作者: Mandyam Srinivasan , Saul Thurrowgood , Dean Soccol
关键词: Computation 、 Artificial intelligence 、 Collision avoidance 、 Mobile robot 、 Perception 、 Robotics 、 Computer vision 、 Obstacle avoidance 、 Computer science 、 Terrain 、 Trajectory
摘要: In this article, we describe how flying insects use vision for guidance, especially in the contexts of regulating flight speed, negotiating narrow gaps, avoiding obstacles, and performing smooth landings. We show that many these maneuvers, which were traditionally believed to involve relatively complex high-level perception, can be achieved through low-level cues simple computation. also tests effectiveness some strategies autonomous guidance small-scale terrestrial aerial vehicles corridor navigation, altitude control, terrain following landing. a novel, mirror- based imaging system is tailored tasks facilitates requisite visual computations.