicle part；
Fig. 5 is shown as a kind of machine vision detection method entire flow figure of vehicle part provided by the invention.
Specific embodiment
Presently filed embodiment is illustrated by particular specific embodiment below, those skilled in the art can be by this explanation
Content disclosed by book understands other advantages and effect of the application easily.
In described below, refer to the attached drawing, attached drawing describes several embodiments of the application.It should be appreciated that also it can be used
Other embodiment, and can be carried out in the case of without departing substantially from spirit and scope of the present disclosure mechanical composition, structure, electrically with
And the operational detailed description changed below should not be considered limiting, and the scope of embodiments herein
Only terms used herein are limited by the claims for the patent announced merely to describing specific embodiment, and be not
It is intended to limitation the application.The term of space correlation, for example, " on ", " under ", "left", "right", " following ", " lower section ", " lower part ",
" top ", " top " etc. can use to illustrate an element shown in figure or feature and another element or spy in the text
The relation of sign.
Although first, second grade of term is used for describing various elements herein in some instances, these elements
It should not be limited by these terms.These terms are only used for distinguishing an element with another element.For example, first is pre-
If threshold value can be referred to as the second predetermined threshold value, and similarly, the second predetermined threshold value can be referred to as the first predetermined threshold value, and
The scope of various described embodiments is not departed from.First predetermined threshold value and predetermined threshold value are to describe a threshold value, still
Unless context otherwise explicitly points out, otherwise they are not same predetermined threshold values.Similar situation further includes first
Volume and the second volume.
Furthermore as used in herein, singulative " one ", "one" and "the" are intended to also include plural number shape
Formula, unless having opposite instruction in context it will be further understood that term "comprising", " comprising " show that there are the spies
Sign, step, operation, element, component, project, species, and/or group, but it is not excluded for other one or more features, step, behaviour
Work, element, component, project, the presence of species, and/or group, appearance or addition term "or" used herein and "and/or" quilt
Be construed to inclusive or mean any one or any combinations therefore, " A, B or C " or " A, B and/or C " mean " with
Descend any one：A；B；C；A and B；A and C；B and C；A, B and C " is only when element, function, step or the combination of operation are in some modes
Under it is inherently mutually exclusive when, be just present with the exception of this definition.
Referring to Fig. 1, the present invention provides a kind of Machine Vision Inspecting System structure diagram of vehicle part；For detecting
Whether qualified state vehicle part, the system comprises：
Transport mechanism 1 including plummer, for carrying vehicle to be measured, makes the plummer along the life on transmission mechanism
Producing line is at the uniform velocity advanced, and sets unique identifier on each plummer；
Wherein, each vehicle part to be measured is due to being placed on different plummers, it is equivalent to each Vehicle mounted station
Vehicle part have unique identifier, such as：Electronic tag or, sequence number etc., if selection sequence number, follow-up correspondence is adopted
With code reader etc.
Image collection module 2 is installed on the detection station of the transport mechanism both sides or/and top, is treated for sensing
When measuring vehicle arrival, the appearance of vehicle all parts to be measured is obtained respectively from multiple angles using multiple CCD cameras
Image；
Specifically, each CCD camera is correspondingly arranged on what triggering CDD video cameras were taken pictures in described image acquisition module
The reader and antenna of sensor and the identification plummer identifier；Each CCD camera in described image acquisition module
It is provided with to compensate the lighting device for light luminance of taking pictures.
Wherein, sensor can be photoelectric sensor or position sensor etc., when it is triggered, then represent the vehicle on plummer
Component has been introduced into region capable of taking pictures, can adjust the sensor mounting location according to the specific environment at scene so that
The image of vehicle part to be measured can have just been obtained during sensing.
CCD camera is industrial camera, the reader be used for read label information equipment, may be designed as hand-held or
It is fixed；The antenna transfers radiofrequency signal between label and reader, and the lighting device is light filling equipment, is taken the photograph for CCD
Light filling is carried out when camera is taken pictures.
Preprocessing module 3 meets the image of default specification for pre-processing described image according to the generation of vehicle different parts；
Specifically, the pretreatment includes filtering and noise reduction, contrast increases processing and marginalisation processing, to improve image
Clarity.
Locating module 4, for utilizing vehicle part generating unit to be measured in outline or template matches positioning described image
Part image；
Specifically, using outline or template matches, the parts to be tested in vehicle part image to be measured is positioned,
To obtain the feature of the critical component of local pattern, the image of the parts to be tested is got by Primary Location, so as to reduce the later stage
Calculation amount ensures the accuracy of image of component position, is detected or analyzes convenient for subsequent image, to improve detection efficiency.
Vision-based detection module 5, for being compared and calculating between the two with its pre-set image by the image of component
Similarity detects the vehicle part according to the similarity and whether there is open defect.
The calculation of similarity includes two kinds：
When the structure for vehicle part to be measured is single, i.e. its corresponding image is simple component, using template
Match somebody with somebody, outline, color rgb value calculate etc. modes, calculate and standard preset value between similarity.
