Intelligent Assessment of Image Matching Method Based on the Cumulative Prospect Theory

Recently years, various image matching methods spring up so that it becomes more and more difficult for decision makers to choose. In this paper, an intelligent and quantitative assessment method of image matching method based on the cumulative prospect theory is proposed. In the method, the theory is induced to deal with the multi outcomes of each measure that assesses the matching results, and there are two novelties. First, before the outcomes of each measure are taken into value function, they are normalized depending on the accepted threshold as well as the desired direction of the decision maker. Second, in the weighting function the probability distributing function of the outcomes of each measure is calculated using the histogram without any assumption of the functional form. Experiment using real matching results obtained by three different methods demonstrated that the proposed method can effectively assess the image matching methods.