At present, the visibility graphs algorithm is mainly implemented in the field of mobile robot’s path planning on unstructured pavement. It only considers the common constraints such as travel time and move distance. The problems of road boundary and vehicle dynamics constraints are not involved. In this paper, a local path planning algorithm based on improved visibility graphs was proposed for intelligent vehicles traveling on structured road. The algorithm combined the geometric constraints of the road boundary with the vehicle dynamics constraints. By predicting the position of the obstacle vehicle in the preview time, the algorithm was used to generate the information of path network. Then, the performance index function was used to evaluate the trajectory which satisfying the maximum curvature constraint in the network. In this way a smooth path with minimum distances satisfying the constraints was obtained. Finally, the validity and efficiency of the algorithm are verified in simulation experiments. The experimental results show that, the planned path achieved in this paper has shorter travel distance and lower curvature, which is more appropriate for the trajectory requirements of intelligent vehicle compared with the results of RRT.