CN1710606A - Non-contact vision paper-counting method and machine thereof - Google Patents
Non-contact vision paper-counting method and machine thereof Download PDFInfo
- Publication number
- CN1710606A CN1710606A CN 200410049579 CN200410049579A CN1710606A CN 1710606 A CN1710606 A CN 1710606A CN 200410049579 CN200410049579 CN 200410049579 CN 200410049579 A CN200410049579 A CN 200410049579A CN 1710606 A CN1710606 A CN 1710606A
- Authority
- CN
- China
- Prior art keywords
- paper
- counting
- image
- striped
- desire
- 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
Images
Landscapes
- Image Processing (AREA)
Abstract
The method forms stripe image for side of paper to be counted, and each piece of stripe is corresponding to a piece of stripe. Then, data process is carried out for the stripe image. The disclosed counting device includes following parts: mechanical platform of image collection in use for placing paper to be counted; a camera installed on side of the said platform is perpendicular to the side of the paper to be counted; illuminated light sources at left and right sides of camera are installed near to the paper to be counted; a photo sensor is in lower part inside the platform; a computer includes a image collection card connected to the computer, and a I/O signal interface card for receiving signal from the photo sensor. When paper to be counted is moved to proper position, the camera takes an image. When paper to be counted is continuous moved to second proper position, the photo sensor sends signal to inform computer to start controlling image collection and detection.
Description
Technical field
The present invention relates to a kind of contactless vision paper counting method.
The invention still further relates to a kind of counting machine that is used to realize said method.
Background technology
At present, paper counting all is to lean on the mode of machine operation and sensor counting to realize, although physical construction is not quite similar, all needs to contact with paper.In order to reach counting rate faster, the speed of machine operation is all very high, and there are several major defects in this type of mechanical system counting.The one, need operating personnel directly to contact the plant equipment that runs up, or paper is placed into and the counting equipment position contacting by operating personnel, there is carelessness can cause industrial injury slightly; When next was counting, plant equipment can produce bigger noise under situation about running up, have a strong impact on the healthy of operating personnel and other staff on the scene; In addition, owing to need artificial the participation, and be to count one by one, under the situation of mechanical high-speed running, it is limited that counting rate remains.
Summary of the invention
The object of the present invention is to provide a kind of contactless vision paper counting method, this method is by the analyzing and processing to the paper edge image gathered, calculate actual number, fundamentally overcome the shortcoming of contact counting, counting rate and precision are greatly improved.
Another object of the present invention is to provide a kind of counting machine that is used to realize said method.
Contactless vision paper counting method provided by the invention mainly is that the paper side with desire counting forms the image striped, and the corresponding a piece of paper of each striped carries out the data processing of following steps then to the image striped:
1. stripe segmentation: the computed image segmentation threshold specially changes image into bianry image;
2. striped search: the position of all stripeds in sequential search and the definite image;
3. striped refinement: all stripeds are carried out thinning processing, remove noise;
4. fringe count:, calculate the paper number by calculating the number of striped.
The present invention is used to realize that the counting machine of said method mainly comprises an image acquisition mechanical platform that is used to put desire counting paper; At least one video camera is installed in a side of image acquisition mechanical platform, is vertical with a side of this desire counting paper; The video camera left and right sides is equipped with lighting source near the paper place; A photoelectric sensor is installed in the below, inside of mechanical platform; One computing machine is equipped with image pick-up card and I/O signal interface card; Image pick-up card is connected with video camera, and the I/O signal interface card receives the signal of photoelectric sensor; By above-mentioned counting machine, desire counting paper is ingested image when moving to camera position earlier, continues then to move to the photoelectric sensor position, and photoelectric sensor sends signal, and the notice computing machine begins to control image acquisition and detection.
Description of drawings
Fig. 1 is an image acquisition part synoptic diagram of the present invention.
Fig. 2 is the mechanical platform surface structure synoptic diagram among Fig. 1.
Fig. 3 A is the original image that image acquisition of the present invention is partly gathered.
