To use this feature you must be a registered user of Scopus.
Document details - A histogram-refinement Histogram-of-Oriented-Gradient Object and Pedestrian Detector
Journal Volume 9, Issue 6, November - December 2020, Article 9752202 Daniel Asher Mitchell, Harvey Barnett Mitchell , "
A histogram-refinement Histogram-of-Oriented-Gradient Object and Pedestrian Detector" , International Journal of Emerging Trends & Technology in Computer Science (IJETTCS) ,
Volume 9, Issue 6, November - December 2020 , pp.
027-031 , ISSN 2278 - 6856.
A histogram-refinement Histogram-of-Oriented-Gradient Object and Pedestrian Detector
Abstract:The Histogram-of-Oriented-Gradient (HOG) is
widelyused forobject and pedestrian detection. The basic idea of
HOG is that theappearance of an object can be characterized by
the spatial distribution of the intensity of the local gradients and
their directions. In this article we describe a new HOG
detectorin which we use the method of histogram refinement to
split the traditional HOG vector into two complementary vectors
which we subsequently fuse together. The new HOG has an
enhanced performance but at the same time is fully compatible
with the traditional HOG. Experimental results on the INRIA
pedestrian detection database shows that for noisy input images
the histogram-refined HOG significantly outperforms the
traditional HOG detector.