DocumentCode
2198908
Title
Research on the Color Reproduction Based on the BP Neural Network Algorithm
Author
Xu, Xiao-zhao ; Zhang, Xin-feng ; Cai, Yi-heng ; Shen, Lan-sun ; Liu, Chang-jiang
Author_Institution
Signal & Inf. Process. Lab., Beijing Univ. of Technol., Beijing
fYear
2008
fDate
20-22 Dec. 2008
Firstpage
256
Lastpage
260
Abstract
This paper proposes an algorithm of BP neural network, which reproduces a color image under D65 light conditions by using a pair of images (one with no D65 light source, and the other with D65 light source). The result of reproduction image which used a fluorescent lamp as an unknown light source is presented. This algorithm successfully removes the effect of the unknown light source color and achieves color constancy at a practical computational cost. The experiment results show that the validity of the proposed algorithm.
Keywords
backpropagation; image colour analysis; neural nets; back-propagation neural network algorithm; fluorescent lamp; image color reproduction; light source; Color; Computer networks; Digital cameras; Image sensors; Light sources; Neural networks; Pixel; Reflectivity; Sensor phenomena and characterization; Temperature sensors; BP neural network; color constancy; color reproduction;
fLanguage
English
Publisher
ieee
Conference_Titel
Advanced Computer Theory and Engineering, 2008. ICACTE '08. International Conference on
Conference_Location
Phuket
Print_ISBN
978-0-7695-3489-3
Type
conf
DOI
10.1109/ICACTE.2008.151
Filename
4736961
Link To Document