This thesis presents a precise material placement method based on real-time compensation using vision inspection to resolve the positional errors that occur in the process of placing material in semiconductor equipment.
The configuration and operating principle of the material placement system are examined. The theoretical background of image-based position measurement and motor-control-based position compensation, as well as the limitations of existing techniques, are analyzed. A compensation system is configured. The system calculates the error between the center position measured by vision inspection at the moment of pickup and the target position. The calculated error is reflected in the control values of the motors that move in the vertical and horizontal directions. The hardware configuration of the system and the image processing algorithm are described.
The placement performance before and after applying the error compensation function is compared and analyzed under various material conditions and positional deviations. The improvement in accuracy is evaluated quantitatively. The feasibility of the applied method is verified through the experimental results.