Texture Analysis Methods for Medical Image Characterisation

Texture analysis refers to the branch of imaging science that is concerned with the description of characteristic image properties by textural features. However, there is no universally agreed-upon definition of what image texture is and in general different researchers use different definitions depending upon the particular area of application In this chapter texture is defined as the spatial variation of pixel intensities, which is a definition that is widely used and accepted in the field. The main image processing disciplines in which texture analysis techniques are used are classification, segmentation and synthesis. In image classification the goal is to classify different images or image regions into distinct groups (Pietikainen, 2000). Texture analysis methods are well suited to this because they provide unique information on the texture, or spatial variation of pixels, of the region where they are applied. In image segmentation problems the aim is to establish boundaries between different image regions (Mirmehdi et al., 2008). By applying texture analysis methods to an image, and determining the precise location where texture feature values change significantly, boundaries between regions can be established. Synthesising image texture is important in three-dimensional (3D) computer graphics applications where the goal is to generate highly complex and realistic looking surfaces. Fractals have proven to be a mathematically elegant means of generating textured surfaces through the iteration of concise equations Conversely the ability to accurately represent a textured surface by a concise set of fractal equations has led to significant advances in image compression applications using fractal methods In Fig. The right image is a grey-level texture map showing the ideal segmentation of the textures

Перевод пока недоступен