A method for converting image data coded with run lengths to the format of a page description language, such as PostScript or PDF, the run lengths identifying how many image points of one color follow one another in an image row, includes utilizing the run lengths of the same color that overlap in successive image rows to form an object, and describing the object with operators from the page description language. Objects of the same color can be combined to form one object. An object is described in the page description language by the image mask operator and an associated bitmap, or by a polygon train that connects reference points on the edge of the object.