CN204440490U - A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing - Google Patents

A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing Download PDF

Info

Publication number
CN204440490U
CN204440490U CN201520017896.4U CN201520017896U CN204440490U CN 204440490 U CN204440490 U CN 204440490U CN 201520017896 U CN201520017896 U CN 201520017896U CN 204440490 U CN204440490 U CN 204440490U
Authority
CN
China
Prior art keywords
bush
type parts
axle type
compressed sensing
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.)
Expired - Fee Related
Application number
CN201520017896.4U
Other languages
Chinese (zh)
Inventor
谢昕
徐殷
李波
黄志刚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
East China Jiaotong University
Original Assignee
East China Jiaotong University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by East China Jiaotong University filed Critical East China Jiaotong University
Priority to CN201520017896.4U priority Critical patent/CN204440490U/en
Application granted granted Critical
Publication of CN204440490U publication Critical patent/CN204440490U/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

Based on a Model For The Bush-axle Type Parts surface imperfection on-line measuring device for compressed sensing, comprise detecting and controlling system, detect servicing unit (1), industrial camera (11,12,13).Detecting and controlling system comprises controller (3), driver (4), industrial computer (2), sense channel (18) and sort channel; Detect servicing unit and comprise stepper motor (17), gear (16), screw mandrel (15) and slide block (14).The utility model adopts three video cameras, part to be measured is carried out to the image taking of different azimuth, arrange image procossing layer and part defect to detect and diagnostic horizon, respectively personalisation process is carried out to different levels, utilizes compressed sensing technology to carry out compression of images to image.The utility model adopts defect training sample dictionary to carry out defects detection.Compared with conventional machines visible sensation method, shorten image processing time, under the prerequisite ensureing enough accuracy, improve the reliability of defect recognition, meet the requirement of real-time of on-line checkingi.

