Automatic detection of obstacles on a runway using computer vision
1. Moscow Aviation Institute (National Research University), 4, Volokolamskoe shosse, Moscow, А-80, GSP-3, 125993, Russia
Onboard systems for automatic detection of obstacles on a runway are highly demanded nowadays. Such systems reduce the potential hazard for the safety of air travel during the landing and the takeoff stages of the flight. Ground based obstacles detection systems are being tested in the biggest airports. Such systems usually use LIDAR or optic sensors.There is a high demand for a mobile obstacle detection system that could be installed on an aircraft. This paper presents an algorithm for detection of obstacles on a runway based on an iterative minimization of a difference of scale photographic maps of a runway. Scale photographic maps are generated from images captured by an onboard camera.
The paper presents the structure of the proposed algorithm and detailed explanation of its major steps: the generation of a photographic scale map, the creation of a scale map difference image and the estimation of geometrical parameters of an obstacle. The paper also concerns a problem of an automatic estimation of the class of the detected obstacle. It is proposed to create a database of 3D-objects of potential obstacles. The class of the discovered obstacle could be estimated by comparison of its dimensions with dimensions of objects in the database.
The proposed algorithm was tested using image sequences of a runway created using a 3D-model on an airport and a scale model of a runway. The algorithm demonstrated reasonable precision and robustness during experiments.The next stage ofresearch is supposed to be the development of more accurate foreign objects classification algorithm and modification for real-time application.
Keywords:aircrafts, foreign object detection system, safety of the flight, digital image processing, photogrammetry
U.S. Federal Aviation Administration. Advisory circular on Airport Foreign Object Debris (FOD) Detection Equipment, 2009.
Weller J.R. FOD Detection System. Wildlife and Foreign Object Debris (FOD) Workshop, Cairo, Egypt, 2014.
Zheltov S.Yu., Sibiryakov A.V. Method of 3d-object detection based on orthophoto difference analysis. The International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3, Amsterdam, 2000.
Vizil’ter Yu.V., ZheltovS.Yu., Bonadarenko A.V., Ososkov M.V., Morzhin A.V. Obrabotka i analiz izobrazhenii v zadachakh mashinnogo zreniya (Processing and analysis of images in computer vision task), Moscow, Fizmatkniga, 2010, 672 p.
Gonzalez R.C., Woods R.E. Tsifrovaya obrabotka izobrazhenii (DigitalImage Processing), Moscow, Tekhnosfera, 2012, 1104 p.
Press W.H., Teukolsky S.A., Vetterling W.T. Flannery B.P. Numerical Recipes: The Art of Scientific Computing, Third Edition, New York: Cambridge University Press, 2007, 1235 p.
Barnard S.T. Interpreting perspective images, Artificial Intelligence 21, 1983, pp. 435-462.
Lobanov A.N. Fotogrammetriya (Photogrammetry), Moscow, Nedra, 1984, 552 p.