Active Contour Based on Local Statistic Information and an Attractive Force for Ultrasound Image Segmentation

This paper presents a new active contour model with local intensities through level set method for ultrasound images segmentation. The method is not affected by the limitation of Gaussian distribution. The model is designed by local intensities, alignment term with a sharpening edge coefficient and regularization. Local intensities have the capability of denoising, and local means and variances are considered. The alignment term with a sharpening edge coefficient can sharpen edge and increase the convergence speed. The numerical schedule is implemented by level set method. Experimental results show that proposed method succeed to segment edges for ultrasound images.