Computer vision in industrial automation

This study focuses on the visional needs of automation processes. In application, some type of image processing algorithm use with camera and computer to make economic, modern and real-time solutions.In the first application, analyses of images captured via stereo cameras for detect the distance of objects. For this aim, we use corner detection, pixel comparison and disparity of cross-correlation algorithm. In the second application, we achieve to detect 6 different colors of ball on conveyer bands and throw outside in required order. Image of balls that moving over conveyer bands captured via digital camera and capturing software. Images analyses in computer with improvement, filtering and color detection algorithms. Sequence of the balls are send to PLC in serial communication and PLC?s digital I/Os use for active pistons. In the third application, we take images from air model helicopter. Helicopter carries a CMUCam2 image processing product. The product can detect region of searched color.During the research and developments of the thesis, the needs of the automation systems which can be solved with vision systems are taken into consideration.Keywords: Stereo vision, image processing, corner detection, pixel comparison, color detection.