Machine vision defect detection

 Chengdu Xinxiwang Technology has rich engineering experience in the field of machine vision defect detection technology (especially line scanning systems), and has profound technical accumulation in precision structures, motion control, pattern recognition, feature search, and parallel computing acceleration. The following is an example of a defect detection system:

1. Defect inspection based on color characteristics

It mainly performs statistical processing on the color features of the surface of the measured object, based on vector features of RGB, YCbCr, HSI and other color spaces, and uses various methods such as pattern classification and feature learning to distinguish defect features from normal features and determine and identify defect locations. Mainly used in printing and packaging, PCB manufacturing, medical equipment and other fields.

2. Defect inspection based on geometric features

Mainly focusing on the geometric shape characteristics of the product, combined with known feature information such as CAD, noise processing, feature extraction, fitting, and parameter extraction are performed on the corresponding areas of the product, and then compared with the design features, the deviation value exceeds the set value. Determine the location of the area.

3. Defect inspection based on statistical features such as texture and blob

Based on the statistics of the surface gray distribution characteristics of the product, high-order statistical factors, edge detection operators and other methods are used to separate defect points from the background and quickly identify defect locations.