Description

A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing
Technical field
The utility model relates to a kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing, belongs to industrial part defects detection research field.
Background technology
Along with global manufacturing center is progressively to China's transfer, while production constantly expands with manufacturing capacity, people it is also proposed more and more higher requirement to product, and product surface quality is to the direct use of product or deep processing important.Model For The Bush-axle Type Parts is in production, assembling process, and the place such as surface, end face of part may produce the defects such as various cut, polishing scratch, pit.The task of defects detection detects defective part exactly, and the clearance of guarantee defect recognition rate, substandard product reaches higher level, and then improves the oeverall quality of product.Therefore how effectively detecting part and process, is the subject matter faced by our current defect inspection process.
Current domestic many manufacturing enterprises still adopt traditional artificial sampling observation method to carry out surface defects of products detection, and not only detection speed causes erroneous judgement, undetected slowly but also very easily, cannot meet the requirement that enterprise improves the quality of products.In recent years, successively the methods such as such as infrared inspection, ultrasonic Detection Method, EDDY CURRENT and magnetic powder inspection are there is, but be difficult to reach well-content effect under industry spot high power pulse noise circumstance, the requirement of modern measure technology to the aspect such as online, accurate, real-time, comprehensive cannot be met.Along with manufacturing industry is towards lean production and the target development pursuing zero-fault, explores objective, effective, reliable quality control measure, improve surface defects of parts vision on-line checkingi level, become the problem that many enterprises are in the urgent need to address.
Summary of the invention
The purpose of this utility model is: cause erroneous judgement, the present situation such as undetected slowly and very easily for Traditional Man detection speed, the utility model discloses a kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing, and this device can detect fast to part.
The technical solution of the utility model is: the image photographed by three video cameras draws last feedback information respectively by detecting and controlling system, judges this part whether existing defects carry out good and the bad classification to part.
Based on a Model For The Bush-axle Type Parts surface imperfection on-line measuring device for compressed sensing, comprise, detecting and controlling system, detection servicing unit, industrial camera and gigabit network cable.Industrial camera connects detecting and controlling system by gigabit network cable; Detecting and controlling system carries out on-line checkingi by detecting servicing unit.
Detecting and controlling system described in the utility model comprises controller, driver, industrial computer, sense channel and sort channel.Model For The Bush-axle Type Parts enters sense channel, the sample image that described industrial computer obtains Model For The Bush-axle Type Parts by gigabit network cable from industrial camera carries out pre-service, utilize compressed sensing technology to compress image, realize the identification to part defect to be measured according to the discrimination standard of compressed sensing; Recognition result is sent to controller by industrial computer, is worked by controller control and drive system, driver implementation controller order Open valve, and Model For The Bush-axle Type Parts is sent into sort channel by testing result.
Detection servicing unit described in the utility model comprises stepper motor, gear, screw mandrel and slide block.Part to be measured just enters into sense channel, push rod on slide block withstands location, its left side, video camera starts pickup image, stepper motor driven gear rotates, gear drive leading screw is moved to the left 1/3 part outer post disc girth to be measured, part to be measured stops after mobile 1/3 outside cylinder disc girth due to self gravitation on slideway, and video camera absorbs the image in second face.Like this through 3 movements of slide block, photograph the picture in 3 faces of part to be measured, by the connection of gigabit network cable and industrial computer, picture is passed to industrial computer, by image processing software, Treatment Analysis is carried out to picture.
Industrial camera described in the utility model has three, is arranged on three orientation of the Model For The Bush-axle Type Parts being arranged in passage respectively, and the second video camera and the 3rd video camera lay respectively at two sides of passage, and the first position for video camera is directly over Model For The Bush-axle Type Parts; The camera of three video cameras all points to Model For The Bush-axle Type Parts; Every platform video camera other configuration light source; The output terminal of three video cameras connects industrial computer by gigabit network cable.
Industrial computer described in the utility model processes in time to the image that three video cameras collect, thus the shooting time of three video cameras is controlled by controller, and the valve in machine electric machine is controlled by driver, thus the on-line checkingi realized part to be measured and classification.
Sort channel described in the utility model is made up of qualified passage and defective passage, these two passages are controlled by valve, if part to be measured is qualified part, valve will not do any variation, if part to be measured is rejected part, then controller will drive valve, and plate is connected to defective passage.
Transmission channel described in the utility model is by setting certain transfer rate by parts transport to be measured to surveyed area.
Detecting and controlling system workflow described in the utility model comprises:
(1) industrial computer carries out pre-service to obtained Model For The Bush-axle Type Parts sample image, comprises image filtering denoising, enhancing, rim detection and defect Segmentation, to obtain defect area; The process such as sampling rate adjusting, size normalization is implemented to the sample image after process, trains sample image according to compressed sensing principle and set up sample redundant dictionary;
(2) the view data synchronous transmission arrived by three camera acquisitions, to industrial computer image processing system, carries out pre-service to obtained part image to be measured; To the three width images with overlapping region photographed directly over part outer post surface to be measured, utilize and carry out image mosaic based on compressed sensing and theory of wavelet transformation, form a complete part outer post surface image to be measured.Both ends of the surface imagery exploitation compressed sensing technology carries out compression of images;
(3) the complete outer cylinder surface image after merging splicing and both ends of the surface image are sampled respectively according to certain sample frequency, select suitable orthogonal basis split-matrix and observing matrix, according to sample redundant dictionary, calculate the rarefaction representation of this three width image respectively, thus realize the identification to part defect to be measured according to the discrimination standard of compressed sensing, uniquely exported data judgment value, automatically control part to be measured whether existing defects classifying according to this output decision content, realize online automatic detection and the sorting of part.
