Vehicle Path Planning Based on Pedestrian Collision Avoidance Action

Although significant progress has been made in developing autonomous vehicle technology, further research is still needed to reduce the risk of collision by planning a suitable path. Drawing paths through pedestrian avoidance action worth study. In this paper, path considering proposed improve driving safety further. First, movement features reflecting different types intention are summarized actual human-vehicle accident video. Then, data representing obtained deep neural network and input into trained Hidden Markov Model predict intention. Finally, planned cubic polynomial combined with The algorithm verified simulation scene. results show that can pedestrians avoid collisions at frames 9, 8, 10, respectively, which 0.6 seconds, 0.4 seconds faster than Long Short Term Memory algorithm. planned, was successfully avoided.