DocumentCode :
588956
Title :
A Print-Scan Modeling Scheme Based on BP Neural Network
Author :
Nan Luo ; Quan Wang ; Kefeng Zhang
Author_Institution :
Inst. of Comput. Sci., Xidian Univ., Xi´an, China
fYear :
2012
fDate :
17-18 Nov. 2012
Firstpage :
558
Lastpage :
562
Abstract :
Aims at the complexity of print-scan process, this paper presents a scheme to construct print-scan model using BP (Back Propagation) Neural Network principles. In this scheme, print-scan process is treated as a black-box. Firstly, divide RGB color space into several color blocks, get the color mapping table between the original colors and that after PS operation. Then obtain the trained net model by utilizing the BP Neural Network algorithm. At last, simulated PS (print-scan) image can be gained from the trained model by taking the original image as the input. Experimental results indicate the modeled image is quite close to the real PS one, which proves the great feasibility of this modeling scheme.
Keywords :
backpropagation; image colour analysis; neural nets; BP neural network algorithm; PS operation; RGB color space; back propagation; black-box; color blocks; color mapping table; print-scan modeling scheme; print-scan process complexity; trained model; Computational intelligence; Security; BP Neural Network; Color Mapping; Print-scan Model;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computational Intelligence and Security (CIS), 2012 Eighth International Conference on
Conference_Location :
Guangzhou
Print_ISBN :
978-1-4673-4725-9
Type :
conf
DOI :
10.1109/CIS.2012.131
Filename :
6406082
Link To Document :
بازگشت