A graph-theoretical approach to region detection

Abstract
This paper presents a general and computationally inexpensive algorithmic scheme that falls in a region detection category. In the scheme, a minimal spanning tree is used as a path of a sequential region grower. The algorithm traverses the spatially adjacent graph while maintaining the structural organization of an image. The graph-theoretical evaluation of heuristics is described and examples of implementation are given.

This publication has 13 references indexed in Scilit: