APPLICATION OF LEVEL SET METHOD IN IMAGE PROCESSING RELATED TO MEDICAL APPLICATIONS Peeyush Tewari, BIT Ranchi, INDIA Santosh Kumar Ray BIT Muscut, Oman, Sanjay Kumar, BIT Muscut, OMAN ABSTRACT While modeling a physical system having ordinary or partial order derivatives, researchers face a problem of finding approximations that provides good accuracy with minimum of computations. For Linear approximation Runge - Kutta methods, Taylor series methods and other method were used in literature due to their simplicity. However for complex problems Involving moving interfaces recently many methods have been developed such as level set methods and fast marching methods. These methods have shown great potentials in developing computer simulation as well as analyzing much complex geometries in medical and non-medical applications. This paper presents an application of level set method in combination with fuzzy technique for image segmentation. A hybrid image segmentation technique is presented based on combination of unsupervised clustering algorithm named as fuzzy c-means and moving boundary capturing technique by level set method. The two phase method takes contours around the desired objects in the first phase using fuzzy c-means and use this input to second phase which makes use of level set method. The proposed method is tested on various set of images and computing time has been found improved. KEYWORDS: Image processing, Level set method, Fuzzy means 1. INTRODUCTION X ray Images have been useful in understanding many medial diseases since last so many years. Image analysis has various constituents including image segmentation. Image segmentation plays an important role in many medical fields e.g. identifying anatomical structures, diagnosis and surgery etc. Generally an image segmentation is a process of partitioning an image into non- overlapping, constituent regions which are homogeneous based on some characteristics e.g. intensity, color, texture, etc. (Pham et al. 2000) [1], (Zuva et al. 2011) [2]. The goal of segmentation is to simplify and/or change the representation of an image into something that is more meaningful and easier to analyze (Shapiro and Stockman, 2001) [3]. Wrong identification or diagnosis can mislead the treatment of a patient including surgery in medical field. There are numerously recent developments in image segmenting methods but still one method is not suitable for all types of image segmentation requirements. Osher and Sethian (1988) [4] proposed one of the most powerful method known as Level set method to track the moving boundary. Later it has been used in image segmentation successfully in various fields of real life like medical, remote sensing, military etc. Sometime traditional level set method does not produce the correct result for image segmentation, especially when the desired object boundaries are weak or unclear in an image.