基于偏最小二乘回归分析的混色数据学习算法研究

PDF(217 KB)

电子学报 ›› 2001, Vol. 29 ›› Issue (3) : 429-431.

论文

基于偏最小二乘回归分析的混色数据学习算法研究

苏晓红, 郭茂祖, 王亚东, 张田文

作者信息 +

Research of Mix-Color Data Learning Algorithm Based on Partial Least-Square Regression

SU Xiao-hong, GUO Mao-zu, WANG Ya-dong, ZHANG Tian-wen

Author information +

文章历史 +

摘要

考虑到仅从单色色块中发现知识的片面性和不准确性,在对400组单色色块进行测试学习的基础上,本文提出用偏最小二乘回归分析从1331组混色数据空间中进一步发现知识,使得从RGB到CMYK转换时误差变小,从而提高色彩匹配的精度,进一步解决色彩匹配的自动化和通用性问题.

Abstract

Considering the one-sidedness and inaccuracy of knowledge discovery only from single-color database,an approach is proposed to discover knowledge from 1331 groups of mix-color database with partial least-square regression,based on measuring and learning 400 groups of single-color database.By this method,the mean error decreases when converting from RGB to CMYK,the precision of color matching is improved,and the automatic and general problem in color matching is further solved.

导出引用

苏晓红;郭茂祖;王亚东;张田文. 基于偏最小二乘回归分析的混色数据学习算法研究[J]. 电子学报, 2001, 29(3): 429-431.

SU Xiao-hong;GUO Mao-zu;WANG Ya-dong;ZHANG Tian-wen. Research of Mix-Color Data Learning Algorithm Based on Partial Least-Square Regression[J]. Acta Electronica Sinica, 2001, 29(3): 429-431.

中图分类号： TP391.4 TP181