Rapid Detection And Identification Of SMD Components

SMT ( Surface Mount Technology ) is a integrated technology including SMD,equipment,jointing,etc.It mounts SMD on PCB according to a design. Machine vision takes 2D image by vision sensor, through image processing and image analyzing, gets the characters of the scenery, and uses the useful information to control the manipulator to motion. Machine vision system is one of key parts in the multi-function surface mount equipment. Chip device recognized is a key issue in the paper.According to the function, software and hardware of today's, machine vision system, vision recognize system in this paper include: research on image preprocessing arithmetic, analysis about SMT and SMD, research on SMD recognized arithmetic and recognized flow. The paper will study some theoretic and practical issures in SMD recognized process, and try to solve them.Image preprocessing is the first step in vision recognized process. Preprocessing method is very different according to originality image information and practical problems. This paper uses the selective local averaging for image smoothness. It includes image enhance based on four-neighborhood, eight-neighborhood and Butterworth high-pass filtering. The experiment result indicates that it is an available solution.SMD recognized arithmetic and flow are key issues in vision recognized system, and recognized precision and speed mainly depend on them. How to get the exact SMD information in the shortest time with precision, this paper tries to give the arithmetic and flow.After image processing, this paper will recognize SMD. For the measurement and identification of the size of its features, then with the standard template matching and judge whether the correct components to achieve the purpose of testing. It tries to make sure every device fitting the design requirement.