Publication Date: 2022/09/28
Abstract: Melanoma is the most dangerous form of skin cancer, develops when unrepaired damage happens to skin cells known as melanocytes. The skin cells rapidly multiply and forms tumors, called as malignant tumors. The main goal of our paper is hair removal from clinical dermoscopic images. It is the initial preprocessing stage for the skin melanoma diagnostics. In dermoscopy image analysis while analyzing the disease the main obstructs will be the hair objects. These hairs will hide the region of interest (ROI) from being detected. There are many computer-aided techniques to assist dermatologists. In this paper, we propose an algorithm which removes hairs without altering the lesion area. The algorithm uses image processing technique for detection and removal of hair from the dermoscopy images. First, hair is excluded from lesion; later image is reconstructed by considering the nearest 8 edge pixels. The algorithm is tested on PH2 database. We have used MATLAB for implementing our algorithm.
Keywords: Melanoma, Skin Cancer, Dermoscopy, lesions;
DOI: https://doi.org/10.5281/zenodo.7117915
PDF: https://ijirst.demo4.arinfotech.co/assets/upload/files/IJISRT22SEP480.pdf
REFERENCES