Freight Analysis Using YOLOv2

9 Pages Posted: 16 Jul 2019 Last revised: 30 Sep 2019

See all articles by Snehal Kadam

Snehal Kadam

Rajarambapu Institute of Technology, Islampur 415414, India

Akash Hatalge

Rajarambapu Institute of Technology, Islampur 415414, India

Abhishek Balip

Rajarambapu Institute of Technology

Avinash Powar

Rajarambapu Institute of Technology, Islampur 415414, India

Date Written: May 18, 2019

Abstract

Monitoring traffic of India and calculating the peak hours and density count in a single day helps to develop a required travel and traffic volume estimates, which is required for satisfying all the needs in the planning of roads, its construction, its maintenance and overall administration of the state. Vehicle counting is an important aspect to understand the traffic load and optimize the traffic signals. Detection of vehicles is expected to be more efficient and robust in number of sceneries. Due to improvement in various algorithms and research work, detection mechanism of traffic data analysis has made a significant improvement over traditional methods. Traditional machine learning algorithms and computer vision for object detection now running under slow response time. This problem can be solved by modern architectures and algorithms based on ANN (Artificial Neural Network), like YOLO (You Only Look Once) without any major losses. YOLO and its versions achieved a jaw-dropping performance in computer vision and had achieved a great success in object detection and classification. In this paper, we are presenting vehicle counting, detection and classification based on YOLOv2. Some video sequences have been taken and tested with the planned algorithm. The results can be a solution for planning of new roads or any other diversions for heavy vehicles can be considered during the peak time. A detection mechanism through YOLOv2 differs from other roadway sensors, such as radar or inductive loops, which provide data only regarding traffic flow and density, and do not provide information about the type of the vehicle in real time.

Keywords: Convolutional Neural Network, YOLOv2 Algorithm, Vehicle Counting, Vehicle Detection

JEL Classification: Y60

Suggested Citation

Kadam, Snehal and Hatalge, Akash and Balip, Abhishek and Powar, Avinash, Freight Analysis Using YOLOv2 (May 18, 2019). Proceedings of International Conference on Communication and Information Processing (ICCIP) 2019, Available at SSRN: https://ssrn.com/abstract=3420232 or http://dx.doi.org/10.2139/ssrn.3420232

Snehal Kadam (Contact Author)

Rajarambapu Institute of Technology, Islampur 415414, India ( email )

Akash Hatalge

Rajarambapu Institute of Technology, Islampur 415414, India ( email )

Abhishek Balip

Rajarambapu Institute of Technology

Urun Islampur
Maharashtra, 415414
India

Avinash Powar

Rajarambapu Institute of Technology, Islampur 415414, India ( email )

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
36
Abstract Views
307
PlumX Metrics