Image segmentation is a fundamental problem in computer vision. Despite many years of research, general purpose image segmentation is still a very challenging task because segmentation is inherently ill-posed. Among different segmentation schemes, graph theoretical ones have several good features in practical applications. It explicitly organizes the image elements into mathematically sound structures, and makes the formulation of the problem more flexible and the computation more efficient. In this paper, we conduct a systematic survey of graph theoretical methods for image segmentation, where the problem is modeled in terms of partitioning a graph into several sub-graphs such that each of them represents a meaningful object of interest in the image. These methods are categorized into five classes under a uniform notation: the minimal spanning tree based methods, graph cut based methods with cost functions, graph cut based methods on Markov random field models, the shortest path based methods and the other methods that do not belong to any of these classes. We present motivations and detailed technical descriptions for each category of methods. The quantitative evaluation is carried by using five indices -Probabilistic Rand (PR) index, Normalized Probabilistic Rand (NPR) index, Variation of Information (VI), Global Consistency Error (GCE) and Boundary Displacement Error (BDE) -on some representative automatic and interactive segmentation methods.

This paper addresses the automatic image segmentation problem in a region merging style. With an initially over-segmented image, in which the many regions (or super-pixels) with homogeneous color are detected, image segmentation is performed by iteratively merging the regions according to a statistical test.There are two essential issues in a region merging algorithm: order of merging and the stopping criterion. In the proposed algorithm, these two issues are solved by a novel predicate, which is defined by the sequential probability ratio test (SPRT) and the maximum likelihood criterion. Starting from an over-segmented image, neighboring regions are progressively merged if there is an evidence for merging according to this predicate.We show that the merging order follows the principle of dynamic programming. This formulates image segmentation as an inference problem, where the final segmentation is established based on the observed image. We also prove that the produced segmentation satisfies certain global properties. In addition, a faster algorithm is developed to accelerate the region merging process, which maintains a nearest neighbor graph in each iteration. Experiments on real natural images are conducted to demonstrate the performance of the proposed dynamic region merging algorithm.

No abstract

No abstract