Similarity quantization: Compressed sensing application

In this paper, a similarity vector, which can be used to estimate the level of the background clutter, between the target image and the background image is obtained by taking advantage of the theories of the compressed sensing. And the similarity vector, in the future, can be applied to develop a clutter metric based on target features, so that we can predict the target acquisition performance without a great deal of priori knowledge about the target image to be detected. Moreover, the experiment results demonstrate its feasibility on the search_2 database.