Detection of Tumor Using Image Segmentation

Abstract

Image segmentation is an important and challenging factor in the medical image segmentation. Detection of cancer tumors has variety of segmentation problems. Filters are being used to remove the noise and generate effective features for segmentation purpose. Breast cancer is currently one of the leading causes of death among women worldwide. This thesis presents an approach for detecting breast tumor not only the detection, an early stage of tumors can also detectable. The difficulty of image segmentation found is the misclassified pixels which lead to ambiguity at correct detection of boundary. The effective classification is required to correct the error and fix the boundary to locate the exact spreading of Cancer Tumor and remove the boundary errors around it. Selection of proper segmentation method is the prime purpose of this work. The thesis is totally concerned to the segmentation of Mammograms of breast and the analysis and performance of different segmentation technique that can be employed in the segmentation process. The different segmentation technique including edge based segmentation, thresholding, region-based approached, clustering are used to generate useful outcome.