CN102565083B - Offline inspection system and method for printed electronic product - Google Patents

Abstract

The invention discloses an offline inspection system for a printed electronic product. The offline inspection system comprises an image acquisition system, an image processing system and an alarm prompt system. The invention additionally discloses an offline inspection method for the printed electronic product. The offline inspection method comprises the following steps of: 1) using the image acquisition system to acquire the image of the finished printed electronic product; 2) using the image processing system to find out the difference between the real-time acquired image and the image of a sample; 3) pointing out defects and using the alarm prompt system to give out an alarm in time; and 4) using the image processing system to conduct statistical analysis to the defects. By using the system and the method to inspect the printed electronic product off line, the inspection efficiency is high, the inspection accuracy is high, the product quality can be effectively improved and the reject rate is reduced at the same time.

Description

A kind of printed electronic product off-line checking system and method

Technical field

The present invention relates to print product detection technique field, refer to a kind of printed electronic product off-line checking system and method especially.

Background technology

No matter international or domestic, printed electronic product is owing to using printing technology (based on serigraphy and ink jet printing), and wherein ink jet printing is the inevitable choice of printing that becomes more meticulous.Therefore, the printing defects problem existed in paper products printing process inevitably exists at the printing process of printed electronic product equally:

● register trouble: the base unit of chip is transistor, transistor has the sandwich constructions such as insulated substrate layer, N-type semiconductor, P-type semiconductor, electrode, passivating film, the material of every Rotating fields is different, needs multiplex priting operation, adopts different ink materials just can be completed for printing respectively.For ensureing consistance and the stability of electric transistor parameter, require that the position of per pass printing process chromatography is very accurate.Because the tension force of organic film affects with variation of ambient temperature, easily cause the problem of printing register trouble, thus cause transistor electricity performance parameter destroy and scrap, as on-line checkingi not in time, huge waste will be caused in the batch printing of volume to volume.

● circuit adhesion: when the distance between circuit and circuit is very near, due to the expansion of electrically conductive ink ink droplet and the change of permeance property, is likely a bit sticked together adjacent two lines at certain, causes signal shorts, destroys circuit function.

● circuit disconnects: when circuit is very narrow, due to the contingency blocking of shower nozzle or shake in jet printing process, and dust granule, winged insect etc. drop on will on the circuit of spray printing, cause spray printing circuit to disconnect, signal cannot pass through, thus destruction circuit function.

● circuit thickness is uneven: cause due to the contingency blocking of shower nozzle or the mechanical shaking of printing platform in jet printing process, affect the electrical parameter of circuit.

● line edges is rough, due to the mechanical shaking of printing platform, or the expansion of ink droplet and the change of permeance property, make the line edges of spray printing there is burr, affect the electrical parameter of circuit.

The quality on-line detection system of current paper printed matter just rises, but can not directly take back for printed electronic product print quality inspection.First, both accuracy of detection requires different, and the former accuracy of detection is relatively low, and the accuracy of detection of the latter requires higher, to algorithm and equipment requirement also higher; Secondly, the circuit surface of the latter has metallic luster to reflect, and has larger interference to imaging, requires higher to image processing algorithm.

Summary of the invention

In view of this, the object of the invention is to propose a kind of printed electronic product off-line checking system and method, it can carry out offline inspection to the finished product of printed electronic product, ensures product percent of pass, and auxiliary optimizing product performance.

Based on above-mentioned purpose a kind of printed electronic product off-line checking system provided by the invention, comprising:

Image capturing system, comprising lighting source, camera, scrambler, image pick-up card, described lighting source irradiates printed electronic product surface, described camera is vertically taken printed electronic product surface, utilize the scrambler be connected with printing equipment controller to control the synchronized movement of the shooting action of printed electronic product and camera, and collect printed electronic product information by the image pick-up card be connected on camera;

Image processing system, comprising computing machine, for searching for the printing defects detection system based on Fast template matching algorithm, the printing defects sorter for printing equipment fault diagnosis of printing defects;

Alarm system, comprising alarm controller and alarm.

Optionally, described printed electronic product to be detected is provided with triggered mark, described image capturing system is provided with trigger sensor, when described trigger sensor senses triggered mark, trigger described camera and carry out image acquisition.