For example, the components such as automobile grille are the single simple components of structure, if necessary to detect the quality of automobile grille,
Extraction module, for extracting the outer profile of automobile grille in image, grid, LOGO characteristics of image；It is specifically included：
Positioning unit, for positioning the band of position of automobile grille in image according to the search of standard form image, in obtaining
Net image；Wherein, the automobile grille in image is positioned using template matching technique search, i.e. the similarity of contrast images can determine
Go out the position of automobile grille, so as to achieve the purpose that positioning, improve to obtain the accuracy and robustness of middle net image；Such as：
The stencil matching algorithm of opencv, Fast Circle matching, the template matches of halcon softwares.
Pretreatment unit for being pre-processed to the middle net image, obtains the middle net image of denoising；Wherein, in advance
Processing unit carries out denoising, any one in the modes such as mean filter, medium filtering, Wiener filtering may be selected and be filtered, go
Except the Gaussian noise in image, the middle net image of denoising is obtained.
Edge detection unit for being handled using edge detection algorithm the middle net image of the denoising, is obtained
Edge binary images；Wherein, edge detection is to realize detection using object and difference of the background on certain characteristics of image, poor
It is different including gray scale, color or textural characteristics；Purpose is that the data of image are substantially reduced in the case where retaining original image attributes
Scale due to needing to consider detection result (precision is high), is preferably based on Canny edge detection algorithms herein.
Extraction unit, for converted based on hough line detection algorithm and geometry detection processing described in
Edge binary images obtain the outer profile of automobile grille in image, grid, LOGO characteristics of image；Wherein, to edge detection algorithm
Obtained bianry image directly takes hough to convert, in detailed process, it is necessary to transformation results deposit hough conversion is cumulative
Hough conversion accumulated values according to given threshold size, are less than threshold value clearing and (think one in not corresponding diagram image field by value
Straight line)；The point of hough conversion accumulated value maximums is searched, which is reset, continue to search for and is recorded is next tired
The point of value added maximum until aggregate-value all in accumulator is all zero, records these points and obtains straight line in image.It can root
According to the statistical property of automobile grille size of the threshold value with resetting field value or the edge pixel using automobile grille are reset to adjust
The location information of point, conllinear endpoint detections are come out, and then, removal noise spot and connection are generated due to edge detection
Discontinuous pixel, so that it is determined that the line segment and shape information of automobile grille, i.e. extract outer profile feature, the net of image respectively
Lattice feature, LOGO characteristics of image.
In the present embodiment, the image comprising automobile grille of acquisition is handled successively by said units, extracted
To outer profile feature, grid search-engine, LOGO characteristics of image, the position of each feature on the one hand can be positioned rapidly, on the other hand,
Robustness is also ensured in extraction position fixing process, it is ensured that outer profile feature, grid search-engine, the essence of LOGO characteristics of image of detection
Accuracy
Detection module, for detect its outer profile, grid, LOGO images geometry it is whether several in default outer profile
What shape, grid geometry, in the range of the geometry of LOGO images, judging the quality of automobile grille according to testing result is
No qualification.
Wherein, the outer profile geometry includes the respective length of upper and lower edge, lateral edges and angle of outer profile；
The size of number of the grid geometry including grid, line number and side shape grid；The geometry bag of the LOGO images
Include position, size, color and the pattern of LOGO images.
In the present embodiment, by directly detect the outer profile of automobile grille, grid, LOGO images geometry whether
In the range of default outer profile geometry, grid geometry, the geometry of LOGO images；When all outer profiles, net
Lattice, LOGO images geometry the default outer profile geometry of correspondence, grid geometry, LOGO images geometry
In form range, then the up-to-standard of automobile grille is judged；When the geometry of outer profile, grid, LOGO images is wherein arbitrary
It is one or several in the range of the default outer profile geometry of correspondence, grid geometry, the geometry of LOGO images, then
Judge the off quality of automobile grille.By way of above-mentioned Image Acquisition, image procossing, image identification and detection, relatively
It in existing artificial or other detection modes, can be needed to develop different functions according to product, need to only correspond to the ginseng of adjustment extraction
Number and default characteristic parameter, it is ensured that the versatility of automobile grille quality testing, meanwhile, also improve the intelligence of detection
Change degree and detection efficiency.
The detection module specifically includes：
First detection unit, for detecting in described image the upper and lower edge of the outer profile of automobile grille, lateral edges each
Length and angle whether in the range of the respective length of upper and lower edge, lateral edges and angle of default outer profile, according to
Testing result judges whether the outer profile geometry of automobile grille is qualified；
Second detection unit, for detecting the number of the grid of automobile grille in described image, line number and side shape grid
Size judges automobile according to testing result whether in the range of the size of the number of default grid, line number and side shape grid
Whether the grid geometry of middle net is qualified；
3rd detection unit, for detecting position, size, color and the figure of the LOGO images of automobile grille in described image
Case judges automobile grille according to testing result whether in the range of the position of default LOGO images, size, color and pattern
LOGO images geometry it is whether qualified；
Judging unit, for whether being closed according to the geometry of outer profile geometry, grid geometry, LOGO images
Lattice judge whether the quality of the automobile grille is qualified.
