PROCEDE ET DISPOSITIF D'EXTRACTION DE CONTOURS D'OBJETS AU MOYEN D'UNE CAMERA VIDEO

摘要

An object viewed by a video camera is identified by dividing the image into a set of lines and columns of points. Each point of the image is analysed in order to determine the points which characterise a change in direction of the contour of the object. The position of each of these points is stored in a first memory line by line. A second memory stores the address of characteristic points preceding and/or following the above point in order to obtain information linking points together. The positions representing successive points are put into order so as to generate signals representing the contours of the object. These signals are then processed to produce signals representing predetermined parameters of the viewed objects. The system is intended for use in robot operated machinery.