CN102495069B - Method for detecting defects of chain belts of zipper on basis of digital image processing - Google Patents
Method for detecting defects of chain belts of zipper on basis of digital image processing Download PDFInfo
- Publication number
- CN102495069B CN102495069B CN 201110403923 CN201110403923A CN102495069B CN 102495069 B CN102495069 B CN 102495069B CN 201110403923 CN201110403923 CN 201110403923 CN 201110403923 A CN201110403923 A CN 201110403923A CN 102495069 B CN102495069 B CN 102495069B
- Authority
- CN
- China
- Prior art keywords
- image
- zip fastener
- detection
- chain band
- chain
- 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
Links
Images
Landscapes
- Slide Fasteners (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method for detecting the defects of chain belts of a zipper on the basis of digital image processing, belonging to a novel technology in the technical field of appearance quality inspection of zippers in the hardware industry. According to the method, a detection device consisting of a target positioning and triggering device, an image acquisition sensor, a conveyor belt driving device, a background light source device, a feed belt, a sorting device and a PC (Personal Computer) machine or an embedded control system is utilized to carry out detection; and a zipper product to be detected is fed into a feed belt platform of the detection device and is then fed to a detection region under the drive of the driving device, then the zipper product is subjected to target positioning by the target positioning and triggering device, a chain belt digital image of the zipper is acquired by the image acquisition sensor, the image is transmitted to the PC machine or the embedded control system, the target image is processed by the PC machine or the embedded control system, and whether a chain belt region of the zipper has the defects is compared and judged according to a processing result and the judgment standard. The method has the characteristics of simple detection process, high detection speed, high detection accuracy and the like.
Description
Technical field
The present invention is a kind of method that adopts Digital Image Processing to come the chain band of slide fastener is carried out defects detection, belongs to the new technology in the slide fastener presentation quality inspection technology field in the hardware industry.
Background technology
Zipper product is the necessity of people's daily life.China is slide fastener producing country the biggest in the world, and zipper industry is after the buffer-type of experience financial crisis increases, and whole zipper industry growth momentum is powerful after 2009.But owing to reasons such as production technologies, the breakage of chain band, the unequal defective of length in the production run of zipper product, can occur, cause the generation of slide fastener substandard products.At present, the presentation quality of zipper product generally adopts manual method, is judged by vision and subjective impression by a large amount of product line workers.On the one hand, detection efficiency is low, and cost is high, and labor strength is large; On the other hand, the human factor impact is very large, the work of large workload easily causes the fatigue that detects the employee, human factor has directly affected the reliability of the examination and test of products, cause product defect to fail to judge and misjudge, make detection less stable, the inefficiency of product quality, restricted the healthy and rapid development of zipper industry.Improvement of production process is realized the upgrading of industrial technology, improves the production efficiency of slide fastener, reduces the production cost of product, become industry development in the urgent need to.
Automatic checkout system based on Digital Image Processing can increase substantially production efficiency and product quality, improve flexibility and the automaticity of producing, can increase substantially production efficiency, reduce production costs, improve to a great extent the level of intelligence of production automation level and detection system, detection technique based on Digital Image Processing has stability by force because of it, good and the precision advantages of higher of operational efficiency, along with the needs with modern production of developing rapidly of digital image processing techniques, the computer picture mode identification technology is in industry, national defence, the various aspects such as scientific research are more widely used.Therefore, in the modern industry production run, be widely used in the fields such as production run condition monitoring, product inspection and quality control based on the automatic checkout system of Digital Image Processing.
At present, adopt Digital Image Processing zip fastener to be carried out the research of defects detection, mainly concentrate on and utilize digital image processing techniques to carry out slide fastener number of teeth context of detection, utilize the technology of histogram equalization, binarization of gray value and centroid algorithm to come the single-edge zipper teeth situation of slide fastener is analyzed and studied.Yet, the detection of the defectives such as, breakage imperfect to the chain band that produces in production run and disappearance and the problem of the differentiation scheme that also is not well solved, the research of zip fastener defective also is in the starting stage.
In sum, the existing detection of adopting digital image processing techniques to finish the slide fastener defective is used, also be in the starting exploratory stage, not yet work out the technical scheme of reasonable, the complete zip fastener defects detection of a cover, mostly can only finish statically a certain starting stage to the detection of specific defects, can not satisfy the demand of the zip fastener defective being carried out complete detection.
Summary of the invention
The object of the invention is to consider zip fastener is carried out the demand that complete edge detects and a kind of detection method of the zip fastener defective based on Digital Image Processing is provided.The method at first drawing of the slide fastener image after gray scale stretches is partly set two analyzed areas, calculate the barycentric coordinates in these two zones, according to two barycentric coordinates original slide fastener image rotation to slide fastener is horizontal, obtain rough chain belt profile by filtering and edge detection algorithm, the rough edge of chain band is carried out the border tracking obtain clearly chain belt edge image; Secondly, by determining edge, the chain band left and right sides and chain band jag initial point position, the zip fastener edge is divided into up and down two analyzed areas, zone and the zip fastener edge image divided are carried out the XOR processing, extract and only comprise respectively, the chain belt edge image of the latter half, according to chain band jag start position setting analyzed area, use image XOR algorithm that the edge image of chain band jag part is extracted from chain band the first half, and after horizontally rotating, respectively drawing part and the jag chain belt edge partly of slide fastener the first half are detected according to the principle of the gray-scale value saltus step of chain belt edge point.At last, use the mirror principle that chain band the latter half edge mirror is arrived the first half, the method at use detection chain band the first half edge is finished the detection to chain band the latter half edge.
