Abstract
Abstract. For autonomous navigation of micro aerial vehicles (MAVs), a robust detection of obstacles with onboard sensors is necessary in order to avoid collisions. Cameras have the potential to perceive the surroundings of MAVs for the reconstruction of their 3D structure. We equipped our MAV with two fisheye stereo camera pairs to achieve an omnidirectional field-of-view. Most stereo algorithms are designed for the standard pinhole camera model, though. Hence, the distortion effects of the fisheye lenses must be properly modeled and model parameters must be identified by suitable calibration procedures. In this work, we evaluate the use of real-time stereo algorithms for depth reconstruction from fisheye cameras together with different methods for calibration. In our experiments, we focus on obstacles occurring in urban environments that are hard to detect due to their low diameter or homogeneous texture.
Highlights
In recent years, micro aerial vehicles (MAVs) such as multicopters have become increasingly popular as a research tool and for applications like inspection tasks
We evaluated the suitability of four different state-of-the-art stereo algorithms for reliable obstacle detection on images from stereo cameras with fisheye lenses
In order to deal with the higher distortion of the fisheye lenses, we used different calibration methods to model the lens distortion and to rectify the stereo images
Summary
Micro aerial vehicles (MAVs) such as multicopters have become increasingly popular as a research tool and for applications like inspection tasks Due to their low costs, small size, and the ability to hover, it is possible to reach locations, which are inaccessible or dangerous for humans or ground vehicles. Most MAVs are remotely controlled by a human operator and when constructing autonomous MAVs, payload limitations are one of the main challenges Due to their small size and low weight, cameras are potential sensors for several tasks: from visual odometry over simultaneous localization and mapping (SLAM) and 3D surface reconstruction to visual obstacle detection. The cameras provide dense measurements with high frequency in comparison to the other sensors: They capture images with 20 Hz while the laser scanner operates with 2 Hz. We use the MAV for autonomous navigation in the vicinity of obstacles (Droeschel et al, 2015). The major challenge hereby is the modeling of the fisheye lenses, which capture highly radial distorted images that need to be rectified in real-time on the onboard computer
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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