In the present embodiment, first detection unit, second detection unit, the 3rd detection unit are all corresponding with respective detection
As a result；It corresponds to outer profile geometry, grid geometry, the geometry of LOGO images respectively, wherein, each detection is single
Member is detected all in accordance with its characteristic parameter specifically included, can be single according to the first detection in the detection process netted in the car
Member, second detection unit, the 3rd detection unit actually detected as a result, judging unit only as each spy in each detection unit
After sign parameter detects qualification, the quality measurements of automobile grille are just qualification.Using said units to each of automobile grille
A feature is detected respectively, not only increases the accuracy of detection, meanwhile, it, can be significantly using the detection mode of mechanical intelligence
Degree improves detection efficiency.
When complicated for vehicle part to be measured, i.e. the component includes more details, and corresponding image is complexity portion
Part, is identified and similarity calculation by the way of deep learning, is such as realized using convolutional neural networks CNN.Method is：
A establishes CNN models, deep learning under B lines.
Sample preparation processes：Artificial acquisition component pattern, and do normative processing so that image size specification is consistent；People
Work labeled bracketing pattern by manually carrying out qualified and underproof division to the pattern of acquisition, and marks；Image preprocessing,
Denoising；Using the pattern of mark as training sample, CNN models are trained.
Specifically, obtain the picture of vehicle each base part to be detected, to training be need to use vehicle part picture by
Its species, model are labeled；Build neural network model：Multiple sub-networks, different characteristic used for vehicles extract,
Preliminary classification is obtained as a result, scoring in the preliminary classification result to sub- network hierarchy, obtains classification results；Training convolutional god
Through network model, network parameter is obtained：Vehicle part input convolutional neural networks model is trained, and training process uses gradient
Descending method learning network parameter；
It reaches the standard grade detection process：Acquisition pattern in real time；Pretreatment；Input CNN models；Vehicle is carried out using convolutional neural networks
Vehicle pictures are input to the convolutional neural networks model trained by sophisticated category, so as to obtain the detected status of vehicle part.
CNN models export the result judgement of acceptance or rejection.Vehicle part is determined according to testing result with the presence or absence of defect, so as to body
Detection agility and accuracy are revealed.
In the present embodiment, by by image similarity between the two compared with the preset value of the component settings,
Accurately judge whether vehicle part to be measured is qualified, greatly shorten some vehicle part and detect cumbersome flow one by one, improve
Detection efficiency.
Referring to Fig. 2, be a kind of Machine Vision Inspecting System complete structure block diagram of vehicle part provided by the invention,
On the basis of above-described embodiment, further include：
Identification module 6 is handled, for when detecting certain component appearance existing defects of the vehicle, according to described image
Threshold value is split processing, and feature the defects of to obtain region corresponding to different threshold values is identified using convolutional neural networks algorithm
The defects of component, is characterized in one or more of in the damage that jumps, dislocation, end face wound, crackle, few tooth, stomata or trachoma.
Specifically, spring damage, dislocation, the detection of end face wound, crackle, few tooth, stomata or trachoma are directly divided using neutral net
Class is divided into seven major classes according to training sample, carries out the training and detection of secondary neutral net, in this way, reduces network
Size, save the trained time, enhance the accuracy of differentiation, training method differs herein referring to the training flow of above-mentioned CNN
One repeats.
The type of defect characteristic can be further detected by above-mentioned detection, convenient for subsequently arrange specialty staff into
Row maintenance is handled, meanwhile, while detection efficiency is not reduced, also improve accuracy of detection.
Management module 7, for all parts testing result according to the corresponding identifier of the plummer and vehicle to be measured
Binding is associated, and is classified by the testing result, underproof component according to its defect type is counted respectively and is given
With report.
Specifically, the plummer label of acquisition and the image of the vehicle part of shooting are associated binding, if inspection
When measuring some component acceptance or rejection, the image institute of the image and some electronic tag corresponding to corresponding label is adjusted
Which the defects of causing has, and classifies convenient for final-period management, and the data such as statistics yield rate, defect rate are conducive to subsequent feedback control
The parameter adjustment of adjustment equipment facilitates later stage sorting subsystem to handle respectively to control the processing quality of vehicle part.
Sort subsystem 8：For identifying vehicle to be measured on each plummer using radio-frequency technique, according to the vehicle
Underproof vehicle to be measured point is passed to reprocessing region by testing result.
Specifically, in the present embodiment, each sorting subsystem identifies different vehicles to be measured according to label is different, from
And whether distinguish the vehicle part qualified, if qualified, along transmission mechanism on production line continue following process；If do not conform to
Its point is then passed to different reprocessing regions according to the unqualified type of detection and carries out anti-work by lattice, is until it detects qualification
Only.
In the production line, set the image collection module of different installation sites that can obtain the different figures of vehicle part to be measured
Picture, i.e. detect different vehicle parts, meanwhile, underproof vehicle part is transmitted respectively by sorting subsystem,
Individually handled, prevent the reprocessabilty on rejected part, cause the wasting of resources, it is time-consuming and laborious the phenomenon that, accurate conduction
It is handled convenient for staff's resolution, reduces the workload of artificial follow-up work.
