Evolutionary algorithms are metaheuristic algorithms that mimic biological mechanisms for the solution of optimization problems. They are widely used in many problems of engineering, including image analysis and pattern recognition. Such methods had shown added value in overcoming certain problems and limitiations of more traditional methods and approaches. This is particularlay interesting in Computer for designing faster and more accurate optimization methods. Despite recent advances in image registration, there is still room for improvement especially with the use of metaheuristic algorithms. In this paper, we present a novel method for solving the automated rigid registration problem by introducing a variant of the relatively new metaheuristic method called harmony search which shows promising results in difficult rigid registration cases.