*Assistant Professor, Dr. Mahalingam College of Engineering and Technology, Pollachi, India
**Professor, Bannari Amman Institute of Technology, Sathyamangalam, India
***Assistant Professor, KSR College of Engineering, Erode, India
****Research Scholar, Bannari Amman Institute of Technology, Sathyamangalam, India
Online published on 9 December, 2016.
Medical images are primarily gray scale with wide range of intensity inhomogeneity. The segmentation of these images gives rise to a lot of problems. Usually active contour models have been widely used to segment images with intensity inhomogeneity. It is defined by a contour and two approximation functions. This energy function is subjected to level set formulation to cope up with the topographical changes. A kernel function is introduced so as to guide the contour on a controllable scale. This method ensures regularity of the level set function and also prevents unwanted re-initialization of the energy term. Experimental results for various types of medical images with varying scale parameters show the performance of our method is desirable.
Region Scalable Fitting energy, Level Set Fitting energy, Computed Tomography, Magnetic-Resonance imaging, Probability Density Function, Gaussian Kernel Function