The beneficial effects of the utility model are, the utility model is when employing three video cameras, part to be measured is carried out to the image taking of different azimuth, arrange image procossing layer and part defect to detect and diagnostic horizon, respectively personalisation process is carried out to different levels, at image procossing layer in order to shorten the processing time of image, compressed sensing technology is utilized to carry out compression of images to image.Detecting with diagnostic horizon in order to ensure the accuracy of defects detection at part defect, adopting defect training sample dictionary to carry out defects detection.Compared with conventional machines visible sensation method, the method carries out reduced overall sampling to image, can omit the process that some are unnecessary, shorten image processing time, under the prerequisite ensureing enough accuracy, improve the reliability of defect recognition, meet the requirement of real-time of on-line checkingi.
The utility model contributes to improving detection efficiency in the application in defects detection field, shortens the cycle of operation of automatic checkout system, greatly can save the production cost of part, realizes the real-time monitoring of part defect detection and automatically controls.The utility model is applicable to industrial processes bottom bracket axle class part defect check processing.
Accompanying drawing explanation
Fig. 1 is detection technique route map of the present utility model;
Fig. 2 is detection servicing unit outboard profile of the present utility model;
Fig. 3 is detection servicing unit front view (FV) of the present utility model;
Fig. 4 is the equipment frame figure of the utility model detecting and controlling system;
Fig. 5 is the compressed sensing Image Fusion process flow diagram of individual layer wavelet transformation;
Fig. 6 is connecting elements defects detection and differentiation process flow diagram.
Picture in picture number represents: 1 is control servicing unit; 2 is industrial computers; 3 is controllers; 4 is drivers; 5 is valves; 6 is parts to be measured; 11 is first video cameras; 12 is second video cameras; 13 is the 3rd video cameras; 14 is slide blocks; 15 is screw mandrels; 16 is gears; 17 is stepper motors; 18 is sense channels; 19 is transport tapes.
Embodiment
The specific implementation method of the utility model method as shown in Figure 1, Figure 2 and Figure 4.Fig. 1 is detection technique route map; Fig. 2 is for detecting servicing unit schematic diagram; Fig. 4 is the equipment frame figure of the utility model detecting and controlling system; Fig. 5 is the compressed sensing Image Fusion process flow diagram based on individual layer wavelet transformation; Fig. 6 is that part defect detects and differentiates process flow diagram.
The hardware device of the present embodiment mainly comprises controller, the industrial camera (not needing capture card) of three band PCI-Express interfaces, a high-performance industrial computer, gigabit network cable 40 meters and detection servicing unit.
Its equipment frame method as shown in Figure 2, Figure 3 and Figure 4, wherein places the first video camera 11 and light source directly over part 6 outer cylinder surface to be measured, places the second video camera 12 and the 3rd video camera 13 and light source respectively at part both sides to be measured end face.Part 6 enters region to be measured from transmission channel.Three industrial cameras are connected with industrial computer 2 respectively by gigabit network cable.
Detect servicing unit and comprise stepper motor 17, gear 16 screw mandrel 15 and slide block 14.
Parts image data fusion flow process to be measured as shown in Figure 5, absorbs part image to be measured by three video cameras, is sent to industrial computer and carries out Image semantic classification, comprise image filtering denoising, enhancing, sharpening, rim detection and defect Segmentation etc.; The compressed sensing image interfusion method of sampling individual layer wavelet transformation, retains the low frequency coefficient after wavelet transformation, only observes high distribution coefficient; Different fusion rules is adopted to merge respectively to low frequency coefficient and high distribution coefficient observed reading subsequently; Wavelet inverse transformation is implemented to the High-frequency and low-frequency signal after merging, finally obtains a complete part outer post surface image to be measured after merging.
Part defect detects and differentiates flow process as shown in Figure 6, the outer cylinder surface image obtained after fusion and both ends of the surface image are sampled according to certain sample frequency, select suitable orthogonal basis split-matrix and observing matrix, according to sample redundant dictionary, calculate the rarefaction representation of three width testing images respectively, thus realize the identification (as long as having a face to be detected defectiveness just judge that this part is as defectiveness part) to part defect to be measured according to the discrimination standard of compressed sensing, uniquely exported data judgment value.If part existing defects to be measured, then drive sorting mechanism to realize online automatic detection and the sorting of part, otherwise be judged to be qualified part.
The idiographic flow of the utility model embodiment whole part defect testing process operation is as follows:
1st step: three industrial cameras are individually fixed in directly over part outer post surface to be measured and in the optimum position of part both ends of the surface to be measured, in pulpit, video camera are connected with industrial computer before fixing.
2nd step: first pre-service is carried out, to obtain defect area to obtained Model For The Bush-axle Type Parts sample image.The process such as sampling rate adjusting, size normalization is implemented to the sample image after process, trains sample image according to compressed sensing principle and set up sample redundant dictionary.
3rd step: the view data synchronous transmission arrived by three camera acquisitions in mechanical motor device, to special industrial computer image processing system, carries out pre-service, to obtain defect area to obtained part image to be measured.To the three width images with overlapping region that video camera directly over piece surface to be measured photographs, utilize and based on compressed sensing technology, image is compressed, obtain the local feature value of image, combined with wavelet transformed principle carries out image co-registration again, forms a complete part outer post surface image to be measured.Both ends of the surface image carries out compression of images by compressed sensing technology.
4th step: utilize compressed sensing and machine vision method, complete outer cylinder surface image after fusion and both ends of the surface image are sampled according to certain sample frequency, select suitable orthogonal basis split-matrix and observing matrix, according to sample redundant dictionary, calculate the rarefaction representation of three width images respectively, thus realize the identification (as long as having a face to be detected defectiveness just judge that this part is as defectiveness part) to part defect to be measured according to the discrimination standard of compressed sensing, uniquely exported data judgment value.Automatically control part to be measured whether existing defects classifying according to this decision content, then return the first step, until detected all parts.

