The performance of existing vehicle detection algorithms are subject to many influences of environment, such as different lighting and weather conditions, moving vehicle shadows, etc. To solve these problems, a novel vehicle detection algorithm was proposed. Different from traditional methods, which use motion features to detect vehicles, the proposed method uses a representation of red colors to find the rear-lights regions of vehicles and uses symmetry measure function to analysis the symmetry of the color distribution, and figures out the accurate position of the symmetry axis. Then, the pair-edges are defined to reconstruct the integrated vehicle edges. Finally, a simple bounding box is fit to detected vehicle regions; the result is the location and dimensions of vehicles in the video. Experimental results show that the proposed method can be accurate and robust in detecting vehicles under different weather and lighting conditions, even at night. This method can also be used in many related applications, such as self-guided vehicles and driver assistance systems. © 2008 Springer-Verlag Berlin Heidelberg.
CITATION STYLE
Lei, G., Chao, L., Ting, F., & Zhang, X. (2008). Vehicle detection based on color and edge information. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5112 LNCS, pp. 142–150). https://doi.org/10.1007/978-3-540-69812-8_14
Mendeley helps you to discover research relevant for your work.