Referring to Fig. 3, a kind of machine vision detection method flow chart of vehicle part is provided for the present invention, for detecting
Whether qualified state vehicle part, including：
When sensing vehicle arrival to be measured, vehicle to be measured is obtained using multiple CCD cameras respectively from multiple angles by step S1
The image of the appearance of all parts；
Step S2 when sensing vehicle arrival to be measured, identifies the identifier of the vehicle to be measured, and the component is corresponding
Image associates one by one with the identifier of vehicle；
Step S3, pretreatment described image meet the image of default specification according to the generation of vehicle different parts；
Step S4 utilizes vehicle part generating unit image to be measured in outline or template matches positioning described image；
The image of component with its pre-set image is compared and calculates similarity between the two by step S5, according to
The similarity detects the vehicle part and whether there is open defect.
In the present embodiment, the testing result of each the parts to be tested image is associated with electronic identifier, convenient for obtaining in real time
The state of all parts in the production line is known, is conducive to follow-up division of labor processing.
Referring to Fig. 4, providing a kind of machine vision detection method step S5 flow charts of vehicle part for the present invention, it is described in detail
It is as follows：
Wherein, described in step S5 by the image of component with its pre-set image is compared and is calculated between the two
The step of similarity, including：
Step S501 detects whether the image of component is the single component of pattern；
Step S502, when the vehicle part to be measured for pattern single component when, using template matches, outline or
Color rgb value calculates the similarity between the image of component and its pre-set image；
Step S503, when the vehicle part to be measured is not the single component of pattern, using convolutional neural networks mode
Calculate the similarity between the image of component and its pre-set image.
In the present embodiment, it can accurately judge that vehicle part to be measured whether there is defect through the above way, also according to
The shape and structure of component to be detected is showed calculates similarity using different calculations, and surveying object using sorting uses difference
Algorithm calculates similarity.During using step S502, for the vehicle part of detection structure complicated (image of pattern complexity), avoid
The problem of accuracy of detection is not high；Meanwhile during using step S503, for the vehicle of detection structure simple (the single image of pattern)
Component, avoids the problem of detection efficiency is not high, therefore, improves detection essence again simultaneously embodiment improves detection efficiency
Degree.
Referring to Fig. 5, a kind of machine vision detection method entire flow figure of vehicle part is provided for the present invention, above-mentioned
On the basis of embodiment, including：
Step S6 when detecting certain component appearance existing defects of the vehicle, is divided according to described image threshold value
Processing is cut, feature the defects of to obtain region corresponding to different threshold values, the defects of component is identified using neural network algorithm
It is characterized in one or more of in the damage that jumps, dislocation, end face wound, crackle, few tooth, stomata or trachoma.
Specifically, drawbacks described above characteristic model is trained respectively according to convolutional neural networks, so as to by vehicle part
Associated disadvantages can be judged whether by being input to the model, since image is generally required by passing sequentially through a variety of moulds when detecting
Type could judge that the vehicle part is that there are a kind of defects or several defects on earth and to coexist, therefore, be equivalent to general stream one by one
The accuracy of detection that journey detects the present embodiment is high.
Step S7 identifies vehicle to be measured on each plummer using radio-frequency technique, according to the testing result of the vehicle
Underproof vehicle to be measured is transferred to reprocessing region by difference.
Specifically, production line sets multiple assembly lines of doing over again and (reprocess), and subsystem is sorted on production line by switching
Path, underproof vehicle part is made to enter the reprocessing region specified, facilitates subsequent production, without artificial detection
And participation, improve the existing intelligent measurement degree of flowing water.
In conclusion the present invention is detected vehicle part to be measured on production line successively, it is first using hierarchical detection mode
First, determine whether the appearance of vehicle part to be measured is defective；Secondly, if zero defect, no longer detect；On the contrary, if there is lacking
It falls into, further detects the type for determining the defect.Inherently reduce detection workload, improve detection efficiency, together
When, subdivision detection improves accuracy of detection.So the present invention effectively overcomes various shortcoming of the prior art and has height and produce
Industry utility value.
The above-described embodiments merely illustrate the principles and effects of the present invention, and is not intended to limit the present invention.It is any ripe
Know the personage of this technology all can carry out modifications and changes under the spirit and scope without prejudice to the present invention to above-described embodiment.Cause
This, those of ordinary skill in the art is complete without departing from disclosed spirit and institute under technological thought such as
Into all equivalent modifications or change, should by the present invention claim be covered.
Claims (10)
1. whether a kind of Machine Vision Inspecting System of vehicle part, qualified for detecting the vehicle part, which is characterized in that
The system comprises：
Transport mechanism including plummer, for carrying vehicle to be measured, makes the plummer even along the production line on transmission mechanism
Speed is advanced, and sets unique identifier on each plummer；
Image collection module is installed on the detection station of the transport mechanism both sides or/and top, for sensing vehicle to be measured
When reaching, the image of the appearance of vehicle all parts to be measured is obtained respectively from multiple angles using multiple CCD cameras；
Preprocessing module meets the image of default specification for pre-processing described image according to the generation of vehicle different parts；
Locating module, for utilizing vehicle part generating unit figure to be measured in outline or template matches positioning described image
Picture；
Vision-based detection module, for being compared the image of component with its pre-set image and calculating between the two similar
Degree detects the vehicle part according to the similarity and whether there is open defect.
