Detection of Faulty Integrated Circuits in PCB with Thermal Image Processing

Abstract

Developing a system which can be used for detection of faulty Integrated circuits (ICs) is one of the major challenge in electronic industry. Heating in ICs due to various reasons which may lead to degradation of performance and can cause serious hazardous effects in Printed Circuit Boards (PCB). Thermal image processing is one of the best non-contact, non-invasive method which can be used for IC fault detection. The paper aims towards detection of faulty ICs in PCB with help of thermal imaging camera and image processing techniques. Thermal images of sequence detector circuit for different fault conditions are collected. Image matching is achieved by comparing features of training images and test image using Speeded-Up Robust Features (SURF) algorithm. The system will indicate image belongs to particular fault class. The proposed method detects and classify possible faulty IC conditions with improved accuracy and can also help in early prevention remotely before complete circuit failure.