Toward Autonomous Miniature Rotorcrafts in Cluttered Environments for Scene Understanding
Miniature rotorcrafts, such as the Quadrotor aerial vehicles, are emerging as a popular platform for unmanned aerial vehicle (UAV) research due to their structural simplicity and small form factor, their ability to carry relatively large payload, and their VTOL (Vertical Take Off and Landing) capability and high maneuverability. They have been used in many military and civilian applications, such as data and image acquisition of targets and affected areas, target localization and tracking, map building, deployment of instrumentation, terrain and utilities inspection, disaster monitoring, environmental surveillance, search and rescue, law enforcement, aerial mapping, traffic surveillance, and cinematography. The objectives of this project are to 1) develop control and navigation algorithms for small scale rotorcrafts to autonomously explore cluttered and obstacle-dense environments via multi-modal sensing; 2) develop innovative approaches for target detection and 3D scene understanding using spatiotemporal image analysis; 3) develop a miniature rotorcraft experimental platform to test and verify the proposed methods.
This project is supported by Army Research Office under Grant No. W911NF0910565
PI: Jizhong Xiao, Co-PI: Yingli Tian
The CityFlyer project was built on a modified Ascending Technologies Pelican Quadrotor Helicopter as a R&D platform to evalaute methods and techniques for 3D mapping and navigation. The CCNY Robotics Lab has developed many software drivers/tools for MAV navigation, visual odometry, 3D SLAM and contributed software packages to ROS.org community (http://www.ros.org/wiki/ccny-ros-pkg), including the AscTec flight control drivers, computer vision and visual odometry stacks, laser scanner tools, Kinect drivers, and ground station interface, which has benefited many research groups worldwide.