Claims (5)

1. based on a Model For The Bush-axle Type Parts surface imperfection on-line measuring device for compressed sensing, it is characterized in that, comprise, detecting and controlling system, detection servicing unit, industrial camera and gigabit network cable; Industrial camera connects detecting and controlling system by gigabit network cable; Detecting and controlling system carries out on-line checkingi by detecting servicing unit.
2. a kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing according to claim 1, it is characterized in that, described detecting and controlling system comprises controller, driver, industrial computer, sense channel and sort channel; Model For The Bush-axle Type Parts enters sense channel, the sample image that described industrial computer obtains Model For The Bush-axle Type Parts by gigabit network cable from industrial camera carries out pre-service, utilize compressed sensing technology to compress image, realize the identification to part defect to be measured according to the discrimination standard of compressed sensing; Recognition result is sent to controller by industrial computer, is worked by controller control and drive system, driver implementation controller order Open valve, and Model For The Bush-axle Type Parts is sent into sort channel by testing result.
3. a kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing according to claim 1, it is characterized in that, described detection servicing unit comprises stepper motor, gear, screw mandrel and slide block; Part to be measured just enters into sense channel, push rod on slide block withstands location, its left side, video camera starts pickup image, stepper motor driven gear rotates, gear drive leading screw is moved to the left 1/3 part outer post disc girth to be measured, part to be measured stops after mobile 1/3 outside cylinder disc girth due to self gravitation on slideway, and video camera absorbs the image in second face; Like this through 3 movements of slide block, photograph the picture in 3 faces of part to be measured, by the connection of gigabit network cable and computing machine, picture is passed to computing machine, by image processing software, Treatment Analysis is carried out to picture.
4. a kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing according to claim 1, it is characterized in that, described industrial camera has three, be arranged on three orientation of the Model For The Bush-axle Type Parts being arranged in passage respectively, second video camera and the 3rd video camera lay respectively at two sides of passage, and the first position for video camera is directly over Model For The Bush-axle Type Parts; The camera of three video cameras all points to Model For The Bush-axle Type Parts; Every platform video camera other configuration light source; The output terminal of three video cameras connects industrial computer by gigabit network cable.
5. a kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing according to claim 2, it is characterized in that, described sort channel is made up of qualified passage and defective passage, these two passages are controlled by valve, if part to be measured is qualified part, valve will not do any variation, if part to be measured is rejected part, then controller will drive valve, and plate is connected to defective passage.
CN201520017896.4U 2015-01-12 2015-01-12 A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing Expired - Fee Related CN204440490U (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201520017896.4U CN204440490U (en) 2015-01-12 2015-01-12 A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201520017896.4U CN204440490U (en) 2015-01-12 2015-01-12 A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing

Publications (1)

Publication Number Publication Date
CN204440490U true CN204440490U (en) 2015-07-01

Family

ID=53608327

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201520017896.4U Expired - Fee Related CN204440490U (en) 2015-01-12 2015-01-12 A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing

Country Status (1)

Country Link
CN (1) CN204440490U (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106091934A (en) * 2016-08-31 2016-11-09 魏志丽 Axial workpiece machine vision size detecting system
CN110009617A (en) * 2019-04-02 2019-07-12 浙江大学 A kind of shaft forgings online quality control method

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106091934A (en) * 2016-08-31 2016-11-09 魏志丽 Axial workpiece machine vision size detecting system
CN110009617A (en) * 2019-04-02 2019-07-12 浙江大学 A kind of shaft forgings online quality control method

Similar Documents

Publication Publication Date Title
CN109239102B (en) CNN-based flexible circuit board appearance defect detection method
CN103499585B (en) Based on noncontinuity lithium battery film defect inspection method and the device thereof of machine vision
CN104063873A (en) Shaft sleeve part surface defect on-line detection method based on compressed sensing
CN102680478A (en) Detection method and device of surface defect of mechanical part based on machine vision
CN102183528A (en) Defect detection device and method of duplicate gear
CN105905132A (en) Intelligent rail state patrolling device and method
CN202793389U (en) On-line routing inspection device for high-speed rail overhead line system
CN103454285A (en) Transmission chain quality detection system based on machine vision
CN204924983U (en) Lens defect automatic checkout device of grating modulation
CN103837087B (en) Pantograph automatic testing method based on active shape model
CN108918539A (en) A kind of apparent Defect inspection device and method of tunnel structure
CN108288288B (en) Method, device and system for measuring precision shaft dimension based on visual identification
CN202058221U (en) Passenger flow statistic device based on binocular vision
CN103983426B (en) The detection of a kind of defect of optical fiber based on machine vision and sorting technique
CN204440490U (en) A kind of Model For The Bush-axle Type Parts surface imperfection on-line measuring device based on compressed sensing
CN109724984A (en) A kind of defects detection identification device and method based on deep learning algorithm
CN105651782A (en) Engine cylinder cover defect automatic detection equipment based on machine vision
CN102175692A (en) System and method for detecting defects of fabric gray cloth quickly
CN105184229A (en) Online learning based real-time pedestrian detection method in dynamic scene
CN110216080A (en) A kind of Quality Monitoring Control System of the PCB processing producing line based on image comparison
CN103745238A (en) Pantograph identification method based on AdaBoost and active shape model
Neethu et al. Role of computer vision in automatic inspection systems
CN105809219B (en) A kind of the prefabricated pipe section quality testing statistical system and method for tunnel prefabricated pipe section production line
CN204630546U (en) A kind of automatic measure on line visualization detection system of wind power equipment securing member
CN105354848B (en) A kind of optimization method of the Cognex Surface Quality Inspection System of hot galvanizing producing line

Legal Events

Date Code Title Description
GR01 Patent grant
C14 Grant of patent or utility model
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20150701

Termination date: 20160112