Abstract—The goal of this work is to accurately detect and localize boundaries in natural scenes using local image measurements. We formulate features that respond to characteristic changes in brightness, color, and texture associated with natural boundaries. In order to combine the information from these features in an optimal way, we train a classifier using human labeled images as ground truth. The output of this classifier provides the posterior probability of a boundary at each image location and orientation. We present precision-recall curves showing that the resulting detector significantly outperforms existing approaches. Our two main results are 1) that cue combination can be performed adequately with a simple linear model and 2) that a proper, explicit treatment of texture is required to detect boundaries in natural images. Index Terms—Texture, supervised learning, cue combination, natural images, ground truth segmentation data set, boundary detection, boundary localization. 1

The information bottleneck (IB) method is an information-theoretic formulation for clustering problems. Given a joint distribution ¢¤£¦¥¨§�©� � , this method constructs a new variable � that defines partitions over the values of � that are informative about �. Maximum likelihood (ML) of mixture models is a standard statistical approach to clustering problems. In this paper, we ask: how are the two methods related? We define a simple mapping between the IB problem and the ML problem for the multinomial mixture model. We show that under this mapping the problems are strongly related. In fact, for uniform input distribution over � or for large sample size, the problems are mathematically equivalent. Specifically, in these cases, every fixed point of the IB-functional defines a fixed point of the (log) likelihood and vice versa. Moreover, the values of the functionals at the fixed points are equal under simple transformations. As a result, in these cases, every algorithm that solves one of the problems, induces a solution for the other. 1

Abstract. A new method for unsupervised image category clustering is presented, based on a continuous version of a recently introduced information theoretic principle, the information bottleneck (IB). The clustering method is based on hierarchical grouping: Utilizing a Gaussian mixture model, each image in a given archive is first represented as a set of coherent regions in a selected feature space. Images are next grouped such that the mutual information between the clusters and the image content is maximally preserved. The appropriate number of clusters can be determined directly from the IB principle. Experimental results demonstrate the performance of the proposed clustering method on a real image database.

The task of fitting parametric curve models to the boundaries of perceptually meaningful image regions is a key problem in computer vision with numerous applications, such as image segmentation, pose estimation, object tracking, and 3-D reconstruction. In this article, we propose the Contracting Curve Density (CCD) algorithm as a solution to the curve-fitting problem.

Abstract. A common approach to the image matching problem is representing images as sets of features in some feature space followed by establishing correspondences among the features. Previous work by Huttenlocher and Ullman [1] shows how a similarity transformation- rotation, translation, and scaling- between two images may be determined assuming that three corresponding image points are known. While robust, such methods suffer from computational inefficiencies for general feature sets. We describe a method whereby the feature sets may be summarized using the Stable Bounded Canonical Set (SBCS), thus allowing the efficient computation of point correspondences between large feature sets. We use a notion of stability to influence the set summarization such that stable image features are preferred. Fig. 1. A) Blob and ridge feature extraction with centroids of blobs and ridges denoted, B) Stable Bounded Canonical Set (SBCS) construction, C) Determine transformation, D) Outline shows transformation determined from SBCS. 1

We consider the problem of image segmentation by clustering local histograms with parametric mixture-of-mixture models. These models represent each cluster by a single mixture model of simple parametric components, typically truncated Gaussians. Clustering requires unsupervised inference of the model parameters, for which we derive a nested variant of the EM algorithm. This learning procedure is designed to deal with the large number of hidden variables required by the model. Results are presented for application of the algorithm to unsupervised segmentation of synthetic aperture radar (SAR) images.

Abstract-This paper details the implementation of three traditional clustering techniques (K-Means clustering, Fuzzy C-Means clustering and Adaptive K-Means clustering) that are applied to extract the colour information that is used in the image segmentation process. The aim of this paper is to evaluate the performance of the analysed colour clustering techniques for the extraction of optimal features from colour spaces and investigate which method returns the most consistent results when applied on a large suite of mosaic images.

This paper presents a new method based on the Expectation-Maximization (EM) algorithm that we apply for color image segmentation. Since this algorithm partitions the data based on an initial set of mixtures, the color segmentation provided by the EM algorithm is highly dependent on the starting condition (initialization stage). Usually the initialization procedure selects the color seeds randomly and often this procedure forces the EM algorithm to converge to numerous local minima and produce inappropriate results. In this paper we propose a simple and yet effective solution to initialize the EM algorithm with relevant color seeds. The resulting selfinitialised EM algorithm has been included in the development of an adaptive image segmentation scheme that has been applied to a large number of color images. The experimental data indicates that the refined initialization procedure leads to improved color segmentation. KEY WORDS Color segmentation, EM, initialization, and diffusion filtering 1.