Vehicle detection from three line scanner image
S. Puntavungkour, Ryosuke Shibasaki · 2004
Nowadays, traffic jam becomes severe rapidly. Therefore we presented the algorithms of vehicle monitoring from Three Line Scanner Imagery. In our contribution, the framework consists of three steps: vehicle detection, moving/ parking vehicle discrimination and vehicle parking identification. Now we focus on vehicle detection by Nadir-Single TLS images. We applied Explicit Model Approach. Firstly, segmented image is created by region and non-vehicle regions are removed We detect the cue of Vehicle from the raw image and generate vehicle candidate from vehicle likely region from segmented image by frame detector expansion technique secondly. At the last step, vehicle in the image have been verified detected by vehicle model matching. Finally, the promising results is shown and listed in this study.