TY - GEN
T1 - Detecting pedestrians at very small scales
AU - Spinello, Luciano
AU - Macho, Albert
AU - Triebel, Rudolph
AU - Siegwart, Roland
PY - 2009/12/11
Y1 - 2009/12/11
N2 - This paper presents a novel image based detection method for pedestrians at very small scales (between 16 × 20 and 32 × 40). We propose a set of new distinctive image features based on collections of local image gradients grouped by a superpixel segmentation. Features are collected and classified using AdaBoost. The positive classified features then vote for potential hypotheses that are collected using a mean shift mode estimation approach. The presented method overcomes the common limitations of a sliding window approach as well as those of standard voting approaches based on interest points. Extensive tests have been produced on a dataset with more than 20000 images showing the potential of this approach.
AB - This paper presents a novel image based detection method for pedestrians at very small scales (between 16 × 20 and 32 × 40). We propose a set of new distinctive image features based on collections of local image gradients grouped by a superpixel segmentation. Features are collected and classified using AdaBoost. The positive classified features then vote for potential hypotheses that are collected using a mean shift mode estimation approach. The presented method overcomes the common limitations of a sliding window approach as well as those of standard voting approaches based on interest points. Extensive tests have been produced on a dataset with more than 20000 images showing the potential of this approach.
UR - http://www.scopus.com/inward/record.url?scp=76249109326&partnerID=8YFLogxK
U2 - 10.1109/IROS.2009.5354463
DO - 10.1109/IROS.2009.5354463
M3 - Conference contribution
AN - SCOPUS:76249109326
SN - 9781424438044
T3 - 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2009
SP - 4313
EP - 4318
BT - 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2009
T2 - 2009 IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2009
Y2 - 11 October 2009 through 15 October 2009
ER -