2. system according to claim 1, which is characterized in that each CCD camera is right in described image acquisition module
The sensor that triggering CDD video cameras are taken pictures should be provided with, and the transmission mechanism both sides are provided with the identification carrying station identification
The reader and antenna of symbol；And compensate the lighting device for light luminance of taking pictures.
3. system according to claim 1, which is characterized in that the similarity bag between the image of component and pre-set image
It includes：
Whether the image for detecting the component is the single component of pattern；
When the vehicle part to be measured is the single component of pattern, calculated using template matches, outline or color rgb value
Similarity between the image of component and its pre-set image；
When the vehicle part to be measured is not the single component of pattern, the component diagram is calculated using convolutional neural networks mode
Picture and the similarity between its pre-set image.
4. system according to claim 1, which is characterized in that including：Identification module is handled, the vehicle is detected for working as
Certain component appearance existing defects when, processing is split according to described image threshold value, to obtain area corresponding to different threshold values
The defects of domain feature, using neural network algorithm identify the component the defects of be characterized in jump damage, dislocation, end face wound, split
It is one or more of in line, few tooth, stomata or trachoma.
5. system according to claim 4, which is characterized in that including：Management module, for being corresponded to according to the plummer
The all parts testing result of identifier and vehicle to be measured be associated binding, and classify by the testing result, general
Underproof component according to its defect type is counted and reported respectively.
6. system according to claim 4, which is characterized in that including：Sort subsystem：For being identified using radio-frequency technique
Underproof vehicle to be measured, according to the testing result difference of the vehicle is transferred to and adds again by vehicle to be measured on each plummer
Work area domain.
7. whether a kind of machine vision detection method of vehicle part, qualified for detecting the vehicle part, which is characterized in that
The described method includes：
When sensing vehicle arrival to be measured, vehicle all parts to be measured are obtained respectively from multiple angles using multiple CCD cameras
The image of appearance；
When sensing vehicle arrival to be measured, the identifier of the vehicle to be measured is identified, by the corresponding image of the component and vehicle
Identifier associate one by one；
Pretreatment described image meets the image of default specification according to the generation of vehicle different parts；
Utilize vehicle part generating unit image to be measured in outline or template matches positioning described image；
The image of component with its pre-set image is compared and calculates similarity between the two, is examined according to the similarity
The vehicle part is surveyed with the presence or absence of open defect.
8. the method according to the description of claim 7 is characterized in that described compare the image of component and its pre-set image
Pair and the step of calculate similarity between the two, including：
Detect whether the image of component is the single component of pattern；
When the vehicle part to be measured is the single component of pattern, calculated using template matches, outline or color rgb value
Similarity between the image of component and its pre-set image；
When the vehicle part to be measured is not the single component of pattern, the component diagram is calculated using convolutional neural networks mode
Picture and the similarity between its pre-set image.
9. the method according to the description of claim 7 is characterized in that including：When certain the component appearance for detecting the vehicle is deposited
In defect, processing is split according to described image threshold value, feature the defects of to obtain region corresponding to different threshold values uses
Neural network algorithm identifies that the defects of component is characterized in the damage that jumps, dislocation, end face wound, crackle, few tooth, stomata or trachoma
Middle one or more.
10. system according to claim 9, which is characterized in that including：It is identified using radio-frequency technique and is treated on each plummer
Underproof vehicle to be measured is transferred to reprocessing region by the vehicle of survey according to the testing result difference of the vehicle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711329856.3A CN108109137A (en) | 2017-12-13 | 2017-12-13 | The Machine Vision Inspecting System and method of vehicle part |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711329856.3A CN108109137A (en) | 2017-12-13 | 2017-12-13 | The Machine Vision Inspecting System and method of vehicle part |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108109137A true CN108109137A (en) | 2018-06-01 |
