ABSTRACT
Texture feature is a measure method about relationships among the pixels in a local area, reflecting the changes of image space gray levels. The homogeneous texture descriptor homogeneous texture descriptor describes directionality, coarseness, and regularity of patterns in images and is most suitable for a quantitative representation of a texture having homogeneous properties. Colour is one of the most widely used image features for many computer vision applications. The colour feature is robust to cluttered background, scaling orientation, perspective, and size of an image. Edge Detection An edge detection problem is considered one of the most important and difficult operations in image processing. Edges are used as image features. In many cases, image segmentation processes need edge detection. The gradient magnitude invariant to image rotation, and so it is independent to the orientation of different image structures. Hough transform Hough transform is an algorithm to identify specific shapes in an image.
