PCB bare board defect detection and recognition system based on machine vision

Abstract

PCB bare board as a carrier of electronic products to achieve the function, its quality directly determines the life of electronic products, and whether function is normal to achieve. This paper introduces a defect detection and recognition system of PCB bare board based on machine vision, and analyzes the overall design and composition principle of the system. The system consists of Renesas G2L embedded platform, X-Y axis mobile platform and high-definition industrial camera. Through the image preprocessing algorithm and the SSIM algorithm, the connected domain comparison is carried out to judge the fault. At the same time, the features such as the area of the connected domain are extracted and input into the BP neural network to judge the fault type, so as to realize the high-precision detection of the fault area, effectively identify the fault type and accurately locate the fault area. The overall system has good robustness.