The present invention, utilization is by the target localization flip flop equipment, the image acquisition sensor, conveyor drive arrangement, background light source device, feed belt, the pick-up unit that sorting equipment and PC or embedded control system consist of detects, production process completed zipper product to be detected in front road is admitted on the feed belt platform of pick-up unit, under the driving of feed belt drive unit, zipper product is transmitted by feed belt and enters surveyed area, the target localization flip flop equipment carries out target localization to zipper product, if zipper product has entered image acquisition center sensor area of visual field, then trigger the LED strip source of lighting background light source device, and start simultaneously the image acquisition sensor, obtain the apparent zip fastener digital picture of region contour, then target image is transferred to PC or embedded control system, PC or embedded control system carry out filtering to target image, slide fastener horizontally rotates, rim detection and boundary profile extraction process, judge relatively according to result and the discrimination standard processed whether the zip fastener zone exists defective, according to testing result qualified zipper product is sent into down and to be packed together production process, with existing the zipper product of chain band defective to send into the substandard products zone by sorting equipment, finish the zip fastener defects detection.
The present invention, described detection may further comprise the steps:
1) adopt the image acquisition sensor to obtain the slide fastener image, and adopt the image partition method of intensity-based threshold value, by threshold value is set, realize cutting apart the metallic bond tooth of slide fastener image and image background, chain band, remove zip fastener and former collection image background, extract the slide fastener coupling element skeleton;
2) according to the rough position of slide fastener drawing part in image, in this zone, set two analyzed areas, obtain respectively the centre of gravity place of the chain tooth in these two analyzed areas according to centroid algorithm, by two regional barycenter coordinate figures, calculate the angle of two regional barycenter points, take the picture centre coordinate as reference point, make the slide fastener image rotation that is in Difference angles to horizontal level with this angle as the anglec of rotation;
3) postrotational image is carried out mean filter and process, smoothed image, noise reduction disturbs, so that follow-up rim detection operational processes;
4) use based on the optimized Canny edge detection operator extraction of functional zip fastener edge;
5) choose zip fastener edge any point, take this point as starting point, be communicated with edge following algorithm according to four and obtain the zip fastener border, extract clearly chain belt profile image with chain band border, from image background, zip fastener is partly split, remove because image rotating causes image not exclusively to be in the impact of target area;
6) by searching boundary chain belt edge pixel by row from left to right, record finds the ordinate value of pixel at first, assert that this point is chain band leftward position; Utilize same procedure to turn left from the right side and search chain band right positions coordinate figure, the difference of calculating the both sides coordinate figure obtains chain band overall length, if overall length surpasses predetermined threshold value, assert that then the chain band is long, judge that this slide fastener is substandard product, jump out program, carry out the subsequent sort operation; If do not exceed predetermined threshold value, continue the subsequent detection operation;
7) in slide fastener drawing zone, choose a row pixel of entire image, take this row pixel as sweep object, calculate from top to bottom the pixel value gradient of this scan columns, predetermined threshold value, whether the Grad that detects this each pixel of scan columns exceeds threshold value, if exceed threshold value, then assert this pixel generation Gray Level Jump, assert that this point is the zip fastener coboundary, record this ordinate of orthogonal axes, continue to carry out above-mentioned detection downwards, until search the zip fastener lower limb, and record this ordinate of orthogonal axes value;
8) the ordinate of orthogonal axes value according to the upper lower limb of zip fastener calculates the up and down medium line at two edges of zip fastener, take this line as center line, zip fastener is divided into up and down two parts, and take medium line as benchmark, search from right to left the Gray Level Jump point, if find, assert that then this Gray Level Jump point is slide fastener jag starting point, take this horizontal ordinate as benchmark, slide fastener the first half is divided into zip fastener drawing part and two analyzed areas of zip fastener jag part;
9) use image XOR algorithm, zip fastener the latter half and whole image are carried out image XOR algorithm process, obtain only to comprise zip fastener the first half image;
10) the chain band left side abscissa value by recording before, according to detecting surplus, obtain chain band left side and detect the abscissa value of reference position, the horizontal ordinate of slide fastener jag starting point is stop value, distance take 10 pixels arranges the detection scan columns as increment circulates from left to right, search from top to bottom the Gray Level Jump situation of chain belt edge point along scan columns, the ordinate value of the marginal point during the record Gray Level Jump, the absolute value of the difference of the ordinate of two neighboring edge check points of calculating, according to the accuracy of detection predetermined threshold value, judge that whether the absolute value of this difference is greater than threshold value, if greater than threshold value, judge that then there is defective in the chain belt edge between these two adjacent check points, jump out program, otherwise enter the detection operation of zip fastener jag part;
11) use image XOR algorithm, the first half of zip fastener drawing and the first half image of chain band are carried out the processing of image XOR, obtain only to comprise the first half image of slide fastener jag;
