CN102262093A

CN102262093A - Machine vision-based on-line detection method for printing machine

Info

Publication number: CN102262093A
Application number: CN2010101794575A
Authority: CN
Inventors: 张爱明
Original assignee: Individual
Current assignee: Individual
Priority date: 2010-05-24
Filing date: 2010-05-24
Publication date: 2011-11-30

Abstract

The invention relates to the field of machine vision, in particular to an on-line detection method for printer vision. According to the on-line detection method, the problems of poor real-time response property, more workers, high labor intensity and low detection level in manual detection of printing defective products are solved. In the an on-line detection method, a printed product image is scanned by using an image scanning device line by line, is transmitted to a controller of the printing machine, is processed and is compared with a corresponding data segment of a printed image file in a vision controller; if flaws such as register trouble, missing print, double image, chromatic aberration and the like exist, an interface circuit of a machine vision controller is used for outputting a fault signal to promote an operator to process or feed back a trouble code to an upper controller; and the upper controller is used for carrying out next control. The on-line detection method disclosed by the invention has high detection response speed and high precision, is used for detecting in real time in printing process and avoiding printing defective products to the maximum extent.

Description

Printing machine online test method based on machine vision

Technical field

The present invention relates to utilize Vision Builder for Automated Inspection to carry out the technical field of online detection, relate in particular to, utilize Vision Builder for Automated Inspection that the flaw of calico is carried out on-line detection method in the calico production scene.

Background technology

At the workshop scene of line production printing and dyeing calico, need carry out online detection to the flaw of calico.In the prior art, the online detection of calico flaw relied on manually detect, be provided with 2-4 people in the both sides of printing machine the flaw of calico is estimated and handled, and establish tracking, the detection that 1 people carries out product quality at oven dry cropping place.The shortcoming of its existence has: on-the-spot humiture is very high, and workman's testing environment is abominable, and labour intensity is big; The workman works long hours and is easy to generate visual fatigue, during especially many, long-time, equipment high-speed cruising in pattern complexity, chromatography, can not in time reflect the qualitative character of PRINTED FABRIC in the production run, occur the stamp defective product easily, thus can not comprehensive implementation quality control.Therefore can't guarantee the qualification rate of product export.

Summary of the invention

Big and be easy to generate visual fatigue, can't guarantee to detect quality and the low problem of product export qualification rate at prior art to the online manual detection labour intensity of calico flaw, the invention provides a kind of Vision Builder for Automated Inspection the online test method of calico flaw has been overcome the instability factor that human eye detection is brought.For example, when we concentrate to observe decalcomania, have only when the contrast colors of decalcomania are relatively stronger, human eye can be found the defective that about 0.3mm is above, and the detection quality that obtains is difficult to keep continual and steady.Because human eye fatiguability under high light can not adapt to long-time detection.And if same colour system, the especially decalcomania of light color system are when detecting, human eye just is difficult to find the defective of depositing, needs 30 above gray level difference at least.Even and have only the difference of a gray level, Vision Builder for Automated Inspection also can be found the defective of 0.1mm size fully aware ofly.In the textile printing production run, when the printing quality defective occurring, Vision Builder for Automated Inspection can be found defect problem immediately, reduces defective product.Managerial personnel can carry out quality monitoring to whole production line according to the data analysis of Vision Builder for Automated Inspection, improve the efficiency of management.

The technical scheme of the inventive method is as follows:

1, a kind of printing machine online test method based on machine vision, the control system that this detection method adopts is contained in the printing machine controller.

2, described online test method may further comprise the steps:

(1). the pattern of determining during with the plate-making of institute stamp type as the standard picture template stores in described printing machine controller;

(2). according to customer requirements the accuracy rating that detects parameter is set, determines the permissible value of flaw area and gradation of image error;

(3). send trigger pip control industrial camera by the printing machine controller, take the printing product pattern that printing machine is printed out in real time, and the image of taking is sent to described printing machine controller;

(4). described printing machine controller carries out series of algorithms such as wavelet analysis, texture analysis, smoothing processing, binary conversion treatment to be handled the graphic image of real-time shooting, extracts characteristics of image, compares with the standard picture template that step (1) is stored;

(5). step (4) judges that then corresponding printing product is a defective product if flaw area of comparing out and gradation of image error exceed the permissible value of step (2), and the printing machine controller is taked corresponding control measures, in time avoids defective product to produce once more

The technique effect of the inventive method:

The present invention adopts Vision Builder for Automated Inspection that printing machine printing process is carried out online detection, can both in time find the flaw that occurs in the printing process when printing each edition, and response speed is fast, has avoided flaw to repeat to produce.Stopped to have printed in the stamp production run in the past last edition just find flaw, sometimes in addition class also can't find the phenomenon of flaw.

