The aim of this paper is to build a joint image segmentation and bias field correction framework that unifies these recent models, and to develop an efficient numerical algorithm to solve the model efficiently. A variational model for intensity in homogeneity correction and segmentation of MR images is developed. The MR image is considered to have noise, bias and an ideal image within it. Main aim is to remove noise and bias. As an initial step the posterior probability distribution of the pixels is determined. Posterior probability is determined as per baye’s rule as a multiplicative structure of likelihood and prior information of pixels. Maximum a posteriori (MAP) is used to determine the optimal segmentation and intensity in homogeneity correction.
Volume 11 | 09-Special Issue
Pages: 469-481
DOI: 10.5373/JARDCS/V11/20192594