JPS6467606A - Converting method for vision coordinates into robot coordinates - Google Patents

Abstract

PURPOSE:To easily convert the vision coordinates into the robot coordinates by using a jig that can perform input of the robot coordinate point to fetch this point into a vision viewfield. CONSTITUTION:A point A equivalent to an original point and a point B slightly distant from the point A are plotted on a screen 19 which is parallel to an XY plane of the robot coordinates by means of a marker of an offset jig 18 put into a robot hand 6. Then the distance of the vision coordinates is obtained from the vision coordinate value that is recognized out of the picture of a TV camera 1 which photographs both points A and B at one time. Thus a conversion coefficient between the vision and robot coordinates is obtained from the difference between the vision coordinate distance and the robot coordinates set at plotting. Furthermore the coordinate value of an optional point on the vision coordinates is converted into the tentative vision coordinate value based on the coordinate values of both points A and B and then into the robot coordinates with addition of the difference between the tentative vision coordinates and the robot coordinates. In such a way, the vision coordinates value is simply converted into the robot coordinate value.