Image-Based Vehicle Verification Using Steerable Gaussian Filter

Authors

  • Heri Prasetyo Department of Informatics, Universitas Sebelas Maret (UNS), Surakarta, Indonesia.
  • Wiranto Wiranto Department of Informatics, Universitas Sebelas Maret (UNS), Surakarta, Indonesia.
  • Winarno Winarno Department of Informatics, Universitas Sebelas Maret (UNS), Surakarta, Indonesia.

Keywords:

Maximum Likelihood Estimation, Steerable Gaussian Filter, Supervised Classification, Vehicle Verification,

Abstract

This paper presents a new feature descriptor for a vehicle verification system. The Steerable Gaussian Filter (SGF) is utilized to generate an image feature descriptor. The descriptor is constructed by concatenating the statistical parameters of the SGF filtered output. The Maximum Likelihood Estimation (MLE) estimates the statistical estimator using a heavy-tailed and bell-shaped distribution assumption such as Gaussian, Laplace, or Generalized Gaussian Distribution (GGD). A classifier assigns a class label of the vehicle hypothesis based on an image descriptor. As documented in the experimental results, the proposed feature descriptor achieves a promising result, and it outperforms the state-of-theart vehicle verification systems, making it a very competitive candidate in the practical applications.

Downloads

Published

2018-07-03

How to Cite

Prasetyo, H., Wiranto, W., & Winarno, W. (2018). Image-Based Vehicle Verification Using Steerable Gaussian Filter. Journal of Telecommunication, Electronic and Computer Engineering (JTEC), 10(2-4), 65–69. Retrieved from https://jtec.utem.edu.my/jtec/article/view/4318

Most read articles by the same author(s)