Boundary Detection in Piecewise Homogeneous Textured Images

We address the problem of scale selection in texture analysis. Two di erent scale parameters, feature scale and statistical scale, are dened. Statistical scale is the size of the regions used to compute averages. We de ne the class of homogeneous random functions as a model of texture. A dishomogeneity function is de ned and we prove that it has useful asymptotic properties in the limit of in nite statistical scale. We describe an algorithm for image partitioning which has performed well on piecewise homogeneous synthetic images. This algorithm is embedded in a redundant pyramid and does not require any ad-hoc information. It selects the optimal statistical scale at each location in the image.

• Type: Conference
• Year: 1992
• Where: ECCV
• Authors: Stefano Casadei, Sanjoy K. Mitter, Pietro Perona
• Keywords: Computer Vision, ECCV 1992, Erent Scale Parameters, Nite Statistical Scale, Optimal Statistical Scale, Scale Selection, Statistical Scale