Application of a Mathematical Morphological Process and Neural Network for Unsupervised Texture Image Classification with Fractal Features
M. Talibi-Alaoui, Abderrahmane Sbihi · 2012
In this paper, we present a new texture image classification algorithm in an unsupervised context, which is based on both Kohonen Maps and Mathematical Morphology. As first part of the proposed algorithm, various features obtained from the fractal dimension computed using differential box counting method, are extracted from the texture image and then applied and projected into a Kohonen map which is represented by the underlying probability density function (pdf). Under the assumption that each modal region of the underlying pdf corresponds to a one homogenous region in the texture image, the second part of the algorithm consists in partitioning the Kohonen map into connected modal regions by making concepts of morphological watershed transformation suitable for their detection. The classification process is then based on the so detected modal regions.