Vehicle Detection in Video Based on the Framework of Kernel Density Estimation
Dan Yang, Yantao Chen, Richen Liu
Available Online December 2013.
- https://doi.org/10.2991/wiet-13.2013.37How to use a DOI?
- Background model; Vehicle detection; Kernel density estimation; Intelligent transportation system
- This paper introduced a method of background subtraction model for detecting vehicles in video, the model was based on the framework of the kernel density estimation. To compute the adjacent frame subtraction on the spatio-temporal by the median filter, when the median value is 0, keep the background pixel value of the point stable, and still use the pixel value of the previous frame of the background to speed up the processing, when the value is not 0, using kernel density estimation to compute the probability, and determine whether update the background pixel value or not. Using histogram to show the statistics of the probability distribution, and find out more accurate threshold of the background adaptively. To solve the problem of deadlock, update the background automatically in a certain period of time. The experiments show that this method can segment the moving vehicles rapidly and accurately from the video.
- Open Access
- This is an open access article distributed under the CC BY-NC license.
Cite this article
TY - CONF AU - Dan Yang AU - Yantao Chen AU - Richen Liu PY - 2013/12 DA - 2013/12 TI - Vehicle Detection in Video Based on the Framework of Kernel Density Estimation BT - AASRI Winter International Conference on Engineering and Technology (AASRI-WIET 2013) PB - Atlantis Press SP - 158 EP - 161 SN - 1951-6851 UR - https://doi.org/10.2991/wiet-13.2013.37 DO - https://doi.org/10.2991/wiet-13.2013.37 ID - Yang2013/12 ER -