Design of a Machine Vision-Based Detection System for Semiconductor Refrigeration Sheets

Abstract

Semiconductor refrigeration sheets are widely utilized across various fields, often in the form of combined semiconductor refrigeration units. The performance of these combined units hinges on the consistency of the electrochemical characteristics of the individual semiconductor refrigeration sheets. Consequently, it is crucial to test and sort these sheets before they leave the factory.Visual inspection involves using machines to measure and judge, replacing human eyes. Visual detection converts the target into image signals via machine vision products, such as image capturing devices like CMOS and CCD sensors, and transmits these signals to a specialized image processing system. This system transforms the pixel distribution, brightness, color, and other information into digital signals. The image processing system then performs various operations on these signals to extract the characteristics of the target. Based on the discrimination results, it controls the equipment's movement on site.This mechanism is valuable for production, assembly, or packaging, providing significant benefits in detecting defects and preventing defective products from reaching consumers. This paper presents the design of a semiconductor detection system based on machine vision to detect and sort semiconductor refrigeration sheets.