Align and fusion two thermal and visual images

Barbara Charchekhandra

Jadavpur University, Department Of Mathematics, Kolkata, India

Abstract

Observation and detection in different environments and weather conditions requires the use of multiple imaging methods in different spectral fields (millimeter, IR, visual, ultraviolet), where the merging process collects useful information from each spectral field and presents it in a single image. Fusion requires scene matching (zoom, rotation, shift). This article forms a good basis for real-time video image alignment and fusion, with a special focus on pixel-level fusion. In this work, we applied the affine transformation to align two images with an accuracy of up to one pixel, and then we applied several images fusion algorithms based on summing, average, linear summation, or taking the maximum or minimum values of the pixel values, and the results of applying linear addition and values algorithms were The pixels are great compared to other methods.

Keywords: Fusion; visual image; fusion algorithm; pixel-level fusion