Fig. 3 B and Fig. 3 C are original image with Fig. 3 A stripe patterns after via data processing of the present invention.
Fig. 4 is a counting machine workflow synoptic diagram of the present invention.
Fig. 5 is a counting algorithm schematic flow sheet of the present invention.
Fig. 6 is the display surface synoptic diagram of counting machine of the present invention.
Embodiment
Below in conjunction with accompanying drawing the present invention is described in detail.
See also Fig. 1 and Fig. 2, be used for paper is carried out the synoptic diagram of noncontact counting machine for the present invention.
One its at least one side is equipped with a video camera 5 in order to put the image acquisition mechanical platform 2 of desire counting paper 7, and the camera lens of this video camera 5 is vertically mounted on the side of paper 7.Can also be furnished with enlarging lens 4 on this video camera 5.For the accuracy that guarantees to count, the present invention respectively installs a video camera 5 in the both sides of mechanical platform 2.Because paper is thinner, differentiate every paper exactly, image acquisition resolution should be able to be told the thickness of every paper.Suppose that every group of paper mostly is 500 most, the thickness maximum is about 50mm, considers the problem on deformation of video camera edge image, and selecting resolution for use is 7500 the black and white linear array CCD camera and the enlarging lens 4 of low distortion.If 5000 picture elements in the middle of getting are as effective image, the acquisition resolution of system can reach 0.01mm.Guaranteeing to have under the situation of enough enlargement factors, can collect the stripe pattern on trimming.
The left and right sides of video camera 5 is equipped with lighting source 3 near the paper place, and lighting source of the present invention requires evenly, and keeps stable and not reflective, and what the present invention adopted is bar shaped light emitting diode (LED) lighting source.
A photoelectric sensor 6 is installed below the inside of mechanical platform 2, and its effect is when desire counting paper 7 arrives the detection position, sends the signal of beginning image acquisition and detection.Also have the gearing that desire counting paper 7 is moved on this mechanical platform 2, because this gearing has been known technology, the present invention seldom is described this.
One computing machine 1 is equipped with image pick-up card and I/O signal interface card.The I/O signal interface card receives the signal of photoelectric sensor 6; Image pick-up card is connected with video camera 5, and the view data that video camera 5 is collected is sent in the computing machine 1.
Precision and consistance in order to ensure image acquisition, desire counting paper is placed on the mechanical platform, video camera and lighting source can move reposefully by the speed of setting relatively, and can provide the trigger pip (delivering to computing machine 1 by photoelectric sensor 6) of control image acquisition.Two video cameras 5 and two groups of LED bar shaped lighting sources 3 are installed in mechanical platform 2 both sides symmetrically, and the distance between the paper 7 that they and desire are counted can be regulated.
Its principle of work of counting machine of the present invention is please in conjunction with Fig. 2, desire counting paper 7 is placed in initial position a place, on mechanical platform 2, press the direction motion shown in the arrow, move forward, and make the paper alignment owing to the guide effect meeting enters among the track b identical with its width paper; When paper continues to move to c place, detection position, promptly enter the position of video camera 5 axis correspondences, and be ingested image; When paper continues to move to d place, position, owing to being blocked, photoelectric sensor 6 sends signal, and notice computing machine 1 begins to control image acquisition and detection.
Because desire counting paper is to pass through photoelectric sensor again through video camera earlier, in the time of can guaranteeing that computing machine begins image acquisition like this, paper has arrived the detection position certainly.
For the accuracy that guarantees to count, the video camera of both sides is respectively gathered piece image, and gets a plurality of positions and count simultaneously in every width of cloth image, the count results of each locational count results and both sides is compared mutually the accuracy of checking counting.
The paper side edge strip line that is absorbed by said apparatus can see from Fig. 3 A as shown in Figure 3A, and one group of paper can present stripe pattern side by side, the corresponding a piece of paper of each striped in image.By analysis, just can calculate actual paper number to these stripeds.
According to the principle of work of above counting machine of the present invention, be readily appreciated that the structure of mechanical platform 2 can change as required, also number system can be directly installed on the streamline, as long as reach the requirement of image capturing system.
