A NEW ALGORITHM OF SPOT COLOR SEPARATION

摘要

The spot color separation is a fundamental technique in pattern processing of many industries such as printing, dyeing and ceramics, etc. The result of the spot color separation has a strong influence on the efficiency of the whole pattern processing. But the results of the methods which are prevailing in pattern processing of these industries are based on the pixels, and contain too much unmeaningful information which increases the difficulty of the latter processing, such as editing and technology processing, and terribly delays the whole course of pattern designing. In this paper, a new method of spot color separation is proposed to resolve these problems. To achieve the aim, the edges detected in the original image are mainly used to prune the regions formed by the image segmentation. Our method can eliminate the noises and the little jumbled regions automatically to get the clear result with smooth boundary; above all. the outline of the meaningful regions is well preserved. After the color separation, we obtained the important information of boundaries and regions that give much convenience to latter processing. Consequently, the efficiency of pattern processing is improved greatly. Our method is very suitable for separation of pure color image or image presented as regions in the pattern processing of the industries.