12) according to the rough position of slide fastener jag part in image, optional two row at the two ends of zip fastener jag part, be listed as searching benchmark with above-mentioned two respectively, search from the top down first Gray Level Jump point, assert that then this point is chain band coboundary if find, and record this point coordinate value, coordinate according to these two Gray Level Jump points, take left side Gray Level Jump point as reference point, the angle between calculating at 2 is rotated chain band jag the first half to horizontal level take this angle as the anglec of rotation;
13) find two ends, the left and right sides abscissa value of chain band jag the first half, with the abscissa value in chain band left side as detecting initial value, the right side horizontal ordinate of chain band is stop value, distance take 10 pixels arranges the detection scan columns as increment circulates from left to right, search from top to bottom the Gray Level Jump situation of chain belt edge point along scan columns, the ordinate value of the marginal point during the record Gray Level Jump, the absolute value of the difference of the ordinate of two neighboring edge check points of calculating, according to the accuracy of detection predetermined threshold value, judge that whether the absolute value of this difference is greater than threshold value, if greater than threshold value, judge that then there is defective in the chain belt edge between these two adjacent check points, jump out program, otherwise enter the detection operation of zip fastener the latter half;
14) take zip fastener up and down the medium line at two edges as the mirror axle, with chain band the latter half mirror to medium line top, obtain only existing the chain band the latter half image that is in the medium line top, jump into 10) to 13) program, if detect chain band defective then jump out program, if do not detect chain band defective then termination routine, show that this zip fastener is intact, be specification product;
15) whether there is defective according to detecting zip fastener, exports different signals.
The present invention can determine according to the method for center of gravity calculation the position of two points on the slide fastener coupling element will be in the slide fastener image rotation of different angles take this angle of 2 as the anglec of rotation to horizontal level; Determine two points on the chain band jag part edge with the principle of searching edge gray scale generation saltus step, so with this partial rotation to horizontal level.
The present invention can obtain first the coarse boundaries of chain band by the mode of rim detection, the mode of using again the border to follow the tracks of is obtained the accurate profile in edge of chain band, extracts clearly chain belt edge from the travelling belt background image, the outstanding target that detects.
The present invention, the accessible region territory is divided with whole stretch-draw chain image and is carried out the XOR processing, being divided into whole zip fastener up and down, two parts detect, use the mode of mirror that the edge mirror of the latter half is arrived the top, the program module of use detection chain band the first half is finished the detection to chain band the latter half, saves program step and size.
The present invention, the accessible region territory is divided with whole stretch-draw chain image and is carried out the XOR processing, be truncated to the chain band image that only contains respectively slide fastener drawing part and jag part, after the chain band of slide fastener jag part horizontally rotated, Grad according to chain belt edge place vertical direction is maximum, judge whether this some the gray-scale value saltus step occurs, if saltus step occur then think that this point is marginal point, determine by the difference of the ordinate of orthogonal axes value of two adjacent spaces marginal points relatively whether the chain belt edge exists defective, finish respectively drawing part and jag detection partly to zip fastener.
The present invention's advantage compared with prior art is:
1, zip fastener image rotation to horizontal level is processed, so that chain band image ratio is more directly perceived, and the chain belt edge all is in the same horizontal line, with the slide fastener image normalization, reduces the complexity of trace routine;
2, carry out the border by the rough profile after the chain belt edge is detected and follow the tracks of, extract the clear contour images of chain belt edge, reduce the difficulty of detection algorithm, significantly improve speed and the accuracy that detects;
3, the point that Gray Level Jump occurs on the default scan columns in the chain band image is searched in employing, determines the coordinate of marginal point, and then realizes the whether judgement of chain band defective, has reduced the complexity of detection algorithm, has greatly improved detection efficiency;
4, use the mirror algorithm that chain band the latter half edge mirror is arrived the top, use the program module realization at detection chain band the first half edge to the detection at chain band the latter half edge, it is big or small to reduce the program committed memory, improves the efficient of detection
5, whole detection method has good versatility and extendability, and the zip fastener edge defect that can apply to easily other patterns detects.
Description of drawings
Figure 1 shows that the block diagram of pick-up unit of the present invention.Among the figure, 1, zipper product to be detected; 2, target localization flip flop equipment; 3, image acquisition sensor; 4, conveyor drive arrangement; 5, background light source device; 6, feed belt; 7, sorting equipment; 8, PC or embedded control system.
Figure 2 shows that zip fastener defects detection process flow diagram; Testing process is: at first intercept the slide fastener image from the image acquisition sensor, the slide fastener that will be in different angles rotates to horizontal level, chain band image after stretching through gray scale is carried out the Canny rim detection, by border track and extract edge contour figure clearly, then carrying out fringe region cuts apart, obtain only comprising respectively the chain belt edge of chain band the first half, the edge of chain band the first half is divided into drawing part and jag part, splitting fork divides and horizontally rotates, respectively drawing part and the jag of chain belt edge are partly carried out the marginal point circulation searching according to the Gray Level Jump principle, judge by the ordinate that calculates adjacent two edges whether the edge exists defective, last output detections result prepares for detecting next zipper product time-delay.
