A novel adaptive level set segmentation method.

Area of Science:

• Computer Vision
• Image Processing
• Medical Imaging

Background:

• Image segmentation is crucial for medical image analysis.
• Existing methods like ADPLS and LBF have limitations in speed, accuracy, or sensitivity to initial contours.
• Intensity inhomogeneity in images poses a significant challenge for segmentation accuracy.

Purpose of the Study:

• To develop a novel adaptive fusing level set method for image segmentation.
• To combine the strengths of the adaptive distance preserving level set (ADPLS) and local binary fitting (LBF) models.
• To improve segmentation accuracy, speed, and stability, particularly for images with intensity inhomogeneity.

Main Methods:

• A new adaptive fusing level set method is proposed.
• The method adaptively adjusts the weights of ADPLS and LBF based on image spatial information.
• The fusion strategy aims to leverage the speed of ADPLS and the accuracy of LBF.

Main Results:

• The proposed method significantly improves comprehensive performance indicators.
• Experimental results demonstrate enhanced accuracy compared to existing methods.
• Improved speed and stability were observed, making it more robust for image segmentation tasks.

Conclusions:

• The novel adaptive fusing level set method offers superior performance for image segmentation.
• This approach effectively addresses the limitations of previous ADPLS and LBF methods.
• The method shows great potential for applications requiring accurate and efficient image segmentation.