Algorithms for autonomous visual flight control

Summary form only given. The authors attempt to utilize optic flow and the geometrical properties of a sensor configuration to perform fast, reliable target tracking from crude visual inputs. Artificial neural networks are used to implement the authors´ algorithms because of their excellent generalization capabilities which allow for quite a good answer in situations where an answer would not be expected of conventional computer algorithms. The advantages offered are speed of computation, simplicity of schemas and algorithms, and robustness as regards the environmental conditions where such tracking is performed.<>