Study on Correlation of Color Components Image in Different Color Spaces

Abstract:

Article Preview

Color image is an information substance with color components, among which exists correlation. In order to investigate the internal relevance of information in color image and to find out the dependence between them, to research the correlation of the color component image is significant. In order to investigate the relationship between the color space and the correlation of the component images, color spaces RGB/ LCH/ LAB/ OHTA/ YCC are selected and the correlation coefficients and cross correlations of the component images are computed and analyzed on MATLAB platform. The Result shows, that the statistical correlation coefficients of component images under RGB color space are the highest, while in OHTA color space the lowest are showed. The correlation coefficients under LAB and LCH are relative lower. In the opposite color spaces, the correlation coefficients of two opposite color components images are higher than the coefficients between the lightness and one of the opposite color component images. For the cross correlation of color component images, it shows a weak negative exponent relationship between pixel distance and cross correlation. The average cross correlation of component images in LCH space is obvious lower than in other spaces, while the levels of cross correlation in other spaces are similar. The relationship between cross correlation and color characteristics of image in RGB color space is closely, while in OHTA space, the difference of cross correlations among component images are usually small. In LCH space, the difference of cross correlations among component images is obvious, the cross correlation among chroma and the other components (lightness and hue) are much lower.

Info:

Periodical:

Edited by:

Ouyang Yun, Xu Min, Yang Li and Liu Xunting

Pages:

86-91

DOI:

10.4028/www.scientific.net/AMM.262.86

Citation:

Y. Jin et al., "Study on Correlation of Color Components Image in Different Color Spaces", Applied Mechanics and Materials, Vol. 262, pp. 86-91, 2013

Online since:

December 2012

Export:

Price:

$38.00

[1] Adrian Ford, Alan Roberts. Colour Space Conversions (Technical Report). London: Westminster University, (1998).

[2] Ohta Y, Kanade T, Sekai T. Color information for region segmentation [J]. Computer Graphics Image Process, 13(3), 1980: 222-241.

DOI: 10.1016/0146-664x(80)90047-7

[3] Zhou Xin Lun, Liu Jian, Liu Hua Zhi. Digital Image Processing [M]. National Defense Industry Press, (1986).

[4] Hiroshi Harashima. Image information compression [M]. Translated by Peiding XUE and Guonai XU, Beijing: Science Press, 2004: 53-61.

[5] Wang Peng fei, Jin Yang, Liu Zhen. Study on Components of Color Space and their Auto-correlation of Image [J]. China Printing and Packaging Study, 3(5), 2011: 34-39.

[6] Cheng Xiao Gang, Chen Qi Mei, Cheng Hao, Ming Wei. Optimal approximation model of autocorrelation function of digital image[J]. Journal on Communications, 32(10), 2011: 185-190.

[7] Sun Ming Lei, Zhang Rong, Zhu Xiao Feng, Zong Guang Hua. Normalized cross correlation computation for geometry image features [J]. Journal of Beijing University of Aeronautics and Astronautics, 34(2), 2008: 1441-1444.

[8] Sun Ming Lei, Zhang Rong, Zhu Xiao Feng, Zong Guang Hua. A Method of Image Resolution Calibration for Automatic Microscopic Zooming Based on NCCO [J]. Acta Optica Sinica, 28(6), 2008: 1117-1123.

In order to see related information, you need to Login.