A New Dynamical Evolutionary Algorithm for Active Contour Optimization

This paper proposes a new dynamical evolutionary algorithm for active contour optimization, called dynamical evolutionary active contour (DEAC). Compared with evolutionary active contour (EAC), DEAC is to improve selecting strategies and to activate all individuals at the most possible with their population evolving, and preserves the diversity of individuals in a population. The experimental results with the synthetic and simple real images demonstrate the validation and robustness. Moreover, the proposed algorithm has been applied to X-ray color image. The results of the comparison with EAC show that the proposed algorithm has a more powerful global exploration capability, and can converge at the less contour energy function value with less time, as well as more accurate boundary.