Research on Digital Testing Technology for Full Surface Defects of Cross scale Heteromorphic Metal Devices

Abstract

Abstract There are many types of cross scale special-shaped metal components in various fields, and the surface quality (defects) of some key components will affect the device performance and even cause the long-term stability and reliability of the system to decline. The existing manual inspection methods are difficult to accurately quantify and establish a digital traceability database, so it is urgent to study the automatic digital inspection technology of full surface defects. However, most of the devices are curved surfaces with different shapes, large size spans, various types of surface defects and diffuse reflection imaging, and the background of machine vision images is complex, which greatly increases the difficulty of defect detection. Based on the principle of machine vision imaging, this paper designs a long depth of field and low distortion imaging system for curved surfaces. Combined with a flexible scanning motion mechanism with multiple degrees of freedom, this paper studies the technology of complex background defect extraction to realize the full surface defect detection of cross scale special-shaped metal devices.