A Novel Vehicle Detection System

Zakaria, Yehia; Ali, Mohamed Ashraf; Hossam El DIn Hassan Abdelmunim; Yousef, Ahmed Hassan; Ghoneima, Maged; Hammad, Sherif;

Abstract


Histogram of oriented gradient (HOG) feature has been widely used in vehicle detection. In this paper, a modified version of HOG is proposed by introducing compass gradient into the HOG calculation. Three different versions of the modified HOG features are used as an input for linear and nonlinear support vector machine (SVM). The modified HOG variants proved to have better classification performance than that of the standard HOG. The classification results of modified HOG and nonlinear SVM are compared to the classification results of YOLO object detector. Finally, a vehicle detection system based on the best performing classifiers is introduced.


Other data

Title A Novel Vehicle Detection System
Authors Zakaria, Yehia; Ali, Mohamed Ashraf; Hossam El DIn Hassan Abdelmunim ; Yousef, Ahmed Hassan ; Ghoneima, Maged; Hammad, Sherif
Keywords CNN;Histogram of Oriented Gradient;HOG;Support Vector Machines;SVM;Vehicle detection;YOLO
Issue Date 2-Jul-2018
Conference Proceedings 2018 13th International Conference on Computer Engineering and Systems Icces 2018
ISBN [9781538651117]
DOI 10.1109/ICCES.2018.8639380
Scopus ID 2-s2.0-85063124736

Recommend this item

Similar Items from Core Recommender Database

Google ScholarTM

Check



Items in Ain Shams Scholar are protected by copyright, with all rights reserved, unless otherwise indicated.