Date of Award
Bachelor of Applied Sciences Honours
School of Arts and Applied Sciences
Dr James Cross
Dr Richard Brightwell
Dr Inta Adams
This dissertation is a natural extension of my undergraduate research project entitled, "Digital Image Processing”. Whilst my undergraduate project dealt with a number of classical digital image filtering techniques such as spatial convolution and frequency domain filtering via the Fourier transform, this dissertation focuses on an alternative approach employing Mathematical Morphology. In contrast to classical filtering techniques, which often geometrically distort the original image, morphological operations, used sensibly, essentially preserve shape and geometry. Mathematical morphology therefore lends itself to image processing applications requiring the identification of objects and object features within an image. Herein basic morphological operations are developed, firstly within the continuous image domain (Euclidean N-space, Rn), and then in the digital domain (Zn). Particular emphasis is placed on the development of digital morphological operators for both binary and grey-tone images, progressing from the rudimentary operations of dilation and erosion to granulometry and topological processing. Subsequent to the development of the theory, a case study is presented. The case study describes a preliminary investigation into the application of morphological operations to photomicrographs of mouse adrenal cortex cells with the aim of identifying specific cell features.
Mehnert, A. (1990). Digital Image Processing Using Mathematical Morphology. Retrieved from http://ro.ecu.edu.au/theses_hons/245