In addition, when counting machine of the present invention needing to occur operating personnel to intervene when workflow, send signal by computer starting warning device 8.
It is to be finished by computing machine that the present invention carries out data processing to the image striped that absorbs, and its main processing procedure is:
1. stripe segmentation: the computed image segmentation threshold specially changes image into bianry image;
2. striped search: the position of all stripeds in sequential search and the definite image;
3. striped refinement: all stripeds are carried out thinning processing, remove noise;
4. fringe count:, calculate the paper number by calculating the number of striped.
If the image striped of gathering is not very good, then step 1 is handled by enhancing and denoising to image earlier, the boundary characteristic of outstanding striped, and then carry out stripe segmentation.
If the striped through step 3 refinement is not very complete, then the striped after the thinning processing is carried out striped according to the trend of position relation and striped and repair, form complete stripe pattern.
Its workflow of device of the present invention is:
(1) open system's general supply after, system state detects automatically, being connected of computer system detection and video camera, sensor enters the wait detected state after normal, waits for the beginning detection signal, step S401; If malunion Chang Ze sends alerting signal, request manual intervention, step S402;
(2) move on mechanical platform when crossing the detection position when paper, sensor sends a signal because of being blocked.
(3) after computing machine enters and waits for detected state, circulation read sensor signal; After reading sensor signal, at first the state of this signal is removed, control image pick-up card and camera acquisition image then, and view data is sent to computing machine, step S403;
(4) after image acquisition was finished, computer starting Counting software module was finished paper counting, carried out figure image intensifying and denoising, image segmentation, striped search, striped refinement, striped repairing and fringe count, step S404;
(5) each result calculated is compared, compare in testing result then, determine the accuracy of count results two check points; If coming to the same thing of both sides picture count then confirmed this counting accurately, step S405; If different, then send cue, need manual intervention or tuple, step S406;
(6) system enters the wait detected state again, repeating step S401-S406
(7) after work is finished, log off and close general supply.
The present invention has also designed parameter setting, visual field equilibrium treatment function and database management function, wherein systematic parameter can be provided with according to different demands, such as revising systematic parameter, adapt to the detection demand at different paper, different rate request etc.
If desired, the normalized that can use the visual field balance module to carry out the visual field illumination reduces the influence of illumination to image acquisition by methods such as shade rectification, background denoisings.
The database management language module realizes classification and the statistics to testing result, particularly to can not accurately detecting the situation of volume, can statistic of classification, and sum up the reason of flase drop, the perfect foundation that provides of system is provided.
If desired, number system can also be connected with relevant network, and count results directly is sent to assigned address, is convenient to electronic management.
The final count results of the present invention, and other duty of system all shows on display interface, sees Fig. 6, makes operating personnel can in time understand the working condition of system, and adjusts timely.
Counting machine of the present invention can be realized following performance index substantially
(1) counting processing time＜2s of opening of every stacker;
(2) image acquisition resolution: 0.01mm;
(3) counting accuracy rate 100%;
(4) discrimination is less than 1%; So-called discrimination is meant that system can accurately discern every group of paper and calculate the probability of accurate number, can reach 99% at present.
When (5) can not discerning, system can provide miscue, requires identification once more or people's number to open, and corresponding image can be preserved for future reference.
With an object lesson counting algorithm of the present invention is elaborated again below in conjunction with Fig. 5.Need to prove that those skilled in the art just is enough to implement the present invention according to above-mentioned explanation, therefore following object lesson can not be used to limit protection scope of the present invention.
The course of work of counting algorithm is described with the example that folded 100 paper are counted:
When collecting 100 paper (Fig. 3 A) behind the striped of trimming clearly, at first carry out stripe segmentation, this step mainly is to utilize image processing method that stripe pattern is handled, step is as follows:
S501: utilize Gauss-Laplce's convolution function to carry out smoothing denoising, utilize standard edge to extract operator then, be such as but not limited to Sobel operator (adoptable edge detection operator except that the Sobel operator, also available Canny operator and Laplace operator etc.) and obtain gradient image after level and smooth.In gradient image, the striped that photographs has obtained enhancing, thereby helps cutting apart.Noise effect when taking in order to reduce, the present invention utilizes two-dimentional Otsu method split image, and this method is compared with general overall binaryzation dividing method, owing to considered the correlativity of neighborhood pixel, robustness is stronger.Fig. 3 B is the stripe pattern after cutting apart.
