| 摘要 |
PROBLEM TO BE SOLVED: To provide a method for compressing a scanned document with color and/or gray scale by dividing a digital image of the scanned document into three images; a foreground image, a background image and a binary mask image. SOLUTION: A threshold image that can locally be varied is formed from a first document, a quantization image in bits is formed by using the first image through quantization (binary processing), the first image specifically interleaved is quantized, a sentence is detected (segmented) from the quantized image and the first image to divide the document into a foreground area and a background area, a specific area, especially a sentence part and a graphic part are attached to the foreground image, the background and the image of the sentence are attached to the background image, the result of the segmentation is stored in a binary mask image whose size is equal to the size of the first image, the foreground image that is formed from the first image and the mask image and whose resolution is decreased represents a color of the foreground area, the background image whose resolution is decreased is generated from the first image and the mask image, and then the binary mask image, the foreground image and the background image are coded by using a known image coder. |