Fig. 3 is slide fastener image level rotational programme figure.In the drawings θ for the zipper metal chain tooth drawing that obtains by the regional barycenter algorithm partly go up about the angle of two points, take left-hand point A as reference point, right-hand point B is last point, utilize the angle calculation algorithm to calculate angle between 2, calculate the anglec of rotation in the size according to angle, the anglec of rotation on the occasion of representative image then take certain point as rotation center by counterclockwise rotation, otherwise, the anglec of rotation be negative value then image be rotated in the direction of the clock.
Fig. 4 is the overhaul flow chart at zip fastener edge.In the drawings, X1 and X2 are respectively the train value at the left and right frontier point place of edge image.The value of scan columns increases with the regulation increment, when along current scan columns scan chain band image, saltus step according to gray-scale value judges whether that current point is the chain belt edge, if the gray-scale value saltus step occur then record the ordinate of orthogonal axes value of this point, calculate the difference of adjacent two marginal point ordinates, whether the absolute value of judging difference exceeds predetermined threshold value is judged whether the chain band exists defective.
The chain belt edge image of Fig. 5 for obtaining by the Canny operator.
Fig. 6 is for using the border to follow the tracks of the clear figure of chain belt edge that extracts from Fig. 5.
Fig. 7 only comprises chain band the first half outline map for what use that the intercepting of XOR algorithm obtains.
Fig. 8, Fig. 9 are zip fastener the first half detection algorithm synoptic diagram.In the drawings, detect the ordinate value of left and right two marginal points of chain band that are horizontal, shown in red dotted line among the figure; The detection side is to for scanning from left to right, shown in blue arrow among the figure; The direction of scanning is for carrying out from top to bottom along current scan columns (among the figure shown in the yellow solid line), shown in figure Green arrow; For example: among Fig. 8, the ordinate of left side edge point adds that detecting surplus is starting point, the ordinate value of chain band jag part starting point is stop value, carry out scan round, the step-length that scan columns increases is decided according to accuracy of detection, if the regulation defect area is more than or equal to 1 square centimeter, then the regulation step-length is 15 pixel distances, until the value of scan columns is more than or equal to final value, otherwise program will circulate always.
Embodiment
With reference to Fig. 1, the pick-up unit that a kind of detection method of the zip fastener defective based on Digital Image Processing realizes is made of target localization flip flop equipment 2, image acquisition sensor 3, conveyor drive arrangement 4, background light source device 5, feed belt 6, sorting equipment 7 and PC or embedded control system 8.Before carrying out the zipper product defects detection, the slide fastener production line has been finished the processing to zipper product 1 to be detected, slide fastener is admitted on the feed belt platform of pick-up unit, under the driving of feed belt drive unit 4, zipper product is transmitted by feed belt 6 and enters surveyed area, 2 pairs of zipper products of target localization flip flop equipment carry out target localization, if zipper product has entered image acquisition center sensor area of visual field, then trigger the LED strip source of background light source device 5, and start simultaneously image acquisition sensor 3, obtain the apparent zip fastener digital picture of region contour, then target image is transferred to PC or embedded control system 8.The processing such as 8 pairs of target images of PC or embedded control system carry out that filtering, slide fastener horizontally rotate, rim detection and boundary profile extraction judge according to result and the discrimination standard processed whether the zip fastener zone exists incomplete defective.According to testing result qualified zipper product is sent into down and to be packed together production process, with existing the zipper product of chain band defective to send into the substandard products zone by sorting equipment (7), finish the zip fastener defects detection.
