文章基本信息

标题：HOL, GDCT and LDCT for Pedestrian Detection
本地全文：下载
作者：Sanaa Tayb ; Youssef Azdoud ; Aouatif Amine 等
期刊名称：Computer Science & Information Technology
电子版ISSN：2231-5403
出版年度：2016
卷号：6
期号：1
页码：87-96
DOI：10.5121/csit.2016.60109
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：In this paper, we present and analyze different approaches implemented here to resolvepedestrian detection problem. Histograms of Oriented Laplacian (HOL) is a descriptor ofcharacteristic, it aims to highlight objects in digital images, Discrete Cosine Transform DCTwith its two version global (GDCT) and local (LDCT), it changes image's pixel into frequenciescoefficients and then we use them as a characteristics in the process. We implementedindependently these methods and tried to combine it and used there outputs in a classifier, thenew generated classifier has proved it efficiency in certain cases. The performance of thosemethods and their combination is tested on most popular Dataset in pedestrian detection, whichare INRIA and Daimler.
关键词：Pedestrian detection; HOL; DCT local; DCT Global; Classification; SVM; Inria Person; Daimler.