Multilook SAR Image Segmentation with an Unknown Number of Clusters Using a Gamma Mixture Model and Hierarchical Clustering

Abstract

This paper presents a novel multilook SAR image segmentation algorithm with an unknown number of clusters. Firstly, the marginal probability distribution for a given SAR image is defined by a Gamma mixture model (GaMM), in which the number of components corresponds to the number of homogeneous regions needed to segment and the spatial relationship among neighboring pixels is characterized by a Markov Random Field (MRF) defined by the weighting coefficients of components in GaMM. During the algorithm iteration procedure, the number of clusters is gradually reduced by merging two components until they are equal to one. For each fixed number of clusters, the parameters of GaMM are estimated and the optimal segmentation result corresponding to the number is obtained by maximizing the marginal probability. Finally, the number of clusters with minimum global energy defined as the negative logarithm of marginal probability is indicated as the expected number of clusters with the homogeneous regions needed to be segmented, and the corresponding segmentation result is considered as the final optimal one. The experimental results from the proposed and comparing algorithms for simulated and real multilook SAR images show that the proposed algorithm can find the real number of clusters and obtain more accurate segmentation results simultaneously.

Flexible Hierarchical Gaussian Mixture Model for High-Resolution Remote Sensing Image Segmentation

“…where Ω l = {µ l , σ l 2 } is the set of parameters for component l in GMM. Additionally, commonly used mixture models include SMM and GaMM [22,25], and their components are defined by the student's t and Gamma distributions, respectively. In this study, a new HGMM is proposed, where the components can accurately model the asymmetric, heavy-tailed, and multimodal distributions of pixel intensities in each object region.…”

“…Its component is a Gamma distribution, which becomes asymmetric and heavy tailed by changing its shape and scale parameters. GaMM is widely used to model the statistical distribution of pixel intensities in SAR image segmentation [24,25]. However, it still fails to solve the problem of modeling the multimodal distribution of pixel intensities.…”

Sea Ice Extraction in SAR Images via a Spatially Constrained Gamma Mixture Model

“…For example, Li et al [27] used GaMM to model the pixel intensity in SAR images and introduced it as a non-similarity measure into Fuzzy C-means (FCM); Akyilmaz [28] combined the GaMM and multilogit spatial interactive models for SAR image segmentation, and its advantage was its robustness. To reduce the impact of speckle noise in SAR images, the weight of GaMM is viewed as a Markov random field (MRF) [29,30], and the spatial information of local pixels can be modeled in prior probability. However, the method may increase the complexity of the model while improving robustness.…”

Shannon Entropy for the $\mathcal {G}^0_I$ Model: A New Segmentation Approach

“…Using multikernel sparse representation, Gu et al [7] furnished new segmenters for SAR imagery. Zhao et al [8] proposed a multilook SAR image segmentation algorithm using gamma mixture model (GaMM) and a Markov random field.…”

“…The proposed methods are compared with other three welldefined segmenters: k-means, GaMM with L known (GaMM-L), and with L estimated (GaMM-LE). These two last methods can be understood as particular cases of Wishart-based segmenter in [17] or resulting of Zhao et al [8]. In order to compare the previous methods, we use two figures of merit: accuracy percentage and kappa coefficient.…”