A neural network-based automated inspection system with an application to surface mount devices

Surface mount technology is becoming popular in electronic assembly. Since most assembly processes are already automated, it is also very important to automate the inspection process to streamline the entire production system. In this paper, an automated inspection system using a Hough transform and a back propagation neural network for surface mount devices is proposed. Experimental results show that the Hough transform can effectively reduce the amount of processing data while preserving all vital edge position and orientation information. The neural network is able to calssify the quality status with high performance. A comparison to a traditional template matching approach clearly shows that the proposed system is better in inspection accuracy. These encouraging results have demonstrated the feasibility of the proposed automated inspection system in potential applications in the electronics industry.