| 摘要 |
Contour vectors are detected by an outline detecting unit based upon a pixel of interest and pixels adjacent thereto read in a raster scanning sequence. A vector-data creating unit receives the contour vector and input data, calculates the difference between starting-point coordinates of continuous vectors, fits the value of the difference in a data packet whose number of digits conforms to the value of the difference and stores the data packet. The data packet is additionally provided with a code representing the length of the packet. An outline smoothing/zooming unit returns the stored difference value to ordinary coordinate expression and performs smoothing or zooming processing. A binary-image reproducing unit converts the contour vector into raster representation, and the resulting data is outputted. The acquisition unit and the extraction unit or creating unit eliminate unnecessary vectors and perform smoothing by an outline (vector sensing/elimination) smoothing and zooming unit. By virtue of this arrangement, the amount of contour-line data is reduced to reduce the memory capacity needed. Further, in a case where contour points of a binary image are delivered to immediately following processing, it is possible to greatly reduce the load of subsequent processing. <IMAGE> |
