Research On Technology Of Machine Vision Inspection And Locating For IC Manufacturing Equipment

With the requirements of intelligence and high accuracy in industry, a new subject,machine vision, has been proposed. And more and more researches on this subject are beingimplemented. This thesis discusses hardware structure and optical components selection ofmachine vision system, and researches the application of visual inspection and location onchips manufacturing equipment, such as QFP chip test equipment and RFID tag packagingequipment.Some optical components are selected after relative calculation according to theapplication requirements of RFID tag packaging equipment. Then a new vision systemconsists of these optical components on the vision test platform. A number of experimentsgive the evidences that prove the vision system can satisfy the equipment’s demands.This thesis designs an automatic inspection system based on machine vision to inspect theappearance of the QFP chip lead. In this thesis, optical path design and hardware structure areintroduced in brief. After that, we discuss the inspection method of standoff and the detectingmethod based on three-point seating plane of co-planarity, which are both based edgedetection. Finally, the test experiments with respect to the QFP chip visual inspection systemare carried out, which achieves satisfactory results very well.This thesis proposes a new decision-making based hausdorff distance matching algorithmto meet the demand of vision locating in RFID tag packaging equipment. When onlytranslation occurs between image and model, there is a proved inequality related to hausdorffdistance. Based on the property of distance inequality, the new matching algorithm isproposed. With applying to the vision locating system of RFID tag packaging equipment, thealgorithm indicates a good robustness and efficiency simultaneously.