A cellular coevolutionary algorithm for image segmentation

Abstract

Abstract-Clustering is inherently a difficult problem, both with respect to the definition of adequate models as well as to the optimization of the models. In this paper we present a model for the cluster problem that does not need knowledge about the number of clusters a priori. This property is among others useful in the image segmentation domain, which we especially address. Further, we propose a cellular coevolutionary algorithm for the optimization of the model. Within this scheme multiple agents are placed in a regular two-dimensional (2-D) grid representing the image, which imposes neighboring relations on them. The agents cooperatively consider pixel migration from one agent to the other in order to improve the homogeneity of the ensemble of the image regions they represent. If the union of the regions of neighboring agents is homogeneous then the agents form alliances. On the other hand, if an agent discovers a deviant subject, it isolates the subject. In the experiments we show the effectiveness of the proposed method and compare it to other segmentation algorithms. The efficiency can easily be improved by exploiting the intrinsic parallelism of the proposed method.

Parallelized segmentation of a serially sectioned whole human brain

“…Definitely, there exist other sophisticated methods, e.g., Markov random field based segmentation [6,44,18,14], multilevel, resp., scale-space approaches [20], anisotropic filtering [3] segmentation based on genetic algorithms [47,13,22], neural networks [48,1], hybrid segmentation techniques [32], statistical procedures [19,49,2] and adaptive weight smoothing [33]. A qualitative and quantitative comparison of conventional segmentation operators was tested for small samples (Fig.…”

Object segmentation using block based patterns

“…As there exists a huge number of objects having different variations among them, these objects are only differentiable if they have different appearance and visually distinct from each other. Since the number of clusters is neither fixed nor manually provided, the minimization of the intra-region variability and the maximization of the inter-region variability in the union of two regions are considered in our algorithm like [16]. However, both the straight minimization of the intra-region variability and the maximization of the inter-region variability lead to undesirable trivial solutions being N regions or 1 region respectively.…”

Object segmentation based on split and merge algorithm

“…Since, the number of clusters in segmentation algorithm is either not fixed or not manually provided, the minimization of the intra-region variability and the maximization of inter-region variability in the union of two regions are considered [8]. In this way, the inter-region variability constraint defines the scale at which two regions can be differentiated from each other.…”