Unsupervised Segmentation of Textured Images

Abstract In this paper, a new unsupervised segmentation algorithm for textured images is developed. The proposed algorithm utilizes the characteristic of the Markov random fields (MRF) for modeling the contextual information embedded in image formation. Textured images are regarded as realizations of the stationary Gaussian MRF on a two-dimensional square lattice, and are modeled by the conditional autoregressive (CAR) equations with a second-order noncausal neighborhood. A hypothesis test over two radially masked areas is involved in our algorithm to detect the location of boundaries. The hypothesis assumes that these two areas belong to the same class of textures, and CAR model parameters are estimated in a minimum-mean-square-error (MMSE) sense. The hypothesis is tested by the analysis of variance (ANOVA)-like technique, and if it is rejected, a measure of dissimilarity is accumulated on the rejected area. This approach produces potential edge maps. From these maps, the boundaries among various textures can be detected without microedges. The performance of the proposed algorithm is demonstrated by some experiments using real textures as well as synthetic ones. The experiments show that the proposed algorithm can produce satisfactory segmentation without any a priori information.

[1]  Ludwik Kurz,et al.  Object detection and experimental designs , 1987, Comput. Vis. Graph. Image Process..

[2]  David B. Cooper,et al.  Simple Parallel Hierarchical and Relaxation Algorithms for Segmenting Noncausal Markovian Random Fields , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  J. Besag.  Spatial Interaction and the Statistical Analysis of Lattice Systems , 1974 .

[4]  Ludwik Kurz,et al.  Two-dimensional object detection in correlated noise , 1991, Pattern Recognit..

[5]  Donald Geman,et al.  Application of the Gibbs distribution to image segmentation , 1984, ICASSP.

[6]  J. Besag.  Efficiency of pseudolikelihood estimation for simple Gaussian fields , 1977 .

[7]  Haluk Derin,et al.  Modeling and Segmentation of Noisy and Textured Images Using Gibbs Random Fields , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[8]  Anil K. Jain.  Fundamentals of Digital Image Processing , 2018, Control of Color Imaging Systems.

[9]  Sridhar Lakshmanan,et al.  Simultaneous Parameter Estimation and Segmentation of Gibbs Random Fields Using Simulated Annealing , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[10]  H. Elliott,et al.  Stochastic boundary estimation and object recognition , 1980 .

[11]  J. Besag,et al.  On the estimation and testing of spatial interaction in Gaussian lattice processes , 1975 .

[12]  George G. Lendaris,et al.  Diffraction-Pattern Sampling For Automatic Pattern Recognition , 1969, Other Conferences.

[13]  B. Ripley.  Statistical inference for spatial processes , 1990 .

[14]  William B. Thompson,et al.  Computer Diagnosis of Pneumoconiosis , 1974, IEEE Trans. Syst. Man Cybern..

[15]  Rama Chellappa,et al.  Estimation and choice of neighbors in spatial-interaction models of images , 1983, IEEE Trans. Inf. Theory.

[16]  Rangasami L. Kashyap,et al.  Characterization and estimation of two-dimensional ARMA models , 1984, IEEE Trans. Inf. Theory.

[17]  Joseph M. Francos,et al.  A unified texture model based on a 2-D Wold-like decomposition , 1993, IEEE Trans. Signal Process..

[18]  Z. Fan,et al.  Textured image segmentation as a multiple hypothesis test , 1988 .

[19]  Ludwik Kurz,et al.  Edge detection in correlated noise using latin square masks , 1988, Pattern Recognit..

[20]  John W. Woods,et al.  Two-dimensional discrete Markovian fields , 1972, IEEE Trans. Inf. Theory.

[21]  Ivan Kadar,et al.  A class of robust edge detectors based on latin squares , 1979, Pattern Recognit..

[22]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[23]  Azriel Rosenfeld,et al.  Computer Methods in Image Analysis , 1977 .

[24]  Calyampudi R. Rao,et al.  Linear Statistical Inference and Its Applications. , 1975 .

[25]  R.M. Haralick,et al.  Statistical and structural approaches to texture , 1979, Proceedings of the IEEE.

[26]  B Julesz,et al.  Experiments in the visual perception of texture. , 1975, Scientific American.

[27]  Ludwik Kurz,et al.  Line Detection in Noisy and Structured Backgrounds Using Græco-Latin Squares , 1993, CVGIP Graph. Model. Image Process..