Segmentation of Images Using Level Set Analysis.
Partha Pratim Mohanta, Arjun Sen, Dipti Prasad Mukherjee · Indian Conference on Computer Vision, Graphics and Image Processing · 2002
We propose an image segmentation technique using level set analysis. Image level set is the binary decomposition of a gray level image. Connected components in the level set, less than a pre-defined size are removed from the level set. Based on level set topology an exposed connected component is defined in the level set. These exposed connected components are merged based on a proximity value derived between the neighboring components. This proximity measure is a function of intensity difference and the shared boundary length between image regions described by the connected components present in the level set. The result obtained using the proposed method is shown and compared with a similar process along with the accuracy measure of the segmentation.