An Analysis of Concealed Object Detection Using Decision Tree and Random Forest Algorithms

Shintre, Pallavi and Mahapatra, Shweatlana and Vincent, Shweta and Kumar, Om Prakash (2018) An Analysis of Concealed Object Detection Using Decision Tree and Random Forest Algorithms. International Journal of Engineering and Technology(UAE), 7 (4.41). pp. 154-157. ISSN 2227-524X

[img] PDF
6032.pdf - Published Version
Restricted to Registered users only

Download (141kB) | Request a copy

Abstract

This paper presents a comparative study of the Decision tree algorithm and Random Forest algorithm, both using Haar wavelet transform to classify a concealed object as a threat or not. This finds its applications in airports and railway stations where passenger security is a major concern. A sub-millimeter wave image of a person having a concealed weapon on his thigh has been treated as the test dataset. The Haar wavelet transform along with the aforementioned algorithms is applied on the image to classify the patches in the images as threat or no threat regions. It is found that the Random Forest algorithm outperforms the Decision tree algorithm in terms of accuracy of detection as well as number of false positive generation

Item Type: Article
Uncontrolled Keywords: Decision tree algorithm, Random forest algorithm, Haar wavelet transform, Millimeter waves
Subjects: Engineering > MIT Manipal > Electronics and Communication
Engineering > MIT Manipal > Mechatronics
Depositing User: MIT Library
Date Deposited: 30 Jan 2019 10:53
Last Modified: 30 Jan 2019 10:53
URI: http://eprints.manipal.edu/id/eprint/153113

Actions (login required)

View Item View Item