Stroboscopic stepped illumination defect detection system

Abstract

The invention discloses a stroboscopic stepped illumination defect detection system, which is used for appearance defect detection of a product and comprises an image extraction unit (101), a brightness adjusting unit (102), a data processing unit (103) and a stroboscopic control unit (104). Wherein the image extraction unit (101) is connected with the stroboscopic control unit (104) and is used for acquiring stable and clear images in various transmission / reflection visual bright fields / dark fields; and the brightness adjusting unit (102) is connected with the stroboscopic control unit (104)and is used for setting various transmission / reflection visual bright fields / dark fields and converting in the various transmission / reflection visual bright fields / dark fields. According to the defectdetection system, under the condition that multiple cameras are adopted, the stroboscopic step control over the light sources is carried out, that is, the starting pulse width and the working currentvalue of the light sources are changed at the same time, and the multi-light-source cooperative work is adopted, so that the purpose of multi-directional imaging of various defects is achieved.

Description

technical field [0001] The invention relates to the technical field of machine vision defect detection, in particular to a machine vision appearance defect detection system and an implementation method with multiple cameras and strobe step lighting. Background technique [0002] The machine vision system converts the captured target into an image signal through an image capture device (CMOS and CCD sensors), and transmits it to a dedicated image processing system. The image system performs various operations on these signals to extract the features of the target. Its basic components usually include cameras, lenses, light sources, image acquisition cards, image processors, image processing software and algorithms. [0003] At present, machine vision systems have been deeply used in industrial inspection systems for identifying product information, measuring product sizes, and detecting product defects. Among them, since users have very high requirements on the appearance ...

AI Technical Summary

Problems solved by technology

3D vision technology includes binocular vision, structured light, TOF (time of flight), triangular distance measurement, laser radar, etc. These technologies are currently widely used in three-dimensional physical recognition, measurement, and positioning, but they are rarely used in defect detection. And it is also difficult to meet the micron-level high-precision requirements of 3C products

[0009] However, in the prior art, the images of the four scenes (reflected bright field, reflected dark field, transmitted bright field, and transmitted dark field) obtained by the detection system of the detected object through the line array camera often reflect limited defects, because the linear array The speed at which the camera takes pictures is fixed, and the driving current of the light source corresponding to the shooting angle is fixed. Therefore, it is often necessary to change the pulse width of the corresponding light source when shooting different defects, and the maximum pulse width of the light source is the exposure of the line scan camera. Time, no matter how wide the pulse width is, it has no effect on the line scan camera, so this greatly limits the number of pictures taken, and also limits the types of defect detection

Moreover, because the current of the driving circuit corresponding to a single light source is fixed, when changing the illumination angle to take pictures of different bright field or dark field, often this light source will not be used for each defect at the same time, so the utilization rate of the light source is very high. Low, fewer types of defects for simultaneous shooting

At this time, the number of light sources and cameras can only be increased to take the required photos or more types of defects, which increases the cost and complexity of the system

Images