Deep Learning of Traffic Volume (Vehicle Recognition & Counting)

Authors

  • Iskandar Naqiuddin Mohd Hisham fkee uthm
  • Mohd Fadzli Abd Shaib

Keywords:

Deep Learning, Neural Network, Vehicle Recognition

Abstract

Neural Network (NN) and Deep Learning are considered to learn abstract representations through their hierarchical architecture, which is inspired by the brain. Today, however, it is not widely understood how this happens. This is the technique through which a moving vehicle is recognized with a camera. Capturing automobile from the monitoring camera in video sequence is a challenging application to improve tracking performance. It increases the number of applications like traffic control, traffic monitoring, traffic flow, security etc. This technology. The expected costs will be much lower with this technique. These data provide three comparisons of the prominent method of recognition by vehicles: Local Binary Pattern (LBP) and Transform Scale-Invariant Features (SIFT) and for my research I’m using Neural Network (NN). For the Local Binary Pattern (LBP), the primary finding is 64.6%, for the Scale-Invariant Feature Transform (SIFT) 78.3%, and with the Deep Neural Network (DNN) 88.4 percent. In addition, the Convolutional Neural Network approach is compatible in the Neural Network method. This comparison is based on the compatibility, accuracy (percentage). These discoveries may also explain how the human brain learns abstract. Proof here that Neural Network (NN) learns abstract representations through a demodulation process. The activation function is introduced and it is used to indicate that the neural network (NN) is best studied and performed for demodulation.

Downloads

Published

15-11-2021

Issue

Section

Articles

How to Cite

Mohd Hisham, I. N., & Abd Shaib, M. F. . (2021). Deep Learning of Traffic Volume (Vehicle Recognition & Counting). Evolution in Electrical and Electronic Engineering, 2(2), 826-833. https://publisher.uthm.edu.my/periodicals/index.php/eeee/article/view/3968

Most read articles by the same author(s)