A Novel Deep Learning Method for Thermal to Annotated Thermal-Optical Fused Images

Abstract

Thermal Images profile the passive radiation of objects and capture them in grayscale images. Such images have a very different distribution of data compared to optical colored images. We present here a work that produces a grayscale thermo-optical fused mask given a thermal input. This is a deep learning based pioneering work since to the best of our knowledge, there exists no other work which produces a mask from a single thermal infrared input image. Our method is also unique in the sense that the deep learning method we are proposing here employ the Discrete Wavelet Transform (DWT) domain instead of the gray level domain. As a part of this work, we also prepared a new and unique database for obtaining the region of interest in thermal images, which have been manually annotated to denote the Region of Interest on 5 different classes of real world images. Finally, we are proposing a simple low cost overhead statistical measure for identifying the region of interest in fused images, which we call as the Region of Fusion (RoF). Experiments on 2 different databases show encouraging results in identifying the region of interest in the fused images. We also show that these images can be processed better in the mixed form rather than with only thermal images.