MA Yu-jie

A New Color Image Quantization Algorithm Based on Fuzzy Kernel Clustering

Full Text (PDF, 280KB), PP.16-23

Views: 0 Downloads: 0

Author(s)

MA Yu-jie ^1,*

1. Department of Computer Science, Shangqiu Normal Univeristy, Shangqiu, China

* Corresponding author.

DOI: https://doi.org/10.5815/ijeme.2011.02.03

Received: 3 May 2011 / Revised: 8 Jun. 2011 / Accepted: 22 Jul. 2011 / Published: 29 Aug. 2011

Index Terms

Image, Quantization, Fuzzy Kernel Clustering, Octree, Fuzzy C-means

Abstract

A new color image quantization algorithm based on fuzzy kernel clustering is studied in this paper. Firstly, the original image is quantized to 256 colors using the octree algorithm. Secondly, based on the quantitative relation of the NBS distance and the color difference of human vision, the initial clustering centers and number are determined automatically. Thirdly, the clustering center color values are modified by use of the fuzzy kernel clustering algorithm in the Munsell space. And then, a color image quantization effect is achieved. At last, simulations are performed on the presented algorithm, and the simulation result shows that the presented algorithm not only can solve the problem of giving the number of quantization in advance but also has better quantization effect than the octree algorithm and fuzzy c-means algorithm in the same quantization number.

Cite This Paper

MA Yu-jie,"A New Color Image Quantization Algorithm Based on Fuzzy Kernel Clustering", IJEME, vol.1, no.2, pp.16-23, 2011. DOI: 10.5815/ijeme.2011.02.03

Reference

[1]Hideo Kasuga. Color quantization using the fast k-means algorithm [J]. Systems and Computers, 2000, 31(8):1120-1128.

[2]Ozdemir D, Akarun, L. A fuzzy algorithm for color quantization of images [J]. Pattern Recognition, 2002, 35, 1785-1791.

[3]Heckbert P. Color image quantization for frame buffer display [J]. Computer Graphics, 1982, 16(2):297-307.

[4]Gervautz M, Purgathofer W. A simple method for color quantization: octree quantization [C] Proceeding of Graphics Gems International. San Diego: Academic Press Professional, 1998, 8(6):219-230.

[5]Wang Xangyang, Hu Fengli, Liu Chunhui. A new adaptive color image quantization algorithm [J] Journal of Liaoning Normal University, 2007, 30(3):310-314.

[6]Wu Zhongdong, Gao Xinbo, Xie Weixin. A study of a new fuzzy clustering algorithm based on the kernel method [J]. Journal of Xidan Uninersity, 2004, 31(4):533-537.

[7]Ma W Y, Manjunath S. Edgeflow: A framework for boundary detection and image segmentation [J]. IEEE Trans on Image Processing，2000，9(8):1375-1388.

[8]Gong Y H, Proietti G. Image indexing and retrieval based on human perceptual color clustering[C]. The International Conference on Computer Vision, Mumbai, 1998.

[9]Zhang D Q, Chen S C.Fuzzy C-means and possibilistic C-means algorithms under kernel based robust metric [J].Pattern Recognition and Artificial Intelligence, 2004, 17(4):390-395.

[10]Wu K L, Yang M S. Alternative c-means clustering algorithm [J]. Pattern Recognition, 2002, 35 (10):2267-2278.

International Journal of Education and Management Engineering (IJEME)