Detection of obstacles on runways using ego-motion compensation and tracking of significant features

摘要

This paper describes a method for obstacle detection on a runway for autonomous navigation and landing of an aircraft, in presence of extraneous features, such as tire-marks. An obstacle is defined as an object, which has a significant motion or height relative to the runway. Suitable features are extracted from the image and warping is performed, using approximately known camera and plane parameters, to compensate the ego-motion as far as possible. Residual disparity after warping is estimated using an optical flow algorithm. Features are tracked from frame to frame to obtain more reliable estimates of their motion. The residual disparities are used to correct the motion parameters with a robust method, where features having large residual disparities are signaled as obstacles. Nelson's optical flow constraint is proposed to separate moving obstacles from stationary ones. A Bayesian framework is used at every stage so that the confidence in the estimates can be determined.