Non-negative matrix factorization for spectral colors using genetic algorithms : Substantially Updated Version
semanticscholar(2017)
摘要
In this paper we introduce a novel method for nonnegative matrix factorization (NMF) using a genetic algorithm. The method finds the optimal basis functions for the spectral colors in both spectral and color spaces. We show that one version of the proposed algorithm works as well as the standard NMF algorithm in spectral space. Further, this algorithm is modified to obtain a functionality to work in color space which the standard NMF is currently not capable of providing. The modification involves optimization in color space reducing the approximation error by a factor of 6 for Macbeth ColorChecker colors. The algorithm can be used in digital camera design. In addition, we propose an algorithm based on multiobjective optimization in both spectral and color space which can be used in digital image archiving.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要