Title :
Bicycle detection using pedaling movement by spatiotemporal Gabor filtering
Author :
Takahashi, Kazuyuki ; Kuriya, Yasutaka ; Morie, Takashi
Author_Institution :
Grad. Sch. of Life Sci. & Syst. Eng., Kyushu Inst. of Technol., Kitakyushu, Japan
Abstract :
This paper proposes an algorithm for detecting bicycles on the road from a sequence of images acquired by a stationary video camera. This algorithm is suitable for monitoring the traffic situation for safety, our algorithm includes shape-based object detection using HOG and SVM and relative motion detection for leg movements in bicycle pedaling. The leg movement, which can be achieved by using spatiotemporal 3D Gabor filtering, can discriminate bicycles from similar objects such as motorbikes.
Keywords :
Gabor filters; image sequences; motion estimation; object detection; support vector machines; traffic information systems; video cameras; video surveillance; HOG; SVM; bicycle detection; images sequence; leg movements; motion detection; pedaling movement; shape-based object detection; spatiotemporal Gabor filtering; stationary video camera; traffic monitoring; bicycle detection; relative motion; spatiotemporal Gabor filter; video surveillance;
Conference_Titel :
TENCON 2010 - 2010 IEEE Region 10 Conference
Conference_Location :
Fukuoka
Print_ISBN :
978-1-4244-6889-8
DOI :
10.1109/TENCON.2010.5686551