A Self-adaptive and Real-time panoramic video mosaicing system

Abstract

In real-time monitoring process, resolution must be set against field vision. There are two problems in the real-time monitoring procedures, which are: the field vision is too small to capture the target completely and a larger field vision with low resolution. In order to solve these problems, a new type of real-time panoramic monitoring systems becomes necessary. In this paper, we proposed a new adaptive video mosaicing algorithm based on SIFT algorithm and template matching method. By real-time adjusting and updating the mosaicing position of adjacent two images to establish the optimal mosaic line, we can both made full use of the accuracy of feature matching and the simplicity of template matching. Experimental results show our algorithm not only has good matching rate but also reduces the time complexity, which make it very suitable for the application which requests high real-time performance.