ABSTRACT
Edge detection plays a major role for region of interest (ROI) extraction in image processing and pattern recognition. This work proposes neutrosophic-based edge detection for COVID-19 computed tomography (CT) images. The classical median filter, hybrid median filter and decision-based median filter are used prior to ROI extraction. The hybrid median filter when coupled with the neutrosophic edge detector yields robust results, when equaled with the other edge detection approaches. Proposed edge detector performance was compared with classical edge detectors and validated by performance metrics. The algorithms are implemented in Matlab 2015a and validated on real-time COVID-19 CT medical images.
