Image detection method for corner defects of rectangular wafer

Abstract

The invention provides an image detection method for corner defects of a rectangular wafer. The method comprises the following steps of respectively detecting four preprocessed corner images of the rectangular wafer, firstly obtaining a corner edge coordinate sequence of the corner images, then obtaining a judgment sequence according to the corner edge coordinate sequence, and finally, judging whether the wafer has the corner defect according to the numerical value of the judgment sequence. According to the method, the judgment sequence is obtained through the same wafer, the method is based on self-comparison of the same wafer, detection errors caused by angle differences of different wafers are overcome, the method can be widely applied to corner defect detection of smooth, semi-corrosion and full-corrosion wafers, and the method is advantaged in that the detection speed is high, accuracy is high, and training and learning are not needed; and requirements of high-speed and high-precision wafer defect detection instruments and equipment can be met.

Description

technical field [0001] The invention belongs to the technical field of image detection, and relates to a detection method of high-speed and high-precision defect detection equipment, in particular to an image detection method of rectangular wafer corner defects. Background technique [0002] In the industrial production process, it is usually necessary to use machine vision to judge whether there are defects on the surface of the product. Due to differences in manufacturing processes, batches, and corrosion degrees, the corners of rectangular wafers are not sharp, but have a certain radian and their radians vary. There are certain differences. The angular radian of a rectangular wafer is not a constant value, but fluctuates within a certain range. On the other hand, the corner defects of rectangular wafers are usually only one or two pixels in size, which is equivalent to the fluctuation range of the corner radian. Corner defect detection is greatly disturbed by angular r...

AI Technical Summary

Problems solved by technology

Traditional defect detection technology mainly relies on filtering and feature extraction methods, which cannot eliminate the influence of angle radian fluctuations, and it is difficult to give effective corner defect detection results

However, the defect detection technology based on learning algorithm has a slow calculation speed and requires a lot of training and learning, which is difficult to meet the speed requirements of high-speed defect detection equipment.

Images