The implementation step of the detection method of a kind of zip fastener defective based on Digital Image Processing of the present invention is as follows:
The size of the slide fastener that 1, provides according to producer and the color of chain band, select suitable light source type and travelling belt color, regulate suitable field range, regulate focal length, light-inletting quantity and the time shutter of image acquisition sensor, make the image acquisition sensor obtain clearly still image;
2, when the driven travelling belt of slide fastener is sent to image acquisition center sensor field range, by triggering the image acquisition sensor, obtain continuous slide fastener image;
3, in two analyzed areas that comprise slide fastener drawing part chain tooth, carry out regional barycenter and calculate, obtain the coordinate of two points (two centers of gravity of two analyzed areas) on the chain tooth (drawing part), calculate the angle θ of slide fastener coupling element skeleton and transverse axis
1, image is horizontally rotated angle θ
1On the occasion of representative image by being rotated along counterclockwise, otherwise angle is that negative value is then by being rotated along clockwise direction;
4, use gray scale stretching algorithm, the chain band image that obtains is carried out gray scale stretch, obtain the obvious zip fastener image of intensity contrast;
5, use based on the optimized Canny edge detection operator extraction of functional zip fastener edge, use edge following algorithm to propose chain belt edge profile;
6, according to slide fastener horizontal center line position the chain belt edge is divided into up and down two parts, at first chain band the first half is processed, according to chain band jag initial point position upper part edge is divided into drawing part and jag part;
7, search the marginal point of the left and right sides of chain band view picture zip fastener image, the difference of calculating the coordinate figure of both sides of the edge point obtains chain band overall length, if overall length surpasses predetermined threshold value, assert that then the chain band is long, judge that this slide fastener is substandard product, jump out program, carry out the subsequent sort operation; If do not exceed predetermined threshold value, continue the subsequent detection operation;
8, the principle of gray-scale value saltus step can occur according to chain belt edge point, detect the marginal point of certain intervals, the ordinate of the marginal point that record detects, the absolute value of the difference of the ordinate of two neighboring edge check points of calculating according to the accuracy of detection predetermined threshold value, judges that whether the absolute value of this difference is greater than threshold value, if greater than threshold value, judge that then there is defective in the chain belt edge between these two adjacent check points, jump out program, otherwise enter the detection operation of zip fastener jag part;
9, search the point of two certain intervals on the chain band jag part edge, and record 2 coordinate figure, take left-hand point as reference point, the angle between calculating at 2 is rotated chain band jag the first half to horizontal level take this angle as the anglec of rotation;
10, repeat the algoritic module in the 8th step, the absolute value of the difference of the ordinate of two neighboring edge check points of calculating, according to the accuracy of detection predetermined threshold value, judge that whether the absolute value of this difference is greater than threshold value, if greater than threshold value, judge that then there is defective in the chain belt edge between these two adjacent check points, jump out program, otherwise enter the detection operation of zip fastener the latter half;
11, take zip fastener up and down the medium line at two edges as the mirror axle, with chain band the latter half mirror to medium line top, obtain only existing the chain band the latter half image that is in the medium line top, jump into the program in the 8th step to the 10th step, if detect chain band defective then jump out program, if do not detect chain band defective then termination routine, show that this zip fastener is intact, be specification product;
12, whether have defective according to detecting zip fastener, export different signals, time-delay is for the detection of next zipper product is prepared.
Claims (5)
1. zip fastener defect inspection method based on Digital Image Processing, it is characterized in that: utilize by target localization flip flop equipment (2), image acquisition sensor (3), feed belt drive unit (4), background light source device (5), feed belt (6), the pick-up unit that sorting equipment (7) and PC or embedded control system (8) consist of detects, the completed zipper product to be detected of front road production process (1) is admitted on feed belt (6) platform of pick-up unit, under the driving of feed belt drive unit (4), zipper product is transmitted by feed belt (6) and enters surveyed area, target localization flip flop equipment (2) carries out target localization to zipper product, if zipper product has entered image acquisition center sensor area of visual field, then trigger the LED strip source of lighting background light source device (5), and start simultaneously image acquisition sensor (3), obtain the apparent zip fastener digital picture of region contour, then target image is transferred to PC or embedded control system (8), PC or embedded control system (8) carry out filtering to target image, slide fastener horizontally rotates, rim detection and boundary profile extraction process, judge relatively according to result and the discrimination standard processed whether the zip fastener zone exists defective, according to testing result qualified zipper product is sent into down and to be packed together production process, send into the substandard products zone with there being the zipper product of chain band defective by sorting equipment (7), finish the zip fastener defects detection, described detection may further comprise the steps:
1) adopt the image acquisition sensor to obtain the slide fastener image, and adopt the image partition method of intensity-based threshold value, by threshold value is set, realize cutting apart the metallic bond tooth of slide fastener image and image background, chain band, remove zip fastener and former collection image background, extract the slide fastener coupling element skeleton;
2) according to the rough position of slide fastener drawing part in image, in this zone, set two analyzed areas, obtain respectively the centre of gravity place of the chain tooth in these two analyzed areas according to centroid algorithm, by two regional barycenter coordinate figures, calculate the angle of two regional barycenter points, take the picture centre coordinate as reference point, make the slide fastener image rotation that is in Difference angles to horizontal level with this angle as the anglec of rotation;
3) postrotational image is carried out mean filter and process, smoothed image, noise reduction disturbs, so that follow-up rim detection operational processes;
