Title: Euclidean distance versus Manhattan distance for skin detection using the SFA database

Authors: Ouarda Soltani; Souad Benabdelkader

Addresses: Electronics Department, Batna 2 University, Batna, Algeria ' Electronics Department, Batna 2 University, Batna, Algeria

Abstract: Skin detection is very challenging because of the differences in illumination, cameras characteristics, the range of skin colours due to different ethnicities and many other variations. New effective and accurate methodologies are developed for skin colour detection to easily identify human's skin colour threw databases which are specifically designed to assist research in the area of face recognition. One of these is the recently built SFA database that showed high accuracy for segmentation of face images. The approach described in this paper exploits skin and non-skin samples provided by SFA for skin segmentation on the basis of the well-known Euclidean and Manhattan distance metrics. Most importantly, the scheme proposed tries to segment facial colour images inside or outside SFA by means of skin samples belonging to SFA. Simulation results in both SFA and UTD colour face databases indicate that detection rates higher than 95% can be achieved with either measure.

Keywords: skin segmentation; skin colour detection; Euclidean distance; Manhattan distance.

DOI: 10.1504/IJBM.2022.119553

International Journal of Biometrics, 2022 Vol.14 No.1, pp.46 - 60

Accepted: 02 Sep 2020
Published online: 09 Dec 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article