Investigation Of Virtual Instrumentation-based Chip Identification System

With the development of chip industry, new chips will continue emerging and distinguishing different chips had became a time-consuming work. The common used template matching technology could not meet the demand of automatic chip identification system for its low accurate and template dependency. Optical character recognition technology which was widely utilized in computer-related and industrial automation fields is an effective way to solve the problem. To realize automatic chip identification, a system, based on image processing technology, is introduced in this thesis. This system could identify chips by read characters on its surface.Chip identification system was based on virtual instrumentation technology. The hardware system was constructed by high-performance PXI image acquisition card functioned in getting the chip image and PXI motion control card served to control the movement of stepper motor. In addition, a stepper motor driver was designed for drive chip carrier, which could enhance the load capacity of the pulse generated by NI motion control card. The software system was divided into two parts: image processing and movement control. The main purpose of image processing was to locate characters. The image processing flow includes: image calibration, edge detection, chip location, text location and so on. The image processing algorithms was combined into movement control system and was used to locate chip so that movement control system could move chip precisely. Combination of movement control and image processing could monitor the chip carrier and prevent movement implementation device from damage which was caused by invalid movement of stepper motor. The software of this thesis used a sequence structure and all the image processing and move control procedures were modularized. So the software could implement image processing step by step. This was helpful for error location of automatic chip identification system. In order to test the functions of the automatic chip identification system, two chips of different sizes and packages were used. Results showed that the automatic chip identification system was robust. The system was able to locate and move chips automatically and identify chip correctly.