![](bg6.png)
learning method using distances to selected K win-
ners, and in the second step those K units compare
their distances factorized by their magnitude values.
The model is parallel and can be executed on-line.
MSCL is compared with other VQ approaches in
four examples of image color quantization with dif-
ferent goals: focused on image foreground, avoiding
mean color, focused on saliency and text image bina-
rization. The results show that MSCL is more versa-
tile than other competitivelearning algorithms mainly
focused on density representations. MSCL forces the
units to distribute their color prototypes following any
desired property expressed by the appropriate magni-
tude of the data image. So, MSCL selects more colors
in the palette to accurately represent certain interest-
ing zones of the image, or generates palettes focused
on less represented, but interesting colors.
ACKNOWLEDGEMENTS
This work is partially supported by Spanish Grant
TIN2010-20177 (MICINN) and FEDER and by the
regional government DGA-FSE.
REFERENCES
Ahalt, S., Krishnamurthy, A., Chen, P., and Melton, D.
(1990). Competitive learning algorithms for vector
quantization. Neural networks, 3(3):277–290.
Atsalakis, A. and Papamarkos, N. (2006). Color reduc-
tion and estimation of the number of dominant col-
ors by using a self-growing and self-organized neu-
ral gas. Engineering Applications of Artificial Intelli-
gence, 19(7):769–786.
Bezdek, J. (1981). Pattern recognition with fuzzy objective
function algorithms. Plenum Press.
Bishop, C., Svens´en, M., and Williams, C. (1998). GTM:
The generative topographic mapping. Neural compu-
tation, 10(1):215–234.
Celebi, M. (2009). An effective color quantization method
based on the competitive learning paradigm. In Pro-
ceedings of the International Conference on Image
Processing, Computer Vision, and Pattern Recogni-
tion, IPCV, volume 2, pages 876–880.
Celebi, M. (2011). Improving the Performance of K-
Means for Color Quantization. Image (Rochester,
N.Y.), 29(4):260–271.
Celebi, M. and Schaefer, G. (2010). Neural Gas Clustering
for Color Reduction. In Proceedings of the Interna-
tional Conference on Image Processing, Computer Vi-
sion, and Pattern Recognition, IPCV, volume 1, pages
429–432.
Chang, C., Xu, P., and Xiao, R. (2005). New adaptive color
quantization method based on self-organizing maps.
Neural Networks, IEEE, 16(1):237–249.
Cheng, G., Yang, J., Wang, K., and Wang, X. (2006). Im-
age Color Reduction Based on Self-Organizing Maps
and Growing Self-Organizing Neural Networks. 2006
Sixth International Conference on Hybrid Intelligent
Systems (HIS’06), pages 24–24.
Dekker, A. (1994). Kohonen Neural Networks for Optimal
Colour Quantization. Network: Computation in Neu-
ral Systems, 3(5):351–367.
Durbin, R. and Willshaw, D. (1987). An analogue approach
to the travelling salesman problem using an elastic net
method. Nature, 326(6114):689–691.
Itti, L. and Koch, C. (2001). Computational modeling
of visual attention. Nature reviews neuroscience,
2(3):194–203.
Koch, C. and Ullman, S. (1985). Shifts in selective visual
attention: towards the underlying neural circuitry. Hu-
man Neurobiology, 4(4):219–227.
Kohonen, T. (2001). Self-Organizing Maps. Springer.
Lazaro, J., Arias, J., Martin, J., Zuloaga, A., and Cuadrado,
C. (2006). SOM Segmentation of gray scale images
for optical recognition. Pattern Recognition Letters,
27(16):1991–1997.
Lloyd, S. (1982). Least squares quantization in PCM. IEEE
Transactions on Information Theory, 28(2):129–137.
Martinetz, T., Berkovich, S., and Schulten, K. (1993).
‘Neural-gas’ network for vector quantization and its
application to time-series prediction. IEEE Transac-
tions on Neural Networks, 4(4):558–569.
Nikolaou, N. and Papamarkos, N. (2009). Color reduction
for complex document images. International Journal
of Imaging Systems and Technology, 19(1):14–26.
Papamarkos, N. (2003). A neuro-fuzzy technique for doc-
ument binarisation. Neural Computing and Applica-
tions, 12(3-4):190–199.
Pelayo, E., Buldain, D., and Orrite, C. (2012). Magnitude
Sensitive Competitive Learning. In 20th European
Symposium on Artificial Neural Networks, Comp. Int.
and Machine Learning, volume 1, pages 305–310.
Treisman, A. and Gelade, G. (1980). A feature integration
theory of attention. Cognitive Psychology, 12:97–136.
Uchiyama, T. and Arbib, M. (1994). Color image seg-
mentation using competitive learning. IEEE Trans-
actions on Pattern Analysis and Machine Intelligence,
16(12):1197–1206.
Vazquez, E., Gevers, T., Lucassen, M., van de Weijer,
J., and Baldrich, R. (2010). Saliency of color im-
age derivatives: a comparison between computational
models and human perception. Journal of the Opti-
cal Society of America. A, Optics, image science, and
vision, 27(3):613–21.
Xu, L., Krzyzak, A., and Oja, E. (1993). Rival Penalized
Competitive Learning for Clustering Analysis, RBF
net and Curve Detection. IEEE Tr. on Neural Net-
works, 4:636–649.
FocusedImageColorQuantizationusingMagnitudeSensitiveCompetitiveLearningAlgorithm
521