ArsenicSkinImageBD

Published: 29 November 2023| Version 2 | DOI: 10.17632/x4hgnjj5gv.2
Contributors:
Nishat Tasnim Niloy, Ismot Ara Emu, Anindya Chowdhury, Bhuyan Md Anowarul Karim, Fatema Tuj Johora, Mahamudul Hasan, Mohammad Rifat Ahmmad Rashid, Tanni Mittra, Taskeed Jabid, Maheen Islam, Sawkat Ali

Description

Type of data: Image file (Dimension: 244p x 244p) Data format: PNG Number of images: The dataset contains 8892 images in "Augmented" folder, and in the "original" folder, each of the healthy and affected folders contain 741 images, in total 1482 distinct images. Diseases considered: Arsenicosis Number of classes: Two (Affected and healthy). Distribution of instances: Both the affected and the healthy category has 4446 images. How data are acquired: Human subjects were photographed using four smartphones with different camera specifications. Data source locations: 1. Village: Betbaria, Ward:08, Union: Balidanga, 2. Village: Balubagan, Ward:08, Union: Maharajpur, 3. Village: Dole para, Ward:09, Union: Maharajapur, 4. Village: Ramchandrapur hat, Ward:08, Union: Dohilpara, Sadar, Chapainawabganj. Where applicable: Suitable in dermatology, to distinguish affected and healthy skin. Folder Structure: Provided in the readme file.

Files

Institutions

  • East West University

Categories

Computer Vision, Dermatology, Image Processing, Image Classification

Licence