Optimal Feature Selection for Pedestrian Detection Based on Logistic Regression Analysis

Author

Jonghee Kim ; Jonghwan Lee ; Chungsu Lee ; Eunsoo Park ; Junmin Kim ; Hakil Kim ; Jaeeun Lee ; Hoeri Jeong

Author_Institution

Sch. of Inf. & Commun. Eng., Inha Univ., Incheon, South Korea

fYear

2013

fDate

13-16 Oct. 2013

Firstpage

239

Lastpage

242

Abstract

This paper describes a pedestrian detection method using feature selection based on logistic regression analysis. As the parent features, Haar-like and Histograms of Oriented Gradients (HOG) features are used manually. For the statistical analysis, stepwise forward selection, backward elimination, and Least Absolute Shrinkage and Selection Operator (LASSO) methods are applied to our Logistic Regression Model for Pedestrian Detection (LRMPD). The experimental results shows that the average of 48.5% of a full model are selected for LRMPD and this classifier shows performance of up to 95% for detection rate with an approximately 10% false positive rate. Processing time for one test image is about 1.22ms.

Keywords

feature extraction; gradient methods; image classification; object detection; pedestrians; regression analysis; traffic engineering computing; HOG; Haar-like features; LASSO; LRMPD; classifier; histograms of oriented gradient features; least absolute shrinkage and selection operator; logistic regression analysis; logistic regression model for pedestrian detection; optimal feature selection; pedestrian detection method; test image; Educational institutions; Equations; Feature extraction; Histograms; Logistics; Mathematical model; Regression analysis; feature selection; logistic regression; multi-feature; pedestrian detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Systems, Man, and Cybernetics (SMC), 2013 IEEE International Conference on

Conference_Location

Manchester

Type

conf

DOI

10.1109/SMC.2013.47

Filename

6721800