A Visual Guided Quadrotor for IMAV 2012 Indoor Autonomy Competition and Visual Control of a Quadrotor for the IMAV 2012 Indoor Dynamics Competition

摘要

This article presents the proposal of the Computer Vision Group to the indoors competitions of the IMAV 2012. The team's hardware consists of a Pelican quadrotor with a down-facing camera, an altitude sensor and a Hokuyo laser range finder. The down-facing camera is used to estimate the horizontal speed of the quadrotor using optical flow. The measurements of the Hokuyo sensor are filtered using a Particle Filter or Monte-Carlo localization algorithm to obtain a position measurement to correct the odometry cumulative error. The utilized system architecture allows the adoption of several configuration schemes, allowing to perform the data processing in different computers.