DocumentCode
3138786
Title
Vector Concepts-Based Spectral Modelling
Author
Lukac, Rastislav ; Plataniotis, Konstantinos N.
Author_Institution
Dept. of Electr. & Compupt. Eng., Toronto Univ., Ont.
fYear
2006
fDate
38838
Firstpage
2009
Lastpage
2012
Abstract
This paper presents a unique spectral modelling approach suitable for digital camera image processing. The proposed vector model combines both the magnitude and directional characteristics of the vectorial RGB color inputs and enforces orientation constraints in an orthogonal color system to produce outputs which can simultaneously match luminance and chrominance characteristics of the captured image. Thus, the model enhances the performance of commonly used practical demosaicking solutions. Moreover, spectral models previously used in demosaicking can be seen as special cases of the proposed vector model
Keywords
cameras; image colour analysis; spectral analysis; vectors; RGB color input; chrominance characteristics; digital camera image processing; image demosaicking; luminance characteristics; orthogonal color system; spectral modeling; vector concept; CMOS image sensors; Color; Colored noise; Digital cameras; Educational institutions; Electronic mail; Laboratories; Semiconductor device modeling; Sensor arrays; Vectors; Color imaging; single-sensor digital cameras; spectral modelling; vector model;
fLanguage
English
Publisher
ieee
Conference_Titel
Electrical and Computer Engineering, 2006. CCECE '06. Canadian Conference on
Conference_Location
Ottawa, Ont.
Print_ISBN
1-4244-0038-4
Electronic_ISBN
1-4244-0038-4
Type
conf
DOI
10.1109/CCECE.2006.277313
Filename
4054783
Link To Document