S502: be striped search location then; In order to orient fringe position in the bianry image that in the first step, obtains, at first to remove and cut apart the assorted point of noise that the back occurs, utilize the Top-Hat conversion in the mathematical morphology, according to striped length setting threshold value, will remove less than the striped of setting threshold.Utilize labelling method to mark remaining striped numbering and position then.
S503: the 3rd step was the striped refinement; In order to obtain the stripe pattern of single pixel, the striped oriented according to eight neighborhoods or relative theory, is adopted range conversion in the mathematical morphology to make skeletonizing and handles, obtain the bar graph (shown in Fig. 3 C) after the refinement.
S504: the 4th step was that striped is repaired; Determine the trend of single striped according to the directional information in the first step gained gradient image, walk upwards at striped the striped of fracture is repaired, form complete stripe pattern.
S505: the 5th step was a fringe count; In the refinement stripe pattern after repairing, utilize labeling acts that the refinement striped is numbered again once more, thereby obtain the number of striped, the number of striped is exactly the number of this stacker Zhang Shiji.
Claims (10)
1. contactless vision paper counting method forms the image striped with the paper side of desire counting, and the corresponding a piece of paper of each striped carries out the processing of following steps then to the image striped:
A) stripe segmentation: the computed image segmentation threshold specially changes image into bianry image;
B) striped search: the position of all stripeds in sequential search and the definite image;
C) striped refinement: all stripeds are carried out thinning processing, remove noise;
D) fringe count:, calculate the paper number by calculating the number of striped.
2. the method for claim 1 is characterized in that, step a strengthens with denoising image earlier to handle, the boundary characteristic of outstanding striped, and then carry out stripe segmentation.
3. the method for claim 1 is characterized in that, again the striped through refinement is carried out striped according to the trend of position relation and striped behind the step c and repairs, and forms complete stripe pattern.
4. counting machine of realizing the described method of claim 1 comprises:
An image acquisition mechanical platform that is used to put desire counting paper;
At least one video camera is installed in a side of image acquisition mechanical platform, is vertical with a side of this desire counting paper;
The video camera left and right sides is equipped with lighting source near the paper place;
A photoelectric sensor is installed in the below, inside of mechanical platform;
One computing machine is equipped with image pick-up card and I/O signal interface card; Image pick-up card is connected with video camera, and the I/O signal interface card receives the signal of photoelectric sensor;
By above-mentioned counting machine, desire counting paper is ingested image when moving to camera position earlier, continues then to move to the photoelectric sensor position, and photoelectric sensor sends signal, and the notice computing machine begins to control image acquisition and detection.
5. counting machine as claimed in claim 4 is characterized in that, the both sides of described desire counting paper all are equipped with video camera.
6. as claim 4 or 5 described counting machines, it is characterized in that described video camera is linear array or Array CCD Camera, and be furnished with enlarging lens.
7. as claim 4 or 5 described counting machines, it is characterized in that the distance between described video camera and the desire counting paper is adjustable.
8. counting machine as claimed in claim 4 is characterized in that, described lighting source is a bar shaped light emitting diode coaxial-illuminating light source.
9. as claim 4 or 8 described counting machines, it is characterized in that the distance between described lighting source and the desire counting paper is adjustable.