Preferably, image capturing system is also provided with for the synchronous photoelectric encoder of image acquisition.

Preferably, described Fast template matching algorithm is modified genetic algorithm, first utilizes low precision search to get rid of impossible position, then searches for defect by high precision.

Optionally, printing defects is divided into wire, point-like, band shape, planar by described printing defects sorter, and carries out statistical study to defect classification.

Present invention also offers a kind of printed electronic product off-line checking method, comprise the following steps:

The printed electronic product image that real-time collection and continual collection printing equipment is produced, its size of adopting graph region is consistent with examination criteria template size;

Adopt image procossing Fast template matching algorithm, real-time image acquisition and examination criteria template carried out mating and contrasts, finding the place inconsistent with examination criteria template to be defective locations;

Utilize printing defects detection system to analyze defective locations, iris out the size and shape of defect, and indicate position and the shape of defect in former figure frame, show on a computer display, provide alerting signal to alarm controller, alarm equipment alarm simultaneously;

Printing defects is divided into wire, point-like, band shape, planar, carries out respectively adding up and depositing, inquire about for the later stage and show picture frame corresponding to all kinds of defect, realizing playback again and analysis.

Optionally, described camera is industrial camera or smart camera.

Optionally, described camera is face battle array industrial camera or linear array industrial camera.

Accordingly, described lighting source is chosen as area array light source or strip-shaped light source.

Preferably, the image acquisition mode of described collected by camera examination criteria template is one-step control mode.

Optionally, described examination criteria template is that the image information by gathering printed electronic product sample obtains, or directly calls existing template from computing machine.

Preferably, described camera and printed electronic product surface keeps vertical, and utilization control with the scrambler that printing equipment controller connects and keep camera to absorb that action and printed electronic product move is synchronous.

As can be seen from above, a kind of printed electronic product off-line checking system provided by the invention and method, by the printed electronic product of this system and method on-line checkingi, its detection efficiency and accuracy of detection are better than common print product detection mode all greatly, and can effectively improve the quality of products and rate of simultaneously reducing the number of rejects and seconds.

Accompanying drawing explanation

Fig. 1 is the printed electronic product off-line checking system schematic diagram of the embodiment of the present invention;

Fig. 2 is template matching algorithm in the embodiment of the present invention, when being K=2, when template size is 4 × 4, carries out the pixel schematic diagram of selection during rough matching;

Fig. 3 is that in the embodiment of the present invention, template matching algorithm improves later algorithm flow chart schematic diagram.

Embodiment

For making the object, technical solutions and advantages of the present invention clearly understand, below in conjunction with specific embodiment, and with reference to accompanying drawing, the present invention is described in more detail.

With reference to figure 1, for the printed electronic product off-line checking system schematic diagram of the embodiment of the present invention, comprising lighting source 1, checkout equipment roll shaft 2, camera 3, scrambler 4, image pick-up card 5, computing machine 6, alarm controller 7, alarm 8, printed electronic product 9, triggered mark 10.

Wherein, the image capturing system triggered mark 10 that includes lighting source 1, camera 3, scrambler 4, image pick-up card 5 and be arranged on printed electronic product 9; Image processing system includes computing machine 6, and 6, computing machine contains printing defects detection system and printing defects sorter; Alarm system includes alarm controller 7 and alarm 8.

Described image capturing system is erected at above checkout equipment, the surface of lighting source 1 oblique illumination printed electronic product 9 at a certain angle, the surface of printed electronic product 9 is provided with triggered mark 10, the rotation of checkout equipment roll shaft 2 drives printed electronic product 9 to travel forward, trigger sensor is provided with accordingly with triggered mark 10 position above checkout equipment, described triggered mark 10 is set to, when the reference position of the required part detected of printed electronic product 9 moves to immediately below camera 3, trigger described trigger sensor and send the signal starting to gather to described camera 3, camera 3 starts the image information absorbing printed electronic product 2 surface, described camera 3 is arranged on directly over the surface of the described printed electronic product 2 needing to carry out image acquisition, described camera 3 is vertically taken printed electronic product 2 surface, the scrambler 4 be connected with printing equipment controller is utilized to control the synchronized movement of the shooting action of printed electronic product 2 and camera 3, and the image information of printed electronic product 2 is collected by the image pick-up card 5 be connected on camera 3.

