Abstract
Real time mapping and exploration without GPS is a challenging problem. For flying robots such as a quadrotor, the choice of sensors is narrowed due to its limited payload and power supply. This paper presents an integrated lightweight system of autonomous mapping and exploration for quadrotor using only a monocular camera as sensor besides IMU. For exploration, we present an original approach named Ambient Environment Based Approach, which is designed for quadrotor in relatively complex environment. The whole system is based on monocular visual SLAM algorithm, camera pose and position obtained from which are used to control the quadrotor, including navigation, obstacle avoidance and autonomous exploration. The new exploration strategy is compared to another state-of-the-art strategy in computer simulation. A small-scaled room demonstration is shown in the paper.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call