Implementation and Optimization of Wavelet and Symmetry Features for Vision-based Pedestrian Detection

S. Schauland and A. Kummert (Germany)


Applications, Pattern Recognition, Pedestrian Detection, Feature Extraction, Classification, Wavelet Features, F Measure, Support Vector Machines


This paper presents two of the most important parts of a vision-based pedestrian detection system: the feature ex traction and the classification module. Wavelet-based fea tures and a combination of symmetry and edge density fea tures are extracted from a monochrome image captured by a vehicle-mounted camera and fed into an SVM-classifier, more precisely a modified version of libSVM [1]. For both types of features an optimization approach based on im age masks is proposed. In order to weight the impact of classifier results (false negatives are preferred over more false negatives in the case of pedestrian detection) the F measure is used as statistical measure. An overview on the advantages and drawbacks of the implemented features and the optimization approach is given, based on the results re ceived from tests using pedestrian and non-pedestrian im ages extracted from video sequences showing urban traffic scenes.

Important Links:

Go Back