Study of morphological operations in image processing.

Mathematical morphology provides an approach, for processing of digital , which is based on shape. Shape can be conveniently studied by the boundary and edges of given image. Also, convexity plays an important role in the study of shapes. The issues of boundary and convexity pose many problems in discrete plane. The definition of boundary of a binary image within the framework of mathematical morphology, and its relation to convexity are presented in this thesis. Also, two algorithms are developed for construction of convex hull. A hardware solution for convex hull is proposed. Since morphological operators are inherently nonlinear, the input image set cannot be completely characterized by the transformed image set. That is, there can be more than one input image which may result into the same transformed image. This problem is solved by defining and characterizing a class of called equivalence class. Procedure for constructing equivalence class for the given operator is developed. More Info

Tags :
Your rating: None Average: 5 (1 vote)