A depth-first search algorithm automatic initialization splitting of snakes
Liang Zhu, Martin D Fox · 2007
For object segmentation, the classical snake algorithms often require laborious human interaction; region growing methods are considerably dependent on the selected homogeneity criterion and initial seeds. In this paper we propose a new segmentation method for multi-object segmentation which is depth-first search algorithm based on GVF. The depth-first search process ends with a set of seeds scored and selected by considering local gradient direction information around each pixel. This step requires no human interaction; it enables our algorithm to segment objects which are separated from the background, while ignoring the internal structures of these objects. We have tested the proposed algorithm with several realistic images and obtained good results.