Family
ID=62216720
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711329856.3A Pending CN108109137A (en) | 2017-12-13 | 2017-12-13 | The Machine Vision Inspecting System and method of vehicle part |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108109137A (en) |
Cited By (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108803518A (en) * | 2018-06-22 | 2018-11-13 | 苏州和必尔斯电子科技有限公司 | A kind of manufacture products detection system for electronics technology |
| CN108818537A (en) * | 2018-07-13 | 2018-11-16 | 南京工程学院 | A kind of robot industry method for sorting based on cloud deep learning |
| CN108985803A (en) * | 2018-08-28 | 2018-12-11 | 深圳市鼎昇贸易有限公司 | The return management method and Related product of e-commerce |
| CN109063708A (en) * | 2018-06-21 | 2018-12-21 | 歌尔股份有限公司 | Industrial picture characteristic recognition method and system based on contours extract |
| CN109242852A (en) * | 2018-10-30 | 2019-01-18 | 长春光华学院 | A kind of automobile engine oil circuit copper sheathing defective vision detection method |
| CN109591917A (en) * | 2018-12-17 | 2019-04-09 | 广汽丰田汽车有限公司 | Vehicle identification detects processing method, vehicle detection apparatus and storage medium |
| CN109685253A (en) * | 2018-12-07 | 2019-04-26 | 上海睿深电子科技有限公司 | Suitable for planetary gear quality control system, method, terminal and storage medium |
| CN110142230A (en) * | 2019-05-23 | 2019-08-20 | 北京阿丘机器人科技有限公司 | A kind of method and device detecting open defect |
| CN110180799A (en) * | 2019-06-28 | 2019-08-30 | 中船黄埔文冲船舶有限公司 | A kind of part method for sorting and system based on machine vision |
| CN110223293A (en) * | 2019-06-21 | 2019-09-10 | 中国神华能源股份有限公司 | The intelligent identification Method and identification device of train body breakage |
| CN110853060A (en) * | 2019-11-14 | 2020-02-28 | 上海眼控科技股份有限公司 | Vehicle appearance detection method and device, computer equipment and storage medium |
| CN110910339A (en) * | 2019-07-10 | 2020-03-24 | 研祥智能科技股份有限公司 | Logo defect detection method and device |
| CN110992336A (en) * | 2019-12-02 | 2020-04-10 | 东莞西尼自动化科技有限公司 | Small sample defect detection method based on image processing and artificial intelligence |
| CN111091548A (en) * | 2019-12-12 | 2020-05-01 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon adapter dislocation fault image identification method and system based on deep learning |
| CN111102920A (en) * | 2019-12-18 | 2020-05-05 | 佛山科学技术学院 | Mechanical component quality inspection method and system based on augmented reality |
| CN111242951A (en) * | 2020-01-08 | 2020-06-05 | 上海眼控科技股份有限公司 | Vehicle detection method, device, computer equipment and storage medium |
| CN111399464A (en) * | 2020-04-08 | 2020-07-10 | 中车青岛四方机车车辆股份有限公司 | Intelligent vehicle quality inspection method and system |
| CN111738367A (en) * | 2020-08-17 | 2020-10-02 | 成都中轨轨道设备有限公司 | Part classification method based on image recognition |
| CN112085709A (en) * | 2020-08-19 | 2020-12-15 | 浙江华睿科技有限公司 | Image contrast method and equipment |
| TWI716923B (en) * | 2018-09-04 | 2021-01-21 | 開曼群島商創新先進技術有限公司 | Car damage image generation method and device based on GAN network |
| CN112691920A (en) * | 2020-12-08 | 2021-04-23 | 湖北博微瑞天信息技术有限公司 | Machine vision recognition system with multi-information fusion |
| CN112819055A (en) * | 2021-01-25 | 2021-05-18 | 北京巅峰科技有限公司 | Vehicle defect identification method, device, equipment and storage medium |
| CN113012157A (en) * | 2021-05-24 | 2021-06-22 | 深圳市菲凡数据科技有限公司 | Visual detection method and system for equipment defects |
| CN113192358A (en) * | 2021-04-26 | 2021-07-30 | 贵州车秘科技有限公司 | Parking management system based on thermal imaging technology in intelligent parking field and use method thereof |
| CN113554024A (en) * | 2021-07-27 | 2021-10-26 | 宁波小遛共享信息科技有限公司 | Method and device for determining cleanliness of vehicle and computer equipment |
| CN113671907A (en) * | 2020-05-14 | 2021-11-19 | 华晨宝马汽车有限公司 | Production monitoring method and device for monitoring vehicle production and storage medium |
| CN114219806A (en) * | 2022-02-22 | 2022-03-22 | 成都数联云算科技有限公司 | Automobile radar detection method, device, equipment, medium and product |
| CN114739293A (en) * | 2022-03-25 | 2022-07-12 | 北京博联众睿机器人科技有限公司 | Vehicle body measuring method, system, device and electronic equipment |
| CN115082508A (en) * | 2022-08-18 | 2022-09-20 | 山东省蓝睿科技开发有限公司 | Ocean buoy production quality detection method |
| JP2023064620A (en) * | 2021-10-26 | 2023-05-11 | ダイハツ工業株式会社 | Object determination device and object determination system |
| US11710158B2 (en) | 2020-01-23 | 2023-07-25 | Ford Global Technologies, Llc | Vehicle damage estimation |
| CN117269180A (en) * | 2023-11-24 | 2023-12-22 | 成都数之联科技股份有限公司 | Vehicle appearance detection method, device, server and computer readable storage medium |
