Data for: Automatic hair removal on ultraviolet-induced fluorescence dermatoscopy images
Description
Data for: Deep learning based hair removal on ultraviolet-induced fluorescence dermatoscopy images This repository contains cropped sub-images from 120 UFVD photographs from 33 patients. The image artifact visible on UVFD include dark hair, black ruler mark, thick light hair, fine light hair, and UV-reflecting particles. Manual segmentation were performed to obtain masks for: a) dark hair and ruler mark, b) light hair and UV particles. After manual segmentation, images were divided into 512x512 pixel squares. Only the cropped sub-images with artifacts present were chosen for latter analysis. For dark hair problem, 2097 cropped images were obtained, while for light hair and UV particles - 805. All masks are saved as PNG images. Data is divided into following folders: images - contains all cropped images, black_masks - contains masks for dark hair and ruler, white_masks - contains masks for light hair and UV particles.
Files
Institutions
- Politechnika SlaskaGliwice