Research On The Fusion Technologies Of Multi-spectral Electro-optical Images

Image fusion is a modern high technology that integrates many sciences, such as sensor technology, image processing, signal processing, computer engineering and artificial intelligent. It is of great application importance to either military or civil fields.In this paper, we discussed the goal and requirements of image fusion, analyzed some of the commonly used algorithms and developed several multiresolution image fusion algorithms. For some human-aided applications, a novel infrared feature and region similarity based image fusion scheme is proposed, which takes full consideration of the requirements on real-time and the fused image, as well as the characteristics of the infrared and visible images which are used in the fusion. The algorithm proposed in this paper is fast, and the produced image is of high contrast and preserves the details in the visible image, as well as the target information in the infrared. Hence it is much easier for the observer to interpret the scene. Finally, the fusion performance evaluation algorithms are studied and a novel similarity based objective evaluation metric, which is consistence with human subjective evaluation, for gray image fusion is proposed.