Vertex location for robotics applications

Machine vision, based on the processing and interpretation of images, is being introduced into many robotics applications. A high speed algorithm for extracting vertex location is an essential part of the performance of the system. The class of algorithms based on edge detection, small mask vertex extraction and matching line to determine the image coordinates of object surface is considered. The method consists of first performing edge detection on the image which obtained from the camera. An image of the object edges and the distorted grid edges is obtained. The resulting image is then thresholded at a pre-determined gray level to obtain a binary image. Using the small mask vertex extraction method, the vertices of the distorted grid are extracted from the binary image. The true image vertex can be located by using the matching line constraint. An important aspect of this algorithm is that the feature extraction is based upon local statistics to determine the vertices' location of the grid patterns. The significance of this technique is that it demonstrates an economical and high speed vertex location method to obtain the distorted grid pattern of the object surface. This technique plays an important role in many high speed surface measurements as well as in single image surface measurements for robotics applications. The technique has direct application to visually guided robotic systems and some medical treatments which require body surface measurements.