Texture describes the surface characteristics object such as plants, skins, terrains etc. This paper is based on the comparative study on various approaches to identify the gray scale texture of an unknown object from a set of texture of a known object available in the database. We have focused on three approaches of comparing the input gray scale texture with the gray scale texture images of various object in a database - image subtraction, gradient based and principal component analysis approaches. The work is carried out on set of the images and the various approaches were then compared and analyzed to find the best method for various textures of different object types.