Rule Based Fuzzy (RBF) membership edge detection on hand phantom radiograph images

Abstract

The paper proposes to prove the application of an adapted Rule Based Fuzzy (RBF) image detection in detecting periosteal and endosteal edges of hand phantom radiograph. The experimental data consists of four sets of hand phantom radiographs. Only the metacarpal 2, 3 and 4 are considered. The Rule Based Fuzzy is slightly modified due to the characteristic of the hand phantom radiograph. Mean and median filters are used as preprocessing tools. The analysis of the edges is done subjectively and statistically. As a whole, the proposed algorithm managed to identify the relevant edges of the periosteal and endosteal.