Computing machine 6 in described image processing system is after the image information receiving image capturing system collection, image information carries out Defect Search by the printing defects detection system of the Fast template matching algorithm adopting a kind of genetic algorithm based on improving in computing machine 6, it first utilizes low precision search to get rid of impossible position, search for matched position by high precision again and find defect, this algorithm principle is as follows:

Only need the position found roughly in genetic algorithm part, and if malposition also can be passed through climbing method and return rear continuation and find, can matching process and adopt rough matching process accurately during coupling.In the matching process of each coordinate points, the point corresponding with original image to points all in template is not needed to compare, but selected part point.Choose pixel horizontal and vertical direction being spaced apart template K to mate, the calculated amount of such genetic algorithm part is original 1/K ², situation when Fig. 1 is K=2.Genetic algorithm fitness function after improvement is as follows:

Wherein:

At genetic algorithm part K=min{P, Q}/10, and at climbing method part K=1, namely still adopt original accurate matching process in climbing method part, the accuracy of algorithm can be ensured like this.

With reference to figure 2, be K=2, when template size is 4 × 4, carry out the pixel of selection during rough matching; In Fig. 2, Grey Point is choose the point carrying out mating.

Climbing method search procedure is again entered in order to prevent the point carrying out climbing method search, a searching route table is added to be kept at the point of process in climbing method process in algorithm, when optimum position is the point in routing table, just no longer enter climbing method search procedure.

Improve later algorithm flow chart with reference to figure 3.

Can find out, this algorithm make use of genetic algorithm fast search speed to reduce match time, utilize climbing method local search ability strong simultaneously, the advantage of exact position can be rapidly converged to ensure matching precision.Secondly, this algorithm also incorporate first thick after the searching idea of thin, Query refinement, utilizing low precision search when starting to search for, getting rid of impossible position fast; Adopt high precision searching and matching method in possible position, find exact position.And compared with traditional matching algorithm, the speed of this algorithm is faster, locate more accurate.

After image processing system finds defect, this information is delivered to alarm system, alarm controller 7 receives alerting signal, and is delivered to alarm 8, and alarm 8 is reported to the police, and defect has appearred in prompting staff print product.

In addition, image processing system carries out fault diagnosis by the printing defects sorter in computing machine 6 to printing equipment, and its principle is as follows:

Extract the profile of defect, carried out ellipse fitting, finally measure the parameter information of defect: defect area, barycentric coordinates, major and minor axis length.

The simplest provincial characteristics is the area in region:

$a=\left|R\right|=\underset{(r,c)\∈R}{\mathrm{\Σ}}1=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{ce}}_i-{\mathrm{cs}}_i+1---\left(1\right)$

From above formula, the area a in region is exactly counting in region | R|.If region one width bianry image represents, so with the area of first summation equation zoning in above formula; If region run-length encoding represents, so with the area of second summation equation zoning in above formula.

Area is a special case in the generalized character of the square being called as region.P >=0, during q >=0, (p, q) rank square is defined as:

$m_{p,q}=\underset{(r,c)\∈R}{\mathrm{\Σ}}{r}^p{c}^q---\left(2\right)$

Note m _0,0it is exactly the area in region.Similar with reference area, only calculate square based on the simple equation of stroke because can derive, just can calculate square expeditiously so use during stroke representation.

Square in equation (2) depends on the size in region.Usually we expect to have some features can not with dimension of object change and change, for obtaining such feature, when p+q >=1, square just obtains normalized square divided by the area in region:

$n_{p,q}=\frac{1}{a}\underset{(r,c)\∈R}{\mathrm{\Σ}}{r}^p{c}^p---\left(3\right)$

The center of gravity in region from the derive the most interesting characteristic that obtains of normalized square, i.e. (n _1,0, n _0,1).It can be used to the position describing region.

