基于DFNN的语义驱动色彩智能设计研究-计算机工程与应用.PDF

Computer Engineering and Applications 计算机工程与应用 2014 ，50（3 ） 145 基于DFNN 的语义驱动色彩智能设计研究 周 晔，余隋怀，初建杰 ZHOU Ye, YU Suihuai, CHU Jianjie 西北工业大学 现代设计与集成制造技术教育部重点实验室，西安 710072 The Key Laboratory of Contemporary Design and Integrated Manufacturing Technology, Ministry of Education,Northwestern Polytechnical University, Xi ’an 710072, China ZHOU Ye, YU Suihuai, CHU Jianjie. Research on semantic-driven intelligent color design based on DFNN. Computer Engineering and Applications, 2014, 50 （3 ）：145-149. Abstract ：Color design is one of the important means through which the designers express their emotion. This paper quan- tifies the expressions of color semantics, and obtains the main factors by extracting the fuzzy rules. Considering those scarce multi-dimensional training samples, this paper constructs a Dynamic Fuzzy Neural Network （DFNN ）which is based on the generalized Radial Basis Function （RBF ）as the neuron activation functions to intelligently simulate the process of designing and choosing a color scheme. After training those data samples, this paper verifies the feasibility of this method by developing an intelligent color design prototype system, which can intelligently assist the designers to design the color of cars with great efficiency at a high-level. Key words ：Radial Basis Function （RBF ）; Dynamic Fuzzy Neural Network（DFNN ）; semantics; color design 摘 要：色彩是表达设计情感的重要手段，进行了色彩语义的量化处理，通过抽取模糊规则获得主要影响因素，根据 色彩方案多维小样本的特点，提出了基于广义径向基函数（RBF ）的动态模糊神经网络（DFNN ）方法，智能模拟色彩 方案的设计、选择过程。根据训练数据开发了色彩智能设计原型系统，结合汽车色彩智能设计进行了可行性验证， 能够在较高层次上辅助设计师进行色彩设计。 关键词：径向基函数；动态模糊神经网络；语义；色彩设计 文献标志码：A 中图分类号：TP391.72 doi ：10.3778/j.issn. 1002-8331.1204-0446 1 引言 在色彩语义的应用实例中主要有两种实现方法： [6-7] 在智能设计领域，色彩的智能化设计是一种相对繁 （1）将成功实例存入数据库，在需要时进行查询匹配 。