Path planning problems are known as one of the most important techniques used in robot navigation. This paper adopts an Improved Artificial Bee Colony (IABC) algorithm and combines with cubic Bezier curve to describe the path, which implements the path optimization. The standard artificial bee colony algorithm has shortcomings of falling into local optima and the convergence speed is slow in the later. To overcome these disadvantages, the proposed algorithm modifies the search methods of employed bees and onlooker bees. Compared with other algorithms, we gain the advantages and disadvantages of the different algorithms in path optimization. The experimental results demonstrate that the IABC algorithm has better search performance in path optimization and is able to get a shorter path.