Improvement of Statistical and Fractal Features for Texture Classification
Texture classification and segmentation have been studied using various approaches. The mean Grey-Level Co-occurrence Matrix, introduced by the authors, gives statistical features relatively insensitive to rotation and translation. On the other hand, texture analysis based on fractals is an approach that correlates texture coarseness and fractal dimension. By combining the two types of features, the discrimination power increases. The paper introduces the notion of effective fractal dimension which is an adapting fractal dimension to classification of texture and is calculated by elimination of a constant zone which appears in all textured images. In the case of colour images, we proposed a classification method based on minimum distance between the vectors of the effective fractal dimension of the fundamental colour components. The experimental results to classify real land textured images validate that effective fractal dimension offers a grater discrimination of classes than typical fractal distance based on complete box counting algorithm.
Keywordstexture fractal dimension box-counting algorithm statistical features image processing texture classification
Unable to display preview. Download preview PDF.
- 1.Tuceryan, M., Jain, A.: Texture Analysis. In: Chen, C.H., Pau, L.F., Wang, P.S.P. (eds.) The Handbook of Pattern Recognition and Computer Vision, 2nd edn., pp. 207–248. World Scientific Publishing Co. (1998)Google Scholar
- 3.Olujic, M., Milosevic, N., Oros, A., Jelinek, H.: Aggressive Posterior Retinopathy of Prematurity: Fractal Analysis of Images before and after Laser Surgery. In: Proc. of 18th Int. Conf. on Control Systems and Computer Science, pp. 877–882. Politehnica Press, Bucharest (2011)Google Scholar
- 4.Shapiro, L., Stockman, G.: Computer Vision. Prentice Hall (2001)Google Scholar
- 8.Peitgen, H.O., Jurgens, H., Saupe, D.: Chaos and Fractals: New Frontiers of Science. Springer, New York (1992)Google Scholar
- 10.Carbone, A.: Algorithm to estimate the Hurst exponent of high-dimensional fractals. Physical Review E 76 056703/1 - 056703/7 (2007) Google Scholar
- 13.Dobrescu, R., Popescu, D.: Image processing applications based on texture and fractal analysis. In: Qahwaji, R., Green, R., Hines, E. (eds.) Applied Signal and Image Processing: Multidisciplinary Advancements, pp. 226–250. IGI Global Publishing (2011)Google Scholar
- 14.Popescu, D., Dobrescu, R.: Carriage road pursuit based on statistical and fractal analysis of the texture. International Journal of Education and Information Technologies 2(11), 62–70 (2008)Google Scholar
- 17.Popescu, D., Dobrescu, R., Angelescu, N.: Fractal Analysis of Textures Based on Modified Box-Counting Algorithm. In: Proc. of 18th Int. Conf. on Control Systems and Computer Science, pp. 894–898. Ed. Politehnica Press, Bucharest (2011)Google Scholar
- 18.Popescu, D., Dobrescu, R., Angelescu, N.: Colour textures discrimination of land images by fractal techniques. In: Proc. 4th Int. Conf. REMOTE 2008, Venice, pp. 51–56 (2008)Google Scholar