Texture based Segmentation Approches for Medical Images

Over the last few years, the atomization of medical diagnosis has increased exponentially. One of the most critical and first step of image analysis is segmentation. The segmentation results play a crucial and critical role in image analysis, such as representation, feature measurement, description, and even in classification and interpretation. The present study proposed various innovative segmentation methods for medical images. The proposed segmentation methods are based on the concepts derived from combinations of morphology, Discrete Wavelet Transform(DWT), Stationary Wavelet Transform(SWT), edge filters, thresholding and texture based methods. The advantages of morphological methods are Noise filtering, Skeletonizing, Thickening, Object marking, Shape simplification, Thinning, Convex hull, Segmenting objects from background, Quantitative description of objects (projections, area, perimeter). In This book several segmentation methods on tuberculosis(TB) images were discussed, TB is one of the chronic bacterial infection, which causes more deaths worldwide than any other infectious disease.