A computer vision system provides a powerful sensory tool for robot control and automated inspection. In this thesis, a simple computer vision system is developed to recognize the position and size of holes drilled in production processes.
The system to digitize the analog image includes a CCTV camera, a hybrid control circuit, and a MINC-11 microcomputer with an A/D converter and a Schmitt Trigger. The hole recognition algorithms adopted are the thresholding using gradient and least squares circular curve fitting method.
Experimental results show that the computer vision system developed is capable of recognizing the position and size of test holes within the error bound of image resolution due to digitization.