An Efficient Approach Technique for Dynamical Infrared/Visible Images Fusion

Abstract

In this paper, we present a novel approach for a real time dynamical infrared/visible images fusion. The advantage of the proposed system compared with previous studies are its huge flexibility and probably lowest cost, since that it only employ both an infrared and a conventional visible cameras. A lot of external applications like security, surveillance..., etc. need this kind of alternative to complete the visible information in some cases like a sudden illumination variation from a bright image to an almost dark situation, or a foggy or smoked environment, where usual computer vision algorithm are not robust anymore. In all those cases, the further image treatments depending on the application are not detailed in this paper, since we just focus this study on the way to make both images fitting in a resulting image and to render the image for a user to easily observe a warm area. We describe in the following paragraphs each step of this treatment, and show the results of this image fusion.