It is necessary to find the largest inner circle of a complex polygon in many applications. The present paper develops a method for finding the largest inner circle of a polygon based on Voronoi diagram, and then improves the algorithm by medial axis simplification (MAS) and parallel computing. Data partition is a key issue in parallel computing of vector data. The algorithm complexity equalization strategy (ACES) is then presented. By several experimental tests of large quantity of lakes in Alaska we conclude that the approach developed in this paper performs effectively and efficiently by using MAS and ACES methods.