Relaxation Labeling Technique Research Articles

Variations in relaxation labeling techniques

Several issues concerning the behavior of relaxation labeling processes are examined in this paper. The performance of region relaxation is shown to be affected by the form of compatibility coefficients and the configuration of local neighborhoods. Three variants of the algorithm are explored on an artificially generated test image: relaxation with simple unary compatibility coefficients, probabilistic relaxation with piecewise-linear compatibility coefficients derived from conditional probabilities, and a form of discrete relaxation without compatibility coefficients. The compatibilities based upon conditional probabilities yield the only version that preserves image detail associated with spatially thin structures.

Decomposition and decentralization techniques in relaxation labeling

The problem of assigning names to objects in a scene is central in image analysis. Relaxation labeling techniques provide ways of taking into account contextual information and evolve from decisions based upon local measurements only to decisions based also on a given “world model”. When the number of objects and the number of possible names grows larger as it does in practice these techniques become more and more difficult to use (the curse of dimensionality). In this paper we propose two classes of algorithms that reduce this difficulty. The main idea is to replace a “global” problem of large dimensionality with a series “local” problems of smaller dimensionality while insuring that the solutioos of the “local” problems are the same as those of the “global” ones. The notion of coordinability is central to this approach. The behavior of these algorithms is presented in a simple example.

Improving Consistency and Reducing Ambiguity in Stochastic Labeling: An Optimization Approach

We approach the problem of labeling a set of objects from a quantitative standpoint. We define a world model in terms of transition probabilities and propose a definition of a class of global criteria that combine both ambiguity and consistency. A projected gradient algorithm is developed to minimize the criterion. We show that the minimization procedure can be implemented in a highly parallel manner. Results are shown on several examples and comparisons are made with relaxation labeling techniques.

Disparity Analysis of Images

An algorithm for matching images of real world scenes is presented. The matching is a specification of the geometrical disparity between the images and may be used to partially reconstruct the three-dimensional structure of the scene. Sets of candidate matching points are selected independently in each image. These points are the locations of small, distinct features which are likely to be detectable in both images. An initial network of possible matches between the two sets of candidates is constructed. Each possible match specifies a possible disparity of a candidate point in a selected reference image. An initial estimate of the probability of each possible disparity is made, based on the similarity of subimages surrounding the points. These estimates are iteratively improved by a relaxation labeling technique making use of the local continuity property of disparity that is a consequence of the continuity of real world surfaces. The algorithm is effective for binocular parallax, motion parallax, and object motion. It quickly converges to good estimates of disparity, which reflect the spatial organization of the scene.

Scene labeling: an optimization approach

This paper deals with an optimization technique applied to the problem of stochastic labeling. We propose a definition of a global criterion on a set of objects to be labeled that combines both ambiguity and consistency with adjustable weights. A projected gradient algorithm is developed to minimize the criterion. Results are shown on a toy example and on the edge detection problem. Comparisons are made with relaxation labeling techniques.