Normalized square is determined by the position in image.Usually, it is very useful for making feature not change with the change in location in region in image.This can realize relative to the square of regional barycenter by calculating.These center squares are calculated by following formula when (p+q >=2):

${\mathrm{\μ}}_{p,q}=\frac{1}{a}\underset{(r,c)\∈R}{\mathrm{\Σ}}{(r-n_{\mathrm{1,0}})}^p{(c-n_{\mathrm{0,1}})}^q---\left(4\right)$

Notice that these center squares are also after normalized.Second-order moment around mean (p+q=2) especially merits attention, and they can be used for the direction of defined range and the scope in region.This first order and second order moments obtaining region from an ellipse by hypothesis realizes.Then, five oval geometric parameters are derived from these five squares.Oval center is consistent with the center of gravity in region.Oval major axis r ₁with minor axis r ₂, and can be calculated by following formula relative to the angle of transverse axis:

$r_1=\sqrt{2({\mathrm{\μ}}_{\mathrm{2,0}}+{\mathrm{\μ}}_{\mathrm{0,2}}+\sqrt{{({\mathrm{\μ}}_{2,0}-{\mathrm{\μ}}_{\mathrm{0,2}})}^2+4{\mathrm{\μ}}_{\mathrm{1,1}}^2})}$

$r_2=\sqrt{2({\mathrm{\μ}}_{\mathrm{2,0}}+{\mathrm{\μ}}_{\mathrm{0,2}}-\sqrt{{({\mathrm{\μ}}_{\mathrm{2,0}}-{\mathrm{\μ}}_{\mathrm{0,2}})}^2+4{\mathrm{\μ}}_{\mathrm{1,1}}^2})}---\left(5\right)$

$\mathrm{\θ}=-\frac{1}{2}\arctan \frac{2{\mathrm{\μ}}_{\mathrm{1,1}}}{{\mathrm{\μ}}_{\mathrm{0,2}}-{\mathrm{\μ}}_{\mathrm{2,0}}}$

By the parameter of ellipse, we can derive another very useful feature: anisotropy r ₁/ r ₂.This characteristic quantity is keep invariable when area zoom, and it can describe the elongated degree in a region.

Classifier design:

If printing image defect area is Area, the ratio R=r of minor axis and major axis ₂/ r ₁, then classifier design is as follows:

If (Area <=4) then defect is point-like;

If (Area > 4) and (R < 0.3) then defect is wire;

If (Area > 4) and (0.3=< R < 0.6) then defect is banded;

If (Area > 4) and (0.6=< R <=1) then defect is planar;

Above classifier design, extract the effective image feature of existing printing defects, construct the sorting criterion function of printing defects, define printing defects sorter, printing defects can be divided into wire, point-like, band shape, planar class, and carry out statistical study, for the fault diagnosis of printing equipment, be convenient to follow-uply to debug for printing equipment defect.

The printed electronic product off-line checking method that printed electronic product off-line checking system described in utilization carries out is as follows:

The first step, on printed electronic product 9 and its sample, the position corresponding to trigger sensor arranges triggered mark 10;

Second step, open lighting source 1, light irradiates the sample of printed electronic product 9 at a certain angle, when trigger sensor senses triggered mark 10, trigger camera 3 and vertically candid photograph pickup image information is carried out to the sample surface of printed electronic product 2, and the two field picture preserving the sample of printed electronic product 9 is to computing machine 6, as examination criteria template, and defines the magnitude range of template area;

3rd step, keep camera 3 vertical with printed electronic product 9 surface, and utilize to control with the scrambler 4 that printing equipment controller connects and keep camera 3 to absorb that action and printed electronic product 9 move is synchronous, the image information of the printed electronic product 9 utilizing camera 3 and image pick-up card 5 real-time collection and continual collection printing equipment to produce, its size of adopting graph region is consistent with examination criteria template size;

4th step, computing machine 6 adopts image procossing Fast template matching algorithm by printing defects detection system, real-time image acquisition and examination criteria template is carried out mating and contrasts, finding the place inconsistent with examination criteria template to be defective locations;

5th step, utilizes printing defects detection system to analyze defective locations, irises out the size and shape of defect, and in former figure frame, indicate position and the shape of defect, show on a computer display, provide alerting signal to alarm controller 7, alarm 8 is reported to the police simultaneously;

6th step, is divided into wire, point-like, band shape, planar by printing defects, carry out respectively adding up and depositing, and inquires about and shows picture frame corresponding to all kinds of defect, realize playback again and analysis for the later stage.

