An Improved Difference Method Based on Three Asymmetric Frames and its Applications in the Vehicle Detection

Abstract:

Original difference method based on three sequential frames is a common method in the moving objective detection. After the introduction of the original method, the formation cause of the internal cavity in vehicle region was analyzed. Then, an improved difference method based on three asymmetric frames, which were abstracted according to the judgment rule of optimal frame interval, has been proposed in this paper. The area of the moving vehicle region is assumed as the criterion to calculate the best frame interval. In the improved method, the best benchmark frame and the best auxiliary frame were firstly abstracted out and used to subtract the current frame separately. Subsequently, the two difference images were transformed to binary images separately. Lastly, the final moving vehicles were detected in the current frame after logical "and" operation was carried out between the two binary images. The experimental results showed that the improve difference method based on three asymmetric frames can effectively reduce the internal cavities in the vehicle region, comparing with the original one.

• Periodical: Advanced Materials Research (Volumes 468-471)
• Pages: 505-509
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.468-471.505
• Online since: February 2012
• Authors: Zhan Wei Liu, Wei Hua Zhao, Wei Wang, Zheng Xu Zhao
• Keywords: Improved Difference Method Based on Three Asymmetric Frames, Judgment Rule of Optimal Frame Interval, Moving Vehicles Detection, Original Difference Method Based on Three Continuous Frames

References

[1] Kouthemy P, Fraancois E. "Motion segmentation and dynamic scene analysis from an image sequence". International Journal of Computer Vision, 1993, 10(2): pp.157-182.

DOI: 10.1007/bf01420735

[2] Memin E, Dence P. "Estimation and object-based segmentation of the optical flow with robust techniques". IEEE Transactions on Image Procession, 1998, 7(5): pp.703-719.

DOI: 10.1109/83.668027

[3] Lipton A, Fujiyoshi H, Patil R. "Moving target classification and tracking from real-time video". IEEE Workshop on Applications of Computer Vision.Princeton NJ, 1998: pp.8-14.

DOI: 10.1109/acv.1998.732851

[4] Foresti G, Murino V, Regazzoni C. "Vehicle recognition and tracking from road image sequences". IEEE Transaction on Vehicular Technology, 1999, 48(1): pp.301-317

DOI: 10.1109/25.740109

[5] Xuan Nie etc. "A Method compound finite difference for a target detection compound finite difference method" , Journal of Northwestern Polytechnic University, 2009.2, 27(1):8