Abstract: Image is a major source of inspiration for color matching design,which is usually done intuitively and experientially at present.This study adopts technical means to aid color extraction from the source image,color selection,color assignment to the color zones in the color matching scheme and color adjustment,in order to improve the efficiency.The color adjacency network model was proposed,which took the extracted colors from the images as the nodes and the degree of adjacency between the color zones as the side.Two networks,namely,source network and target networks were built based on the source image and the target scheme respectively.Color matching problem was mapped into a sub-network that could best match the target network by searching in the source network.The degree of matching was assessed using the combination of two indices that were reverse ordinal number and number of different connections.The color matching schemes were optimized by using Interactive Genetic Algorithm(IGA).A color matching design system based on graphic design software was developed,which could be applied to the main sections of color matching design from color extraction to reuse.The proposed method was tested in a task of color matching for automobile appearance,and the results demonstrated the efficiency and effectiveness of the method.

Key words: product color design, color matching, color adjacent network, design reuse

摘要： 图像是设计师进行配色设计的重要灵感来源,通过技术手段对源图色彩提取、色彩选择到目标设计方案的色区赋色、调整等过程进行辅助,可以提高设计效率。提出了色彩邻接网络模型,以图像提取色为节点、以色区间的邻接紧密程度为边,为源图和目标配色对象分别建立源网和目标网两个网络模型。将配色问题映射为从源网中搜索可以最佳匹配目标网的子网,利用逆序数和异连接数两个指标的组合来评价匹配强度,并基于交互式遗传算法等技术实施配色方案的优化。基于平面设计软件开发了配色设计系统,实现了从色彩提取到重用的配色设计关键环节的辅助。以汽车外观配色为载体进行了应用测试,证明了所提方法的有效性。

关键词: 产品配色设计, 配色, 色彩邻接网络, 设计重用