Image Compared by Election Campaign Algorithm

摘要

Usually CBIR (Content-based image retrieval) is an image retrieval method that exploits the feature of the image as the retrieval index, which is based upon the content, including colors, textures, shapes and distributions of objects in the image. After the feature detecting, the composition of the similarity matching image set is found, then detecting the most matching image still need to be process in the higher level analysis and retrieval. It is a difficult and slow process. So, if we take an opposite approach, detecting the not-match image from the similarity matching image set but comparing all the images in the set, it can be more easily to achieve. In this paper, we propose an new image comparison method base on Election Campaign Algorithm, which provide parallel and fast optimum feature detecting, to detect the not-match images from the similarity matching image set, then another method would be use to find the most-match images. With this method, the image comparison process is fast, the size-reduce image set is quickly to be received.