Can find out, by the printed electronic product of this system and method offline inspection, because the Fast template matching algorithm which employs modified genetic algorithm is searched printing defects, substantially increase the efficiency of Defect Search, and precision is high, thus coordinating the detection efficiency of this System and method for of the print product offline inspection of this algorithm high, accuracy of detection is high, and can effectively improve the quality of products and rate of simultaneously reducing the number of rejects and seconds; Meanwhile, adopt printing defects sorter to carry out fault diagnosis to printing equipment, defect can be carried out statistic of classification and analyze, search printing defects for follow-up, and corresponding debugging is carried out to printing equipment play good booster action, save time and efforts.

It is to be noted, the optional industrial camera of described camera 3, smart camera or other machine vision can be utilized to carry out the camera taken, preferably face battle array industrial camera or linear array industrial camera, corresponding lighting source 1 can select area array light source or strip-shaped light source, and the image acquisition mode of collected by camera examination criteria template is then preferably one-step control mode.

In addition, when conditions permit, described examination criteria template also directly can call in computing machine 6 the existing Template Information stored, and all need not carry out image acquisition to the sample of printed electronic product at every turn; The setting of triggered mark 10 and trigger sensor can be arranged as required, and the effect arranging triggered mark 10 is to reject the part not needing to gather image information, when between the part needing collection image, interval is less, also the mode of continuous acquisition image can be taked, then without the need to arranging separately triggered mark 10 again.

Those of ordinary skill in the field are to be understood that: the foregoing is only specific embodiments of the invention; be not limited to the present invention; within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (10)

1. a printed electronic product off-line checking method, is applied to printed electronic product off-line checking system, and this system comprises: image capturing system, image processing system and alarm system; Described image capturing system comprises lighting source, camera, scrambler, image pick-up card, and described image processing system comprises computing machine, for searching for the printing defects detection system based on Fast template matching algorithm, the printing defects sorter for printing equipment fault diagnosis of printing defects; It is characterized in that, comprise the following steps:

The printed electronic product image that real-time collection and continual collection printing equipment is produced, its size of adopting graph region is consistent with examination criteria template size;

Adopt image procossing Fast template matching algorithm, real-time image acquisition and examination criteria template carried out mating and contrasts, finding the place inconsistent with examination criteria template to be defective locations; Described Fast template matching algorithm is modified genetic algorithm, first utilizes low precision search to get rid of impossible position, then searches for matched position by high precision and find defect; Described modified genetic algorithm fitness function is as follows:

Wherein:

Utilize printing defects detection system to analyze defective locations, iris out the size and shape of defect, and indicate position and the shape of defect in former figure frame, show on a computer display, provide alerting signal to alarm controller, alarm equipment alarm simultaneously;

Printing defects is divided into wire, point-like, band shape, planar, carries out respectively adding up and depositing, inquire about for the later stage and show picture frame corresponding to all kinds of defect, realizing playback again and analysis.

2. printed electronic product off-line checking method according to claim 1, it is characterized in that, printed electronic product to be detected is provided with triggered mark, described image capturing system is provided with trigger sensor, when described trigger sensor senses triggered mark, trigger described camera and carry out image acquisition.

3. printed electronic product off-line checking method according to claim 1, is characterized in that, image capturing system is also provided with for the synchronous photoelectric encoder of image acquisition.

4. printed electronic product off-line checking method according to claim 1, is characterized in that, printing defects is divided into wire, point-like, band shape, planar by described printing defects sorter, and carries out statistical study to defect classification.

5. printed electronic product off-line checking method according to claim 1, is characterized in that, described camera is industrial camera or smart camera.

6. printed electronic product off-line checking method according to claim 5, is characterized in that, described camera is face battle array industrial camera or linear array industrial camera.

7. printed electronic product off-line checking method according to claim 6, is characterized in that, described lighting source is area array light source or strip-shaped light source.

8. printed electronic product off-line checking method according to claim 1, is characterized in that, described examination criteria template is that the image information by gathering printed electronic product sample obtains, or directly calls existing template from computing machine.

9. printed electronic product off-line checking method according to claim 8, is characterized in that, the image acquisition mode of described collected by camera examination criteria template is one-step control mode.

10. the printed electronic product off-line checking method according to claim 1 to 9 any one, it is characterized in that, described camera and printed electronic product surface keeps vertical, and utilization control with the scrambler that printing equipment controller connects and keep camera to absorb that action and printed electronic product move is synchronous.