This paper deals with an innovative approach for achieving an automatic vision system for road surface texture classification. A road surface is composed by aggregates with a particular grain size distribution and the mortar matrix. From a vision point of view, road surface images can be described as a set of objects of high intensities puts on a uniform background of low intensity. Image analysis is based on the calculation of an original morphological pyramid, whose links structure is exploited through the computation of a multiresolution co-occurrence matrix. Such a matrix is computed between two consecutive pyramid levels and is analyzed in order to extract a threshold value. We obtain a set of threshold values along the pyramid. Analyzing a sequence of binary images carries out classification algorithm.