Forearm and Hand Vein Detection System for an Infrared Image Database

Abstract

This paper presents a system that allows defining the vein patterns of a person's forearm and hand. In order to accomplish this, infrared (IR) images of the region of interest were registered. The main goal is to help in vein detection, to aid in procedures like intravenous catheter or venipuncture, in a non-invasive way. In the image acquisition protocol, the anterior and posterior compartments of the left and right forearms of each subject were considered to create an image database each one containing the anthropometric data of the subject. A GUI was developed to allow recording an image by gender, age, weight, high and blood group and to allow two filtering options; the standard histogram equalization or fuzzy equalization. Finally, the user is able to choose and apply one of the two classification methods: one using the Fuzzy C-means algorithm, and the other using a Bayesian probabilistic model.