4) use based on the optimized Canny edge detection operator extraction of functional zip fastener edge;
5) choose zip fastener edge any point, take this point as starting point, be communicated with edge following algorithm according to four and obtain the zip fastener border, extract clearly chain belt profile image with chain band border, from image background, zip fastener is partly split, remove because image rotating causes image not exclusively to be in the impact of target area;
6) by searching boundary chain belt edge pixel by row from left to right, record finds the ordinate value of pixel at first, assert that this point is chain band leftward position; Utilize same procedure to turn left from the right side and search chain band right positions coordinate figure, the difference of calculating the both sides coordinate figure obtains chain band overall length, if overall length surpasses predetermined threshold value, assert that then the chain band is long, judge that this slide fastener is substandard product, jump out program, carry out the subsequent sort operation; If do not exceed predetermined threshold value, continue the subsequent detection operation;
7) in slide fastener drawing zone, choose a row pixel of entire image, take this row pixel as sweep object, calculate from top to bottom the pixel value gradient of this row pixel, predetermined threshold value, whether the Grad that detects this each pixel of row pixel exceeds threshold value, if exceed threshold value, then assert this pixel generation Gray Level Jump, assert that this point is the zip fastener coboundary, record this ordinate of orthogonal axes, continue to carry out above-mentioned detection downwards, until search the zip fastener lower limb, and record this ordinate of orthogonal axes value;
8) the ordinate of orthogonal axes value according to the upper lower limb of zip fastener calculates the up and down medium line at two edges of zip fastener, take this line as center line, zip fastener is divided into up and down two parts, and take medium line as benchmark, search from right to left the Gray Level Jump point, if find, assert that then this Gray Level Jump point is slide fastener jag starting point, take this horizontal ordinate as benchmark, zip fastener the first half is divided into zip fastener drawing part and two analyzed areas of zip fastener jag part;
9) use image XOR algorithm, zip fastener the latter half and whole image are carried out image XOR algorithm process, obtain only to comprise zip fastener the first half image;
10) the chain band left side abscissa value by recording before, according to detecting surplus, obtain chain band left side and detect the abscissa value of reference position, the horizontal ordinate of slide fastener jag starting point is stop value, distance take 10 pixels arranges the detection scan columns as increment circulates from left to right, search from top to bottom the Gray Level Jump situation of chain belt edge point along scan columns, the ordinate value of the marginal point when Gray Level Jump occurs record, the absolute value of the difference of the ordinate of two neighboring edge check points of calculating, according to the accuracy of detection predetermined threshold value, judge that whether the absolute value of this difference is greater than threshold value, if greater than threshold value, judge that then there is defective in the chain belt edge between these two neighboring edge check points, jump out program, otherwise enter the detection operation of zip fastener jag part;
11) use image XOR algorithm, the first half of zip fastener drawing and the first half image of chain band are carried out the processing of image XOR, obtain only to comprise the first half image of zip fastener jag;
12) according to the rough position of zip fastener jag part in image, optional two row at the two ends of zip fastener jag part, be listed as searching benchmark with above-mentioned two respectively, search from the top down first Gray Level Jump point, assert that then this point is chain band coboundary if find, and record this point coordinate value, coordinate according to these two Gray Level Jump points, take left side Gray Level Jump point as reference point, angle between calculating at 2 is rotated chain band jag the first half to horizontal level take this angle as the anglec of rotation;
13) find two ends, the left and right sides abscissa value of chain band jag the first half, with the abscissa value in chain band left side as detecting initial value, the right side horizontal ordinate of chain band is stop value, distance take 10 pixels arranges detection row datum line as increment circulates from left to right, search from top to bottom the Gray Level Jump situation of chain belt edge point along the row datum line, the ordinate value of the marginal point during the record Gray Level Jump, the absolute value of the difference of the ordinate of two adjacent check points of calculating, according to the accuracy of detection predetermined threshold value, judge that whether the absolute value of this difference is greater than threshold value, if greater than threshold value, judge that then there is defective in the chain belt edge between these two adjacent check points, jump out program, otherwise enter the detection operation of zip fastener the latter half;
14) take zip fastener up and down the medium line at two edges as the mirror axle, with chain band the latter half mirror to medium line top, obtain only existing the chain band the latter half image that is in the medium line top, jump into 10) to 13) program, if detect chain band defective then jump out program, if do not detect chain band defective then termination routine, show that this zip fastener is intact, be specification product;
15) whether there is defective according to detecting zip fastener, exports different signals.
2. detection method according to claim 1 is characterized in that: determine the position of two points on the slide fastener coupling element will be in the slide fastener image rotation of different angles take this angle of 2 as the anglec of rotation to horizontal level according to the method for center of gravity calculation; Determine two points on the chain band jag part edge with the principle of searching edge gray scale generation saltus step, so with this partial rotation to horizontal level.
3. detection method according to claim 1, it is characterized in that: the coarse boundaries that obtains first the chain band by the mode of rim detection, the mode of using again the border to follow the tracks of is obtained the accurate profile in edge of chain band, extracts clearly chain belt edge from the travelling belt background image, the outstanding target that detects.
4. detection method according to claim 1, it is characterized in that: divide with whole stretch-draw chain image by the zone and carry out the XOR processing, being divided into whole zip fastener up and down, two parts detect, use the mode of mirror that the edge mirror of the latter half is arrived the top, the program module of use detection chain band the first half is finished the detection to chain band the latter half, saves program step and size.
