Paper Title
KL Transform Based Compressive Segmentation Using Modified Multiplicative Intrinsic Component Optimization

Abstract
Image segmentation is a base, to analyze a physical property of tissue in medical applications. Segmentation is a challenging aspect for researches in medical field. This paper proposes a KL Transform based compressive segmentation using modified Multiplicative model with energy minimization. For effective segmentation and Bias correction simultaneously with less iterations are achieved on the basis of lossless compression by extracting significant energy components. With this proposal noisy components present in the Magnetic Resonance (MR) Image can also be reduced. The proposed method represents an Image into two Multiplicative components, first component is a reflected intensity variations of the image that characterizes the physical property of tissues and is defined as a random variable varying from zero to one, which results soft segmentation. The second component is the intensity in homogeneity in the bias field and is defined as a linear combination of optimal coefficients with some basis function. This methods provides better tissue segmentations in the energy minimization process with existing methods. Observed results are shows that it obtained better bias field correction and 3 iterations are enough for proper segmentation for different percentage of compression. Keywords� KL Transform, MRI, Bias field estimation, correction and Tissue segmentation.