10. counting machine as claimed in claim 4 is characterized in that, described image acquisition mechanical platform is provided with the parallel orbit with desire counting paper same widths.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410049579 CN1710606A (en) | 2004-06-18 | 2004-06-18 | Non-contact vision paper-counting method and machine thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 200410049579 CN1710606A (en) | 2004-06-18 | 2004-06-18 | Non-contact vision paper-counting method and machine thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1710606A true CN1710606A (en) | 2005-12-21 |
Family
ID=35706845
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 200410049579 Pending CN1710606A (en) | 2004-06-18 | 2004-06-18 | Non-contact vision paper-counting method and machine thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1710606A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100416434C (en) * | 2006-09-21 | 2008-09-03 | 上海大学 | Method for stabilizing images of TV pickup system in optokinetic reflection based moving carrier |
| CN101045374B (en) * | 2006-03-31 | 2011-05-11 | 海德堡印刷机械股份公司 | Method and device for processing picture taking for edges of moving object |
| CN102456155A (en) * | 2010-10-25 | 2012-05-16 | 苏州启成精密工业有限公司 | Automatic material slice counting device |
| WO2014063469A1 (en) * | 2012-10-25 | 2014-05-01 | 广州广电运通金融电子股份有限公司 | Financial self-service device, banknote recognition module thereof and recognition method therefor |
| CN104268623A (en) * | 2014-09-18 | 2015-01-07 | 成都印钞有限公司 | Precise counting device and method based on microimaging method |
| CN105590318A (en) * | 2015-12-16 | 2016-05-18 | 中科院广州电子技术有限公司 | Laminated paper counting method and device |
| CN106408081A (en) * | 2015-07-30 | 2017-02-15 | 澄洋企业股份有限公司 | Image board quantity counting device and method |
| CN104240204B (en) * | 2014-09-11 | 2017-02-15 | 镇江苏仪德科技有限公司 | Solar silicon wafer and battery piece counting method based on image processing |
| CN107437109A (en) * | 2016-05-25 | 2017-12-05 | 北京弘浩千瑞科技有限公司 | A kind of method and apparatus by optoelectronic scanning to paper accurate metering |
| CN108090925A (en) * | 2017-12-13 | 2018-05-29 | 德淮半导体有限公司 | Paper number of pages computational methods, device, computer readable storage medium and terminal |
| CN108292357A (en) * | 2015-11-12 | 2018-07-17 | 飞利浦照明控股有限公司 | Image processing system |
| CN109712132A (en) * | 2018-12-28 | 2019-05-03 | 中国科学院长春光学精密机械与物理研究所 | A method of it is counted using electronic device counting device automatic screening |
| CN111429646A (en) * | 2020-05-07 | 2020-07-17 | 中国工商银行股份有限公司 | Banknote counter, banknote counting method, device, system and medium |
| CN112140721A (en) * | 2020-09-27 | 2020-12-29 | 湖北泰友纸业有限公司 | Printing quantity self-adjusting counting system for corrugated board printing |
-
2004
- 2004-06-18 CN CN 200410049579 patent/CN1710606A/en active Pending
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101045374B (en) * | 2006-03-31 | 2011-05-11 | 海德堡印刷机械股份公司 | Method and device for processing picture taking for edges of moving object |
| CN100416434C (en) * | 2006-09-21 | 2008-09-03 | 上海大学 | Method for stabilizing images of TV pickup system in optokinetic reflection based moving carrier |
| CN102456155A (en) * | 2010-10-25 | 2012-05-16 | 苏州启成精密工业有限公司 | Automatic material slice counting device |
| CN102456155B (en) * | 2010-10-25 | 2014-04-23 | 苏州启成精密工业有限公司 | Automatic material slice counting device |
| WO2014063469A1 (en) * | 2012-10-25 | 2014-05-01 | 广州广电运通金融电子股份有限公司 | Financial self-service device, banknote recognition module thereof and recognition method therefor |
| US9406185B2 (en) | 2012-10-25 | 2016-08-02 | Grg Banking Equipment Co., Ltd. | Financial self-service device, banknote recognition module thereof and recognition method therefor |
| CN104240204B (en) * | 2014-09-11 | 2017-02-15 | 镇江苏仪德科技有限公司 | Solar silicon wafer and battery piece counting method based on image processing |
| CN104268623A (en) * | 2014-09-18 | 2015-01-07 | 成都印钞有限公司 | Precise counting device and method based on microimaging method |
