Vision Positioning Method And System Development For Wire Bonder

Aimed at achieving a machine vision system with high speed, high accuracy and a friendly human-computer interface for IC packing, in this case, the algorithm of vision system is studied in detail. At the same time, it involves in the development and achievement of vision system. The main works and contributions in this dissertation are as follow:(1) Based on analyzing vision system of wire bonder, the enterprise and structure of the vision system were designed. Furthermore, according to the system’s requirement, the workflow of the system was designed.(2) Based on analyzing the requirements of wire bonder and the feature of the chip’s image, an image segmentation method based on the 1-D histogram was present, it got the reasonable threshold by OTSU and genetic algorithm; an image segmentation method based on the 2-D histogram and an fast realization method for it were present, it got the reasonable threshold by 2-D OTSU.(3) Based on analyzing the integration method for LabVIEW, the complex algorithms was called and achieved in LabVIEW by using Dynamic-Link Library.(4) According to the requirement of wire bonder, the requirement of the software was present. The vision system software was developed by using the G language and the C++ language. An easy operating workflow and a friendly human-computer interface were designed.(5) The contrast experiment of the different methods was completed on the XY table. Results showed that both the method based on GA and OTSU and the method based on 2-D OTSU could meet the requirement of the wire bonder.