Development of Visual-Based Automatic Assembly Microfactory

• Main Authors: Ching-Chang Chen, 陳慶昌
• Other Authors: Ren-Jung Chang
• Format: Others
• Language: zh-TW
• Published: 2006
• Online Access: http://ndltd.ncl.edu.tw/handle/04431384213737776963

碩士 === 國立成功大學 === 機械工程學系碩博士班 === 94 === In this thesis, the micro-compliant gripper developed by “Opto-Mechatronic System Control Laboratory” is utilized to develop a PC-based visual servoing system, which utilizes regional edge statistics( RES) algorithm, for automatic assembly in microfactory. At first, the micro gripper system has a piezo-drived compliant micro gripper which can grasp an object with diameter 60~80μm. Next, the visual system utilizes the DSP C6000 of TI to develop an image processing and regional edge statistics algorithm. Through using Real-Time Data eXchange (RTDX ), The DSP and IPC can exchange data immediately and used for visual servoing control . Finally, the motion system including motion stage, assembly stage and glue stage and is constructed to provide a motion space with 7mm×5.76mm×15mm. The resolution of the motion stage, with maximum positioning accuracy of 1.3μm and the minimum of 0.5μm is expected for precise positioning and assembly. In this thesis, organic glue is utilized to assemble two components with diameter 60μm and 380μm, respectively and the RES algorithm is employed to realize the visual-based automatic assembly system. In addition, the minimum distance classifier is also adopted and used for successful object recognition and classification in assembly task.