Home // International Journal On Advances in Software, volume 10, numbers 3 and 4, 2017 // View article

Algorithms for Face Detection on Infrared Thermal Images

Authors:
Ricardo Ribeiro
Antonio J. R. Neves

Keywords: Face detection; Infrared images; Image processing; Robotics; Object detection.

Abstract:
Face detection on digital images is an area with immense practical potential which includes a wide range of commercial and research applications, and it continues to be one of the most active research areas on computer vision. Even after several decades of intense research, the state-of-the-art in this topic continues to improve, benefiting from advances in a wide range of different fields. In an attempt to overcome some of the limitations in face detection using visible light, the use of thermal imaging has emerged as a particularly promising research direction. This is possible because nowadays the infrared sensors have reached a new technological level. In this work we propose several methods for face detection on infrared thermal images.The well known algorithm developed by Paul Viola and Michael Jones, using Haar feature-based cascade classifiers, is used to compare the traditional algorithms developed for visible light images when applied to thermal imaging. Moreover, we present three algorithms for face detection, using image segmentation a pre-processing step to obtain a binary image. Our first approach is based on the use of an edge detection algorithm applied to the binary image. The face detection is based on the analysis of the obtained contours. As a second approach, the use of a template matching method searches and try to find the best location of a template image with the shape of human head in the binary image.In a third approach, a matching algorithm is used. This algorithm correlates a template with the distance transform of the edge image. This algorithm incorporates edge orientation information resulting in the reduction of false detection and the cost variation is limited.We performed tests taking into consideration different environments. Two of them were events at the University of Aveiro with the objective to simulate real environments, were several images were recorded while people were looking to posters. Some other laboratory tests were performed with image captured processed in real time. The results show that the proposed methods have promising outcome, but the second method is the most suitable for the performed experiments.

Pages: 499 to 512

Copyright: Copyright (c) to authors, 2017. Used with permission.

Publication date: December 31, 2017

Published in: journal

ISSN: 1942-2628