DocumentCode :
1910596
Title :
Local information statistics of LBP and HOG for pedestrian detection
Author :
Brehar, Raluca ; Nedevschi, Sergiu
Author_Institution :
Comput. Sci. Dept., Tech. Univ. of Cluj-Napoca, Cluj-Napoca, Romania
fYear :
2013
fDate :
5-7 Sept. 2013
Firstpage :
117
Lastpage :
122
Abstract :
We present several methods of pedestrian detection in intensity images using different local statistical measures applied to two classes of features extensively used in pedestrian detection: uniform local binary patterns - LBP and a modified version of histogram of oriented gradients - HOG. Our work extracts local binary patterns and magnitude and orientation of the gradient image. Then we divide the image into blocks. Within each block we extract different statistics like: histogram (weighted by the gradient magnitude in the case of HOG), information, entropy and energy of the local binary code. We use Adaboost for training four classifiers and we analyze the classification error of each method on the Daimler benchmark pedestrian dataset.
Keywords :
gradient methods; image processing; learning (artificial intelligence); pedestrians; statistical analysis; traffic engineering computing; Adaboost; Daimler benchmark pedestrian dataset; HOG; LBP; gradient image; gradient magnitude; histogram of oriented gradients; intensity images; local binary patterns; local information statistics; pedestrian detection; statistical measurement; Detectors; Entropy; Feature extraction; Histograms; Support vector machines; Training; Vectors; Histogram of Oriented Gradients; Local Binary Patterns; Local Energy; Local Entropy; Local Information; Pedestrian detection;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Computer Communication and Processing (ICCP), 2013 IEEE International Conference on
Conference_Location :
Cluj-Napoca
Print_ISBN :
978-1-4799-1493-7
Type :
conf
DOI :
10.1109/ICCP.2013.6646093
Filename :
6646093
Link To Document :
بازگشت