An Enhanced Level Set Segmentation for Multichannel Images Using Fuzzy Clustering and Lattice Boltzmann Method

Abstract

In the last decades, image segmentation has proved its applicability in various areas like satellite image processing, medical image processing and many more. In the present scenario the researchers tries to develop hybrid image segmentation techniques to generates efficient segmentation. Due to the development of the parallel programming, the lattice Boltzmann met hod (LBM) has attracted much attention as a fast alternative approach for solving partial differential equations. In this project work, first designed an energy functional based on the fuzzy c-means objective function which incorporates the bias field that accounts for the intensity in homogeneity of the real-world image. Using the gradient descent method, corresponding level set equations are obtained from which we deduce a fuzzy external force for the LBM solver based on the model by Zhao. The method is speedy, robust for denoise, and does not dependent on the position of the initial contour, effective in the presence of intensity in homogeneity, highly parallelizable and can detect objects with or without edges. For the implementation of segmentation techniques defined for gray images, most of the time researchers determines single channel segments of the images and superimposes the single channel segment information on color images. This idea leads to provide color image segmentation using single channel segments of multi chann el images. Though this method is widely adopted but doesn’t provide complete true segmentation of multichannel ie color images because a color image contains three different channels for Red, green and blue components. Hence segmenting a color image, b y having only single channel segments information, will definitely loose important segment regions of color images. To overcome this problem this paper work starts with the development of Enhanced Level Set Segmentation for single channel Images Using Fuzzy Clustering and Lattice Boltzmann Method. For the implementation of the proposed method over color images the input color image will be first divided into red, green and blue single channels then by applying the proposed single channel technique on each channels of color image will lead to three different segmentation information for red, green and blue channels. After combining all the three segmentation information true color image segments will be obtained.

For the comparative analysis of the proposed segmentation scheme three segmentation parameters have been utilized, they are Probabilistic Rand Index (PRI), Variation of Information (VOI) and Global Consistency Error (GCE). A huge comparative analysis is performed for comparison on the basis of these three parameters. The results obtained clearly indicates that the proposed technique for color image segmentation is efficient as compare to existing technique in all the aspects of segmentation.