| CN117911411A (en) * | 2024-03-19 | 2024-04-19 | 南京认知物联网研究院有限公司 | Computer vision detection method and device based on parallel detection of picture streams |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102253048A (en) * | 2011-04-29 | 2011-11-23 | 惠州市钧悦科技有限公司 | Machine vision detection method and system for detection of various products |
| CN104280406A (en) * | 2014-09-16 | 2015-01-14 | 中国科学院广州能源研究所 | Machine vision system for detecting surface defects of copper part |
| CN106290392A (en) * | 2016-08-05 | 2017-01-04 | 宁波达尔机械科技有限公司 | A kind of little micro-bearing surface pitting defects online test method and system thereof |
| CN107274398A (en) * | 2017-06-16 | 2017-10-20 | 重庆越畅汽车科技有限公司 | Based on automobile grille quality detecting system and method |
-
2017
- 2017-12-13 CN CN201711329856.3A patent/CN108109137A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102253048A (en) * | 2011-04-29 | 2011-11-23 | 惠州市钧悦科技有限公司 | Machine vision detection method and system for detection of various products |
| CN104280406A (en) * | 2014-09-16 | 2015-01-14 | 中国科学院广州能源研究所 | Machine vision system for detecting surface defects of copper part |
| CN106290392A (en) * | 2016-08-05 | 2017-01-04 | 宁波达尔机械科技有限公司 | A kind of little micro-bearing surface pitting defects online test method and system thereof |
| CN107274398A (en) * | 2017-06-16 | 2017-10-20 | 重庆越畅汽车科技有限公司 | Based on automobile grille quality detecting system and method |
Non-Patent Citations (1)
| Title |
|---|
| 彭博: "汽车车灯玻壳视觉检测系统的设计与实现", 《万方学位论文库》 * |
Cited By (44)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063708A (en) * | 2018-06-21 | 2018-12-21 | 歌尔股份有限公司 | Industrial picture characteristic recognition method and system based on contours extract |
| CN108803518A (en) * | 2018-06-22 | 2018-11-13 | 苏州和必尔斯电子科技有限公司 | A kind of manufacture products detection system for electronics technology |
| CN108818537B (en) * | 2018-07-13 | 2023-04-14 | 南京工程学院 | Robot industry sorting method based on cloud deep learning |
| CN108818537A (en) * | 2018-07-13 | 2018-11-16 | 南京工程学院 | A kind of robot industry method for sorting based on cloud deep learning |
| CN108985803A (en) * | 2018-08-28 | 2018-12-11 | 深圳市鼎昇贸易有限公司 | The return management method and Related product of e-commerce |
| TWI716923B (en) * | 2018-09-04 | 2021-01-21 | 開曼群島商創新先進技術有限公司 | Car damage image generation method and device based on GAN network |
| CN109242852A (en) * | 2018-10-30 | 2019-01-18 | 长春光华学院 | A kind of automobile engine oil circuit copper sheathing defective vision detection method |
| CN109685253A (en) * | 2018-12-07 | 2019-04-26 | 上海睿深电子科技有限公司 | Suitable for planetary gear quality control system, method, terminal and storage medium |
| CN109591917A (en) * | 2018-12-17 | 2019-04-09 | 广汽丰田汽车有限公司 | Vehicle identification detects processing method, vehicle detection apparatus and storage medium |
| CN110142230A (en) * | 2019-05-23 | 2019-08-20 | 北京阿丘机器人科技有限公司 | A kind of method and device detecting open defect |
| CN110223293A (en) * | 2019-06-21 | 2019-09-10 | 中国神华能源股份有限公司 | The intelligent identification Method and identification device of train body breakage |
| CN110180799A (en) * | 2019-06-28 | 2019-08-30 | 中船黄埔文冲船舶有限公司 | A kind of part method for sorting and system based on machine vision |
| CN110910339B (en) * | 2019-07-10 | 2022-11-22 | 研祥智能科技股份有限公司 | Logo defect detection method and device |
| CN110910339A (en) * | 2019-07-10 | 2020-03-24 | 研祥智能科技股份有限公司 | Logo defect detection method and device |
| WO2021004130A1 (en) * | 2019-07-10 | 2021-01-14 | 研祥智能科技股份有限公司 | Logo defect detection method and device |
| CN110853060A (en) * | 2019-11-14 | 2020-02-28 | 上海眼控科技股份有限公司 | Vehicle appearance detection method and device, computer equipment and storage medium |
| CN110992336A (en) * | 2019-12-02 | 2020-04-10 | 东莞西尼自动化科技有限公司 | Small sample defect detection method based on image processing and artificial intelligence |
| CN111091548A (en) * | 2019-12-12 | 2020-05-01 | 哈尔滨市科佳通用机电股份有限公司 | Railway wagon adapter dislocation fault image identification method and system based on deep learning |
| CN111102920A (en) * | 2019-12-18 | 2020-05-05 | 佛山科学技术学院 | Mechanical component quality inspection method and system based on augmented reality |
| CN111242951A (en) * | 2020-01-08 | 2020-06-05 | 上海眼控科技股份有限公司 | Vehicle detection method, device, computer equipment and storage medium |
| US11710158B2 (en) | 2020-01-23 | 2023-07-25 | Ford Global Technologies, Llc | Vehicle damage estimation |
| CN111399464A (en) * | 2020-04-08 | 2020-07-10 | 中车青岛四方机车车辆股份有限公司 | Intelligent vehicle quality inspection method and system |
| CN113671907A (en) * | 2020-05-14 | 2021-11-19 | 华晨宝马汽车有限公司 | Production monitoring method and device for monitoring vehicle production and storage medium |
| CN111738367A (en) * | 2020-08-17 | 2020-10-02 | 成都中轨轨道设备有限公司 | Part classification method based on image recognition |
| CN111738367B (en) * | 2020-08-17 | 2020-11-13 | 成都中轨轨道设备有限公司 | Part classification method based on image recognition |
| CN112085709A (en) * | 2020-08-19 | 2020-12-15 | 浙江华睿科技有限公司 | Image contrast method and equipment |
| CN112085709B (en) * | 2020-08-19 | 2024-03-22 | 浙江华睿科技股份有限公司 | Image comparison method and device |
