Moving Vehicle Detection of Airborne Video on SURF

Abstract:

Due to the high mobility, a wide range of monitoring, air mobile platform-based vehicle detection and tracking system is becoming core of the investigation and the monitoring. Self-motion of the camera and external interference caused by the low-level platform led to instability of the obtained video and affect the correct detection of moving targets and subsequent analysis. For the characteristics for low-level video, an image stabilization algorithm based on SURF combined with normal vector of optical flow is proposed to solve moving vehicle detection low-altitude video. From the experimental results can be seen: (1) compared to other moving vehicle detection methods, the method proposed can get better detection efficiency and detection accuracy; (2) in the complex context, this method can effectively detect moving vehicles. The experiments show that this method has some theoretical and application value of space-based video moving target detection.

• Periodical: Advanced Materials Research (Volumes 734-737)
• Pages: 2815-2818
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.734-737.2815
• Online since: August 2013
• Authors: Hui Liu*, Chun Xian Gao, Xing Hao Ding, Zhe Zeng
• Keywords: SURF Image Stabilization Moving Vehicle Detection

References

[1] Stauffer, C., Grimson, W, IEEE conference on computer vision and pattern recognition. Fort Collins, CO (1999), p.246–252

[2] Barron, J., Fleet, D., Beauchemin, S, Int. J. Comput. Vis. 12(1), (1994)42-77

[3] Isaac Cohen, G´erard Medioni, Proc. of the IEEE Computer Vision and Pattern Recognition 99, Fort Collins, June 1999. 2

[4] Anand C.Shastry and Robert A.Schowengerdt, IEEE Transaction on Intelligent Transportation Systems, Vol.6. NO.4. December (2005)

[5] Bay, Herbert, Ess, Andreas, Tuytelaars, Tinne, Van Gool, Luc, Computer Vision and Image Understanding, v 110, n 3, June 2008, pp.346-359

DOI: 10.1016/j.cviu.2007.09.014

[6] Lowe D.G, International Journal of Computer Vision , 2004 , 60 (2) : 912110