Swarm intelligence based image fusion for thermal and visible images

Image fusion is the process of merging multiple images in order to accomplish various multimedia analysis. Recently, Multi-resolution approach (MRA) based image fusion approaches gained the attention of researchers. The reason is because of the advancement of sensor based technology and high storage capacity. The existing MRA based image fusion techniques do not preserve the edges along the curve. Secondly, coeffiecients are fused either using maximum selection, minimum selection or averaging which results in information loss since it takes only some frequency coefficients for fusion and not all of them. Curvelet transform handles both thermal image and visual image by preserving the edge information which is not preserved in other state of-art methods and particle swarm optimization is used to find the optimal weights with which optimized fusion results are obtained. The proposed method is compared with gradient method, laplacian pyramid, ratio of laplacian and shift invariant discrete wavelet transform. The proposed algorithm is proved to be than better than other image fusion techniques. Several characteristic issues that influence a image fusion process such as surveying and mapping, reduced accuracy, noise, blurring and identifying complex situations are addressed and highlighted using the proposed approach by combining curvelet tranform and PSO.