Selection of Feature Regions Set for Digital Image Using Optimization Algorithm

Abstract

Image transformation techniques are used for extraction of local features which are used in many applications such as object recognition, database retrieval, and motion tracking. The effectiveness of a digital image is identified by the robustness of image candidate regions. This paper aims to obtain most robust and minimal region set which preserves image quality. It first performs a simulated attacking procedure on non overlapped region set to evaluate the robustness of every candidate feature region. According to the evaluation results, it then adopts a track-with-pruning procedure to search a minimal primary feature set which can resist the most predefined attacks. In order to enhance its resistance to undefined attacks under the constraint of preserving image quality, the primary feature set is then extended by adding into some auxiliary feature regions. This work is formulated as a multidimensional knapsack problem and would be solved by Optimization Algorithms such as Genetic Algorithm as well as Particle Swarm Optimization and Simulated Annealing based approach. The experimental result uses some benchmark images. Comparing results of different optimization techniques on primary feature region set, the proposed method determines the best choice of optimization techniques for selecting most robust and minimal region set under the constraint of preserving image quality.