5. detection method according to claim 1, it is characterized in that: divide with whole stretch-draw chain image by the zone and carry out the XOR processing, be truncated to the chain band image that only contains respectively slide fastener drawing part and jag part, after the chain band of slide fastener jag part horizontally rotated, Grad according to chain belt edge place vertical direction is maximum, judge whether this some the gray-scale value saltus step occurs, if saltus step occur then think that this point is marginal point, determine by the difference of the ordinate of orthogonal axes value of two adjacent spaces marginal points relatively whether the chain belt edge exists defective, finish respectively drawing part and jag detection partly to zip fastener.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110403923 CN102495069B (en) | 2011-12-07 | 2011-12-07 | Method for detecting defects of chain belts of zipper on basis of digital image processing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201110403923 CN102495069B (en) | 2011-12-07 | 2011-12-07 | Method for detecting defects of chain belts of zipper on basis of digital image processing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102495069A CN102495069A (en) | 2012-06-13 |
| CN102495069B true CN102495069B (en) | 2013-03-20 |
Family
ID=46186910
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201110403923 Expired - Fee Related CN102495069B (en) | 2011-12-07 | 2011-12-07 | Method for detecting defects of chain belts of zipper on basis of digital image processing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102495069B (en) |
Families Citing this family (33)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102721703A (en) * | 2012-07-03 | 2012-10-10 | 苏州天准精密技术有限公司 | Illuminating device and imaging method for online detection |
| JP6088855B2 (en) * | 2013-02-28 | 2017-03-01 | モレックス エルエルシー | Appearance inspection apparatus and appearance inspection method |
| CN103837542B (en) * | 2014-03-25 | 2015-11-18 | 福建浔兴拉链科技股份有限公司 | A kind of metallic bond sound of baby talk point defect pick-up unit and detection method thereof |
| CN106461566B (en) * | 2014-06-24 | 2019-05-17 | Ykk株式会社 | The inspection condition enactment method and slide fastener coupling element inspection method of slide fastener coupling element |
| JP6486050B2 (en) * | 2014-09-29 | 2019-03-20 | 株式会社Ｓｃｒｅｅｎホールディングス | Inspection apparatus and inspection method |
| CN104677272B (en) * | 2014-12-30 | 2017-10-13 | 东莞市美厚塑磁有限公司 | A kind of permanent magnetism chip detection method and implementation this method detection device |
| CN105057229B (en) * | 2015-08-24 | 2017-08-01 | 东莞劲胜精密组件股份有限公司 | A kind of sprayed surface defect inspection machine |
| CN105335963B (en) * | 2015-09-24 | 2018-07-03 | 凌云光技术集团有限责任公司 | A kind of edge defect detection algorithm and device |
| CN106022331B (en) * | 2016-05-23 | 2017-06-27 | 山东科技大学 | The tea packaging sky bag method of inspection based on machine vision |
| CN106442556A (en) * | 2016-11-16 | 2017-02-22 | 哈尔滨理工大学 | Device and method for detecting surface defects of perforated plate workpiece |
| CN106890798B (en) * | 2017-01-11 | 2019-04-09 | 广西师范大学 | A kind of bamboo and wood item detects separation system and method automatically |
| CN106706656B (en) * | 2017-03-14 | 2024-06-14 | 杭州汇萃智能科技有限公司 | Zipper detection device and method based on machine vision |
| CN107067421B (en) | 2017-05-12 | 2020-02-28 | 京东方科技集团股份有限公司 | Method, device and system for detecting residual materials of substrate |
| CN110276229A (en) * | 2018-03-14 | 2019-09-24 | 京东方科技集团股份有限公司 | Target object regional center localization method and device |
| CN108413844B (en) * | 2018-03-30 | 2020-10-16 | 福建晋江浔兴拉链科技有限公司 | Nylon zipper tooth pitch measuring tool and tooth pitch measuring method |
| CN108627520A (en) * | 2018-05-25 | 2018-10-09 | 北京英斯派克科技有限公司 | A kind of on-line detecting system and method for heterogeneous solid material presentation quality |
| CN108426892A (en) * | 2018-06-12 | 2018-08-21 | 广东理工学院 | A kind of machine vision device of detection household plate surface defect |
| CN108802052A (en) * | 2018-07-23 | 2018-11-13 | 西安工程大学 | A kind of detecting system and its detection method about slide fastener defect |
| CN109030741A (en) * | 2018-08-01 | 2018-12-18 | 长兴曼尔申机械科技有限公司 | A kind of metal tooth defect inspection method |
| CN109712114A (en) * | 2018-11-30 | 2019-05-03 | 无锡维胜威信息科技有限公司 | A kind of system and its detection method applied to zipper defects detection |
| CN109916914B (en) * | 2019-04-10 | 2021-07-02 | 清华大学深圳研究生院 | Product defect detection method and device |
| CN110334692B (en) * | 2019-07-17 | 2021-07-23 | 河南科技大学 | Blind road identification method based on image processing |
| CN111047592B (en) * | 2020-03-13 | 2020-06-05 | 武汉理工大学 | Monitoring system and image communication processing method thereof |
| CN111981990B (en) * | 2020-08-03 | 2022-02-01 | 无锡鼎茂机械制造有限公司 | Machine vision-based self-sealing zipper quality online detection method |
| CN111929315A (en) * | 2020-09-04 | 2020-11-13 | 北京远大宏略科技股份有限公司 | Wound cable skin detection device and detection method thereof |
| CN113706525B (en) * | 2021-10-20 | 2022-02-15 | 武汉飞恩微电子有限公司 | Intelligent printed circuit board patch defect identification method based on image processing |
