Mathematical Morphology for Edge and Overlap Detection for Medical Images

Abstract Mathematical morphology operations have become increasingly popular in the past two decades in the field of pattern recognition because of their flexibility and ability to be implemented in hardware. In this paper, we propose a generalized approach using these operations for shape recognition of medical images, specifically human lungs generated by a gamma ray camera. We present a new approach towards overlap detection in such images. Moreover, we propose the use of morphology for edge detection and present results of some other algorithms like Deriche's algorithm and the Laplacian of Gaussian algorithm. Mathematical morphology is very flexible and can be implemented in hardware to achieve a real-time result. In the present system, only the data acquisition is in real time, however, the subsequent analysis is done off-line because mathematical morphology is inherently computationally intensive since it deals with sets. Therefore our objective is to achieve real-time results by implementing this system in hardware.