| CN112691920A (en) * | 2020-12-08 | 2021-04-23 | 湖北博微瑞天信息技术有限公司 | Machine vision recognition system with multi-information fusion |
| CN112819055A (en) * | 2021-01-25 | 2021-05-18 | 北京巅峰科技有限公司 | Vehicle defect identification method, device, equipment and storage medium |
| CN113192358A (en) * | 2021-04-26 | 2021-07-30 | 贵州车秘科技有限公司 | Parking management system based on thermal imaging technology in intelligent parking field and use method thereof |
| CN113012157B (en) * | 2021-05-24 | 2021-07-20 | 深圳市菲凡数据科技有限公司 | Visual detection method and system for equipment defects |
| CN113012157A (en) * | 2021-05-24 | 2021-06-22 | 深圳市菲凡数据科技有限公司 | Visual detection method and system for equipment defects |
| CN113554024A (en) * | 2021-07-27 | 2021-10-26 | 宁波小遛共享信息科技有限公司 | Method and device for determining cleanliness of vehicle and computer equipment |
| JP7425032B2 (en) | 2021-10-26 | 2024-01-30 | ダイハツ工業株式会社 | Object determination device and object determination system |
| JP2023064620A (en) * | 2021-10-26 | 2023-05-11 | ダイハツ工業株式会社 | Object determination device and object determination system |
| CN114219806A (en) * | 2022-02-22 | 2022-03-22 | 成都数联云算科技有限公司 | Automobile radar detection method, device, equipment, medium and product |
| CN114219806B (en) * | 2022-02-22 | 2022-04-22 | 成都数联云算科技有限公司 | Automobile radar detection method, device, equipment, medium and product |
| CN114739293A (en) * | 2022-03-25 | 2022-07-12 | 北京博联众睿机器人科技有限公司 | Vehicle body measuring method, system, device and electronic equipment |
| CN115082508A (en) * | 2022-08-18 | 2022-09-20 | 山东省蓝睿科技开发有限公司 | Ocean buoy production quality detection method |
| CN115082508B (en) * | 2022-08-18 | 2022-11-22 | 山东省蓝睿科技开发有限公司 | Ocean buoy production quality detection method |
| CN117269180A (en) * | 2023-11-24 | 2023-12-22 | 成都数之联科技股份有限公司 | Vehicle appearance detection method, device, server and computer readable storage medium |
| CN117269180B (en) * | 2023-11-24 | 2024-03-12 | 成都数之联科技股份有限公司 | Vehicle appearance detection method, device, server and computer readable storage medium |
| CN117911411A (en) * | 2024-03-19 | 2024-04-19 | 南京认知物联网研究院有限公司 | Computer vision detection method and device based on parallel detection of picture streams |
| CN117911411B (en) * | 2024-03-19 | 2024-05-24 | 南京认知物联网研究院有限公司 | Computer vision detection method and device based on parallel detection of picture streams |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108109137A (en) | The Machine Vision Inspecting System and method of vehicle part | |
| Aquino et al. | Automated early yield prediction in vineyards from on-the-go image acquisition | |
| CN106875373B (en) | Mobile phone screen MURA defect detection method based on convolutional neural network pruning algorithm | |
| CN109308447A (en) | The method of equipment operating parameter and operating status is automatically extracted in remote monitoriong of electric power | |
| CN103279765B (en) | Steel wire rope surface damage detection method based on images match | |
| CN111047655B (en) | High-definition camera cloth defect detection method based on convolutional neural network | |
| CN109886298A (en) | A kind of detection method for quality of welding line based on convolutional neural networks | |
| CN109550712A (en) | A kind of chemical fiber wire tailfiber open defect detection system and method | |
| CN110084165A (en) | The intelligent recognition and method for early warning of anomalous event under the open scene of power domain based on edge calculations | |
| CN108491892A (en) | fruit sorting system based on machine vision | |
| CN108090434B (en) | Rapid ore identification method | |
| Sahu et al. | Identification and classification of mango fruits using image processing | |
| CN113324864B (en) | Pantograph carbon slide plate abrasion detection method based on deep learning target detection | |
| CN104268505A (en) | Automatic cloth defect point detection and recognition device and method based on machine vision | |
| CN109918971A (en) | Number detection method and device in monitor video | |
| CN109711322A (en) | A kind of people's vehicle separation method based on RFCN | |
| CN103454285A (en) | Transmission chain quality detection system based on machine vision | |
| CN109741314A (en) | A kind of visible detection method and system of part | |
| CN110675395A (en) | Intelligent on-line monitoring method for power transmission line | |
| CN104198497A (en) | Surface defect detection method based on visual saliency map and support vector machine | |
| CN109215015A (en) | A kind of online visible detection method of silk cocoon based on convolutional neural networks | |
| CN111199556A (en) | Indoor pedestrian detection and tracking method based on camera | |
| CN107886131A (en) | One kind is based on convolutional neural networks detection circuit board element polarity method and apparatus | |
| CN112069985A (en) | High-resolution field image rice ear detection and counting method based on deep learning | |
| CN106548131A (en) | A kind of workmen's safety helmet real-time detection method based on pedestrian detection |
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: 20180601 |