Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
393176 | Information Sciences | 2015 | 21 Pages |
In this paper, we address the problem of simulating the human perception of texture properties in images. In particular, we have focused our study on the properties of coarseness, contrast and directionality, that play a fundamental role in the human perception of texture. The objective is not to precisely identify, classify or discriminate between different textures as a whole, but to be able to assess the presence of each texture property in the image. For this purpose, fuzzy sets defined on the domain of different groups of measures are employed in order to model each property by using parametric membership functions. The corresponding parameters are obtained by learning a functional relationship between the computational values given by the measures and the human perception. The performance of each fuzzy set is analyzed and tested with the human assessments, and a ranking of subsets of measures is obtained according to their ability to represent the perception of the property, allowing us to identify the most suitable combination of measures.