Aircraft's visual navigation is a navigation technology of extracting navigation information from a train of images, which gives motion information of the aircraft concerned. To date, visual method provides an alternative of inertial and satellite navigation techniques. Aircraft's visual navigation without manmade landmarks is studied in this paper. Since there is no artificial landmark available, resorting to environmental characteristics of flight is necessary to obtain structural information. Using a multi-scale approach to extract such structural information is a highly preferable choice. Since its hierarchical nature makes it efficient to compute and accurate enough to satisfy practical use. The feature estimation function and the feature innovation are proposed in this paper. Using a hierarchical algorithm to extract navigation information is realized in the end
[1]
Peter G. Ifju,et al.
Sky/ground modeling for autonomous MAV flight
,
2003,
2003 IEEE International Conference on Robotics and Automation (Cat. No.03CH37422).
[2]
Zhaoying Zhou,et al.
Micro air vehicle: configuration, analysis, fabrication, and test
,
2004
.
[3]
Dong Sun,et al.
Micro air vehicle: configuration, analysis, fabrication, and test
,
2004,
IEEE/ASME Transactions on Mechatronics.
[4]
E.N. Johnson,et al.
Vision-only aircraft flight control
,
2003,
Digital Avionics Systems Conference, 2003. DASC '03. The 22nd.
[5]
Eric N. Johnson,et al.
Vision-Only Aircraft Flight Control Methods and Test Results
,
2004
.