| 摘要 |
<p>A method for conversion of grayscale images with a non-white background into halftone images is disclosed. The illustrated embodiment of the present invention addresses the difficulties current conversion methods have in converting multi-bit grayscale images with non-white backgrounds into single bit halftone images. Through manipulation of the grayscale image data, the edges of text and line art features in the grayscale image data are located and extracted from the non-white background by performing calculations on subsets of pixels in horizontal rows (4) and vertical columns (6) cantered around a focus pixel. The extracted edge information is used to determine a weight factor to drive a threshold conversion algorithm (16), together with running horizontal (10), vertical (12), and local (14) pixel averages, resulting in a more defined halftone image output (22).</p> |