| CN115222709B (en) * | 2022-07-26 | 2022-12-13 | 南通市通州区锦都拉链有限公司 | Zipper cloth belt defect detection method based on artificial intelligence |
| CN115201207B (en) * | 2022-07-28 | 2024-10-01 | 浙江伟星实业发展股份有限公司 | Zipper defect detection method, device, equipment and medium |
| CN115307817A (en) * | 2022-10-12 | 2022-11-08 | 南通元景电气科技有限公司 | Pressure sensor batch test system and method thereof |
| CN117770585A (en) * | 2022-11-11 | 2024-03-29 | 张化 | Zipper shaping device capable of being used for detecting product quality |
| CN116000747A (en) * | 2022-11-28 | 2023-04-25 | 中国汽车技术研究中心有限公司 | Polishing and bonding system and method for skin zipper of leg of dummy |
| CN116580031B (en) * | 2023-07-13 | 2023-09-29 | 山东勇嘉包装科技股份有限公司 | Solvent-free composite printing anomaly detection method |
| CN117456519B (en) * | 2023-12-22 | 2024-03-12 | 江西科益高新技术有限公司 | Simulation photoelectric target device for industrial control safety and simulation method thereof |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1475795A (en) * | 2002-08-12 | 2004-02-18 | 广西师范大学 | Glass Bottle and can detecting method and detecting device |
| CN2795852Y (en) * | 2005-05-12 | 2006-07-12 | 佛山市德科机电设备有限公司 | Automatic detector for glazed tile surface property |
| CN101556239A (en) * | 2008-04-07 | 2009-10-14 | 福建浔兴拉链科技股份有限公司 | Zipper tooth detector, zipper tooth detection method and zipper injection machine using zipper tooth detector |
| CN201909762U (en) * | 2010-12-31 | 2011-07-27 | 广东辉丰科技股份有限公司 | Device for automatically detecting defect of metal zipper |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008190867A (en) * | 2007-01-31 | 2008-08-21 | Nakayama:Kk | Nodular graphite cast iron quality measuring device |
-
2011
- 2011-12-07 CN CN 201110403923 patent/CN102495069B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1475795A (en) * | 2002-08-12 | 2004-02-18 | 广西师范大学 | Glass Bottle and can detecting method and detecting device |
| CN2795852Y (en) * | 2005-05-12 | 2006-07-12 | 佛山市德科机电设备有限公司 | Automatic detector for glazed tile surface property |
| CN101556239A (en) * | 2008-04-07 | 2009-10-14 | 福建浔兴拉链科技股份有限公司 | Zipper tooth detector, zipper tooth detection method and zipper injection machine using zipper tooth detector |
| CN201909762U (en) * | 2010-12-31 | 2011-07-27 | 广东辉丰科技股份有限公司 | Device for automatically detecting defect of metal zipper |
Non-Patent Citations (1)
| Title |
|---|
| JP特开2008-190867A 2008.08.21 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102495069A (en) | 2012-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102495069B (en) | Method for detecting defects of chain belts of zipper on basis of digital image processing | |
| CN102495076B (en) | Method for detecting defects of metal zipper teeth of zipper on basis of machine vision | |
| CN109115785B (en) | Casting polishing quality detection method and device and use method thereof | |
| CN102636490B (en) | Method for detecting surface defects of dustproof cover of bearing based on machine vision | |
| CN105388162B (en) | Raw material silicon chip surface scratch detection method based on machine vision | |
| CN102288613B (en) | Surface defect detecting method for fusing grey and depth information | |
| CN109724984A (en) | A kind of defects detection identification device and method based on deep learning algorithm | |
| CN104537651B (en) | Proportion detecting method and system for cracks in road surface image | |
| CN110111303A (en) | A kind of large-scale carrier strip tearing intelligent fault detection method based on dynamic image | |
| CN110146516B (en) | Fruit grading device based on orthogonal binocular machine vision | |
| CN104574389A (en) | Battery piece chromatism selection control method based on color machine vision | |
| CN103454285A (en) | Transmission chain quality detection system based on machine vision | |
| CN107876425A (en) | A kind of bearing defect detecting system device of view-based access control model | |
| CN104697998A (en) | Inclusion detection in polished gemstones | |
| CN104483320B (en) | Digitized defect detection device and detection method of industrial denitration catalyst | |
| CN111266315A (en) | Ore material online sorting system and method based on visual analysis | |
| CN106295705B (en) | A kind of more color materials under movement background screen number system | |
| CN109087286A (en) | A kind of detection method and application based on Computer Image Processing and pattern-recognition | |
| CN106824806A (en) | The detecting system and detection method of the low module plastic gear based on machine vision | |
| CN106018422A (en) | Matching-based visual outline defect inspection system and method for specially-shaped stamping parts | |
| CN104634791A (en) | Omnidirectional and visual on-line detection system of surface defects of strip-shaped workpiece | |
| CN102175692A (en) | System and method for detecting defects of fabric gray cloth quickly | |
| CN110108712A (en) | Multifunctional visual sense defect detecting system | |
| CN106780526A (en) | A kind of ferrite wafer alligatoring recognition methods | |
| CN102890093B (en) | Machine vision detecting device of surface quality of master cylinder piston of automobile brake |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130320 Termination date: 20181207 |