Biologically Inspired Strategies, Algorithms and Hardware for Visual Guidance of Autonomous Helicopters

摘要

This project has designed, developed and field-tested biologically inspired strategies, algorithms and hardware for vision-based guidance of aircraft for the performance of a variety of tasks, listed below. (i) Control of aircraft altitude and attitude, and terrain following are achieved by using a novel principle of 'virtual optic flow' to monitor and regulate the height and the orientation of the aircraft relative to the ground. (ii) Obstacle detection and avoidance are achieved by using virtual optic flow to sense and steer away from objects that are dangerously close to the flight path. (iii) Algorithms have been developed and tested for stabilizing and controlling aircraft attitude by monitoring the horizon profile. (iv) Algorithms, based on analysis of optic flow, have been developed for the detection of moving targets by a moving vision system.