Skip Navigation Links
Journal of Applied Nonlinear Dynamics
Miguel A. F. Sanjuan (editor), Albert C.J. Luo (editor)
Miguel A. F. Sanjuan (editor)

Department of Physics, Universidad Rey Juan Carlos, 28933 Mostoles, Madrid, Spain


Albert C.J. Luo (editor)

Department of Mechanical and Industrial Engineering, Southern Illinois University Ed-wardsville, IL 62026-1805, USA

Fax: +1 618 650 2555 Email:

Recognition and Threat Level Estimation of FOD Based on Image Content Features and Experiment Analysis

Journal of Applied Nonlinear Dynamics 4(2) (2015) 197--213 | DOI:10.5890/JAND.2015.06.008

Gang Xiao$^{1}$, Yu Li$^{1}$, Xiao Yun$^{1}$, Jinhua Xie$^{2}$

$^{1}$ School of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China

$^{2}$ AVIC Radar and Avionics Institute, Wuxi Jiangsu 214063, China

Download Full Text PDF



Foreign Object Debris (FOD) is debris or article alien, which may have fallen onto a runway or taxiway, would potentially cause damage to aircraft. FOD surveillance system is used to FOD surveillance system as a significant technology. However, FOD is a kind of relatively random and diverse target that makes it difficult to analyze and recognize. In the reality, the primary goal of FOD recognition cannot recognize FOD accurately, because airport staffs do not care much about what exactly the FOD is but how great the FOD could threat to airplane. Therefore, based on FOD image features definition and threat level model, we proposed a simple and efficient method to estimate the FOD threat level. By comparison with BP neural network model, the error of new FOD threats coefficient model are no more than 15%. There are five sections discussed in the paper. First, the FOD surveillance system was introduced briefly. Second, the features of FOD including color, texture and shape and related extraction algorithm had been discussed. The HSV color model, texture model and normal shape feature like area, perimeter and eccentricity which are all used to describe FOD targets. Fourteen typical FOD targets such as stone fragment, metal chip, piece of rubber, pieces of paper, plastic and small plant are selected as experiment objects. Thirdly, a new FOD threat coefficient model was proposed, in terms of the values of features based on the huge amount of experiment results and human subjective judgment. Next, the threat level classifier was constructed by the BP neural network. The classifier was trained by some experiment FOD targets with given threat level. The others FOD targets were tested. Finally, the recognition and estimation results indicated that the proposed threat level model were effective and efficient, satisfied the robustness and real-time requirement.


The authors would like to thank the anonymous referees for their valuable comments and suggestions, which helps to improve the quality of this paper greatly. This paper was supported by the National Basic Research Program of China (2014CB744903), National Natural Science Foundation of China (60904096), China Aviation Science Fund (2012ZC15005), China Aviation Support Program (61901060202), the China AVIC industrial projects (CXY2012SJ37) and China Scholarship Council (201208310570).


  1. [1]  Jim Patterson, Jr (2008), Foreign Object Debris (FOD) detection research, International Airport Review, 23(2), 22-27.
  2. [2]  FAA. AC 150/5220-24(2009), Airport Foreign Object Debris Detection Equipment, 1-13.
  3. [3]  Steffi Baumgarten (2014), Case studies international-Incidents at international airports, ZRH Safety Newsletter FOD hazards, 2014(4), 10.
  4. [4]  Beasley, Patrick. Tarsier (2007), Unique radar for helping to keep Debris off Airport Runways, the Future of Civil Radar 2006, The Institution of Engineering and Technology Seminar, London, 12-22.
  5. [5]  Feil P. (2008), Foreign Objects Debris Detection (FOD) on Airport Runways Using a Broadband 78 GHz Sensor, Proceedings of the 38th European Microwave Conference, 1608-1611.
  6. [6]  Ding, Lufei et al (2009), Radar Principles (fourth edition), Electronic Industry Press, Beijing.
  7. [7]  Ferri, M., Giunta, G., Banelli, A., and Neri, D.(2009), MillimetreWave Radar Applications to Airport Surface Movement Control and Foreign Object Detection, Proceedings of the 6th European Radar Conference. Roma, 1-8.
  8. [8]  Rafael, C., Gonzalez, Richard E. Woods (2006), Digital Image Processing (Second Edition), Publishing House of Electronic Industry, 290-344.
  9. [9]  Smith, J. and Chang, S. (1995), Tools and techniques for colour image retrieval, SPIE Proceedings: Storage & Retrieval for Image and Video Databases IV, 2670, 426-437.
  10. [10]  Charl Coetzee, et al. (1998), PC Based Number Plate Recognition Systems, In Proc. IEEE International Symposium on Industrial Electronics, 605-610.
  11. [11]  Lippmann, R. (1989), Pattern Classification Using Neural Networks, IEEE Communication on Mag, 7, 47-64.
  12. [12]  Veltkamp, R. and Tanase, M. (1999), State-of-the-art in shape matching, Technical Report UU-CS-1999-27, Utrecht University, the Netherlands, 105-110.
  13. [13]  Huang, T.S. and Yong, R. (1997), Image retrieval: past present and future, International Symposium on Multimedia Information Processing, Taiwan, 45-55.
  14. [14]  Gunsel, B. and Tekalp, A. (1998),Shape similarity matching for query by example, Pattern Recognition, 3l (7), 931-944.
  15. [15]  Persoon, E. and Fu, K.S. (1986), Shape discrimination using Fourier descriptors, IEEE Transactions on Pattern Analysis and Machine Intelligence, 8, 388-397.
  16. [16]  Yu Li, Gang Xiao (2011), Airport runway foreign object detection system design and research, Laser and infrared, 41(8), 909-915.
  17. [17]  Jorge J. MorĂ© (1978), The Levenberg-Marquardt algorithm: Implementation and theory, Lecture Notes in Mathematics, 630, 105-116.