A fast level set method for inhomogeneous image segmentation with adaptive scale parameter.

Area of Science:

• Medical Image Analysis
• Computer Vision

Background:

• Image segmentation is challenging with intensity inhomogeneity, where object regions overlap.
• Existing methods struggle with accurate segmentation and bias field estimation in such scenarios.

Purpose of the Study:

• To present a novel level set method for segmenting inhomogeneous images.
• To simultaneously estimate the bias field and improve segmentation accuracy.

Main Methods:

• A new region-based pressure force function within a level set formulation.
• Derivation of an optimal segmentation plane based on an inhomogeneous image model.
• Introduction of a new bias field initialization and adaptive scale parameter for kernel function.

Main Results:

• The proposed method accurately segments inhomogeneous images.
• Simultaneous estimation of the bias field is achieved.
• Improved robustness to initial contour and accurate bias field estimation demonstrated.

Conclusions:

• The novel level set method offers superior accuracy, efficiency, and robustness for inhomogeneous image segmentation.
• The technique effectively handles intensity overlap and bias field variations.
• Experimental results validate the method's performance on synthetic and real images.