A Novel Informative Autonomous Exploration Strategy With Uniform Sampling for Quadrotors

In this article, a novel informative autonomous exploration strategy with uniform sampling is proposed to efficiently reduce the unknown volume of 3-D environments and provide an accurate truncated-signed-distance-function-based reconstruction for quadrotors. Different from existing methods, the proposed method locally samples candidate viewpoints uniformly and achieves a global coverage efficiently by backtracking history nodes when there is no unmapped space near the quadrotor. Specifically, by sampling viewpoints with a fixed radius and assigning priority, the best one is selected to which a collision-free path will be planned. Then, a history list is designed to store the executed nodes, which is backtracked to find the unmapped space, resulting in a complete exploration of the environment. In addition, a hierarchical information gain is proposed to balance the space exploration efficiency and reconstruction accuracy by considering the uncertainty of the surface and volume of the unknown space. Comparative experiments in simulation show that the proposed method outperforms the sampling-based state-of-the-art method in terms of both the exploration time and reconstruction accuracy. Moreover, the effectiveness of the proposed approach is further evaluated in a real-world experiment.