| CN106408081A (en) * | 2015-07-30 | 2017-02-15 | 澄洋企业股份有限公司 | Image board quantity counting device and method |
| CN106408081B (en) * | 2015-07-30 | 2018-09-28 | 澄洋企业股份有限公司 | Image-type plate body quantity counting device and method |
| CN108292357A (en) * | 2015-11-12 | 2018-07-17 | 飞利浦照明控股有限公司 | Image processing system |
| CN105590318A (en) * | 2015-12-16 | 2016-05-18 | 中科院广州电子技术有限公司 | Laminated paper counting method and device |
| CN107437109A (en) * | 2016-05-25 | 2017-12-05 | 北京弘浩千瑞科技有限公司 | A kind of method and apparatus by optoelectronic scanning to paper accurate metering |
| CN108090925A (en) * | 2017-12-13 | 2018-05-29 | 德淮半导体有限公司 | Paper number of pages computational methods, device, computer readable storage medium and terminal |
| CN109712132A (en) * | 2018-12-28 | 2019-05-03 | 中国科学院长春光学精密机械与物理研究所 | A method of it is counted using electronic device counting device automatic screening |
| CN109712132B (en) * | 2018-12-28 | 2021-03-19 | 中国科学院长春光学精密机械与物理研究所 | Method for automatically screening and counting by using electronic device counting device |
| CN111429646A (en) * | 2020-05-07 | 2020-07-17 | 中国工商银行股份有限公司 | Banknote counter, banknote counting method, device, system and medium |
| CN111429646B (en) * | 2020-05-07 | 2021-12-03 | 中国工商银行股份有限公司 | Banknote counter, banknote counting method, device, system and medium |
| CN112140721A (en) * | 2020-09-27 | 2020-12-29 | 湖北泰友纸业有限公司 | Printing quantity self-adjusting counting system for corrugated board printing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1710606A (en) | Non-contact vision paper-counting method and machine thereof | |
| CN103471512B (en) | A kind of glass plate width detecting system based on machine vision | |
| CN104729426B (en) | Angle steel automatic on-line detecting system and method based on machine vision | |
| CN111951237A (en) | Visual appearance detection method | |
| CN110068578B (en) | Apparent defect detection method and device for PVC (polyvinyl chloride) floor and terminal equipment | |
| CN104034638A (en) | Diamond wire particle online quality inspection method based on machine vision | |
| US10586321B2 (en) | Automatic detection, counting, and measurement of lumber boards using a handheld device | |
| CN104034637B (en) | Based on the online quality inspection device of diamond wire particle of machine vision | |
| US6298171B1 (en) | System for reading data glyphs | |
| CN105574845A (en) | Cigarette pack lamination layer number measurement method and device by multi-camera array | |
| US20070116350A1 (en) | Method for detecting the alignment of films for automated defect detection | |
| CN104597057A (en) | Columnar diode surface defect detection device based on machine vision | |
| CN108489394A (en) | A kind of large-scale sheet metal works almost T-stable automatic detection device and method | |
| CN107966102A (en) | A kind of plate production six-face detection device | |
| CN103630544A (en) | Online visual detection system | |
| CN111982933A (en) | Coating defect detection system and device | |
| CN107742119B (en) | Object contour extraction and matching device and method based on back-image imaging | |
| CN111724445A (en) | Method and system for identifying large-view small-size identification code | |
| CN1575477A (en) | Automatic digitization of garment patterns | |
| CN108230385B (en) | Method and device for detecting number of ultra-high laminated and ultra-thin cigarette labels by single-camera motion | |
| CN208042989U (en) | A kind of large-scale sheet metal works almost T-stable automatic detection device | |
| CN107403429B (en) | Method for quickly and automatically acquiring parameters of periodic sequence image model | |
| RU2435661C2 (en) | Scanner system of loading device | |
| CN113256568A (en) | Machine vision plate counting general system and method based on deep learning | |
| CN109387525A (en) | On a kind of film in film defect determination method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |