A feature based �Selection of feature region set for digital image using Optimization Algorithm� is proposed here. The work is based on simulated attacking and optimization solving procedure. Image transformation techniques are used to extract local features. Simulated attacking procedure is performed to evaluate the robustness of every candidate feature region. According to the evaluation results, a track-with-pruning procedure I adopted to search a minimal primary feature set which may resists the most predefined attacks. In order to enhance its resistance capability against undefined attacks, primary feature set is then extended by adding some auxiliary feature regions in it. This work is formulated as a multidimensional knapsack problem and solved by optimization algorithms such as Genetic Algorithm, Particle Swarm Optimization and Simulated Annealing.