Image Separations For Graphic Design And Collaborative Technology Research,

With the rapid development of the computer technology, computer pattern design technology, because of its high quality and fast speed, is replacing traditional manual pattern design and drawing in various fields. Computer aided pattern design systems are widely used in such industries as publishing, advertising, decorating, textile, chinaware, printing and dyeing, carpeting, weaving etc. Research and development on CAD, process planning automation and production information management in such fields have been carried out in succession home and abroad since 1980s. A series of new process plannings including draft designing and CAPP aiming at different products are proposed. Manual and experience-based process planning is transformed into modern IT-based process planning with integration of design, technique, plate making and management.Aiming at the large dimension and high accuracy pattern design, these exist a few key techniques hard to futher solve in theory and application. 1) Improve the precision of image color separation. Th image color separartion of the pattern design now have more and more high requests, not only requesting the color gorgrous, layering enrich, pattern clear, but also wanting to scan the kind draft handle varied, such as paper, photo and cloth etc, the image layer complications, increase the diffculty of the color separation. Some common image color separation techniques have some problems like limited color layers and blurred color bounds, hence making following design more difficult. 2) Satisfy the new arts and crafts requirement. Along with the development of pringting and dyeing techniques, The new arts and crafts such ascontinuous tone image color separation, spot color separation are appeared in printing and dyeing, porcelain and ceramics stamp-tax, silk screen printing etc. The color separation techniques based on pattern design are put forward the new challenge. 3) Improve the efficiency of the pattern design. The pattern design systems are now mostly the single-user deisgn mode. Because the size of pattern is too large and high precision, the efficiency lowers of pattern design, the longer of design life cycle, and which is difficult to meet the needs of enterprises for the model of "small quantity, diverse kinds". Traditional collaborative design technology can not satify the requirement of pattern design. Therefor, there is need to provide multi-user collaborative design environment, for the purpose of assurance the quantity of design, increase the efficiency of design availably.The dissertation is under the applied background, aiming at the above-mentioned problems, these two key techniques are significantly around and image color separation of the high accuracy pattern design to launch the research. The project is supportted by National Nature Science foundation of China (69873028) and National High-Tech Research and Development Program of China (863 Program No. 863-511-820-027) and.The contributions of this dissertation are:1.An algorithm of automatic image color separation by smoothing technique based on histogram is proposed. Given the number of image colors, the histogram of image is produced, by smoothing, center color and pixel mapping is automatically determined, and the color is accurately quantized. The algorithm is solving the problem of image color separation, improving color precision.2.An alogorithm of color images separation and printing for spotprinting is proposed. A combination of inks is selected by genetic algorithm according to the information of both original color images and given inks. Then a numerical optimization algorithm is used to compute the separations. With the ability of producing the same or more vivid reproduction than four processes color printing with just a small number of inks, this method can be widely applied in many industries such as, printing, dyeing and chinaware, etc.3. Aiming at the application of the large dimension pattern design, the collaborative design model based on pattern design is proposed. The model is adopted in hybrid architecture, the whole system is composed of edit workstation and application server, the large pattern data is stored distributedly in application servers. While collaborative designing, copy of pattern data is stored, the temporary data of the local pattern is stored in the local. The application server is responsible for managing and coordinating design between designers. The single index server mode is used, it manages pattern library indexes of all application server in a collaborative environment. Collaborative communication protocols are proposed. The method of adaptive area managing is used to solve concurrent control. It also describes the push and pull's functions to ensure the data integrity and consistency.The dissertation has focused on study on two key techniques based on distributed pattern design system, which are, high precision image color separation and collaborative design. These techniques have been used successfully in printing and dyeing CAD, weave CAD, chinaware CAD, and carpet CAD in the textile industry. The project has won the National Science and Technology Progress Award (Grade Two) in 2002.