A Spatial Segmentation Method

The problem of partitioning images into homogenous regions or semantic entities is a basic problem for identifying relevant objects. Visual segmentation is related to some semantic concepts because certain parts of a scene are pre-attentively distinctive and have a greater significance than other parts. However, even if image segmentation is a heavily researched field, extending the algorithms to spatial has been proven not to be an easy task. A true spatial segmentation remains a difficult problem to tackle due to the complex nature of the topology of spatial objects, the huge amount of data to be processed and the complexity of the algorithms that scale with the new added dimension. Unfortunately there are huge amount of papers for planar images and segmentation methods and most of them are graph-based for planar images. There are very few papers for spatial segmentation methods. The major concept used in graph-based spatial segmentation algorithms is the concept of homogeneity of regions. For color spatial segmentation algorithms the homogeneity of regions is color-based, and thus the edge weights are based on color distance. Early graph-based methods use fixed thresholds and local measures in finding a spatial segmentation. Complex grouping phenomena can emerge from simple computation on these local cues. As a consequence we consider that a spatial segmentation method can detect visual objects from images if it can detect at least the most objects. The aim in this paper is to present a new and efficient method to detect visual objects from color spatial images and to extract their color and geometric features, in order to determine later the contours of the visual objects and to perform syntactic analysis of the determined shapes. In this paper we extend our previous work for planar segmentation by adding a new step in the spatial segmentation algorithm that allows us to determine regions closer to it. The key to the whole algorithm of spatial segmentation is the honeycomb cells.