The resolution of described industrial camera is 1600 * 1200 o'clock, and the detection of dynamic precision can reach 0.02mm, and human eye can only be found the defective that about 0.3mm is above, and accuracy of detection is greatly improved and the flaw loss is 0.

The online test method to calico of having deposited generally is to select complete display according to customer requirements on the calico of operation, flawless image is stored in computing machine as the standard picture template, or set up complicated flaw model storehouse in advance as the standard picture template according to customer requirements, the pattern that the present invention determines during with the plate-making of institute stamp type as the standard picture template stores in described printing machine controller, graphic image to real-time shooting is carried out wavelet analysis, texture analysis, smoothing processing, after series of algorithms such as binary conversion treatment are handled, extract characteristics of image, compare with the standard picture template of being stored, more suit the designing requirement of product, guaranteed the accuracy that detects.

Description of drawings

Fig. 1 is a hardware system block diagram of the present invention;

Fig. 2 is a process flow diagram of the present invention;

Embodiment

As shown in Figure 1, this hardware components is made up of industrial camera 1, light source 2, printing machine controller 3.The industrial camera of present embodiment adopts the In-Sight1403C of Cognex, and resolution is 1600 * 1200, can carry out the high resolving power inspection to multiple colored the application.Light source adopts the annular light source that surrounds video camera.Control system is built in the printing machine controller, is made up of a plurality of modules such as trigger pip generation module, the pre-storing module of image masterplate, absorption image pretreatment module, image comparing module, feedback warnings.Being illustrated in figure 2 as the process flow diagram of online detection, below is the detailed description of this flow process.

1, when plate-making the pattern determined as the standard picture template stores in described printing machine controller;

2, set the permissible value of flaw area and gradation of image error;

3, the trigger pip generation module 4 by the printing machine controller triggers industrial camera 1 shooting printing product real-time image, by cable input printing machine controller 3

4, image pre-service: image gray processing, expeling noise, grey level stretching, binaryzation and morphologic filtering.

Image gray processing, the image that camera is taken in real time is 16 bitmaps (RGB565 form) data, handles for the ease of follow-up rapid image, needs view data is changed, and makes coloured image become 256 grades of gray-scale maps.

Remove noise, inevitably contain noise in the image, adopt medium filtering that image is carried out the pre-service grey level stretching thereupon,, image is carried out grey level stretching in order to strengthen the contrast of background area and character zone.Binaryzation is carried out binary conversion treatment to gray level image, adopts the method for maximum between-cluster variance and infima species internal variance ratio, and self-adaptation is calculated gray scale door screen value, thinks the target area less than the zone of this door screen value, greater than the background area of thinking in the zone of this late value.

Morphologic filtering carries out morphologic filtering to bianry image and handles, the synthetic operation that adopts expansion, burn into ON operation and closed operation to combine.

The target extraction module, target selection selects graphics field or close position to have the part of remarkable rigidity characteristic as localizing objects in the standard picture template.

Feature extraction utilizes the Canny operator to extract the closed contour feature and the line feature of non-closure in localizing objects, according to scale size the geometric properties that extracts is sorted.

Location training is made as each yardstick according to from big to small order and cuts apart door screen value, carrying out the quick characteristic matching of the overall situation in the large-scale characteristics image template of the geometric properties that is extracted.

4, the pixel location in the template compares, and whether the pixel location, color, gray scale of judging calico be in the error allowed band of setting.

5, discovery flaw printing machine controller takes appropriate measures and adjusts automatically or stopping alarm.Avoid producing continuously flaw.

Claims

1. the printing machine online test method based on machine vision is characterized in that the control system that this detection method adopts is contained in the printing machine controller.

2. printing machine online test method according to claim 1 is characterized in that may further comprise the steps:

(5). step (4) judges that then corresponding printing product is a defective product if flaw area of comparing out and gradation of image error exceed the permissible value of step (2), and the printing machine controller is taked corresponding control measures, in time avoids the defective product continuity.