Perceptual meaning of the estimated parameters of the autoregressive model

This paper proposes an interpretation of the estimated parameters of the autoregressive model, used to model texture content of images, that corresponds to human perception. First, we define briefly the AR model and the parameters estimation process. A perceptual interpretation of the AR estimated parameters is then proposed and discussed. In particular, a computational measure to estimate the degree of randomness/regularity of textures is proposed based on the estimated parameters. Experimental results, obtained on the well-known Brodatz database of textures, support the proposed perceptual interpretation.