Abstract: A method of processing an image takes a picture of an object by CCD camera, and calculates a center position of an edge of the image of the object formed on an image field to determine the size of the object. A differential between picture elements is formed from image data to detect an edge region. A differential for each picture element located with the edge region is converted into an edge existence width estimate which is centered about the associated picture element and into an edge probability estimate that that picture element is an edge center. These estimates are used to weight the x position of the respective picture element in the edge region. A center x position Ejcp of an assemblage of weighted positions of all the picture elements in the edge region is calculated and is defined to be the center of the edge region.