Phd student pilots smart traffic-surveillance system for HCM City

February 13, 2021 - 08:23

During his years at International University under the Việt Nam National University-Hồ Chí Minh City, Phan Ngọc Hưng graduated with top honours and was the valedictorian in both his undergraduate and postgraduate degree programmes.

 

Phan Ngọc Hưng, 26, was the valedictorian in both undergraduate and postgraduate programmes at the International University in HCM City. Photo courtesy of International University

Gia Lộc

HCM CITY— During his years at International University under the Việt Nam National University-Hồ Chí Minh City, Phan Ngọc Hưng graduated with top honours and was the valedictorian in both his undergraduate and postgraduate degree programmes.

The 26-year-old, who has won a full scholarship to pursue a doctoral degree at Norway's Tromoso University - The Arctic University, has developed, along with other teammates, a pilot smart traffic-surveillance system for the city.     

“Luckily, I was chosen to take part in the city’s pilot project on the development of a smart traffic-surveillance system which has been installed on Võ Văn Kiệt Avenue. I learned how to apply academic knowledge to solve real problems,” Hưng said.

Hưng was in charge of developing a module for detecting and classifying moving vehicles in crowded scenes. The surveillance system extracts data that assists users in regulating and managing traffic conditions at monitoring sites through static surveillance cameras. 

The system consists of three main parts: an image signal processing centre, simulation centre and traffic operating centre.

The centres provide support for traffic management systems in urban areas where the roads are burdened by steadily increasing traffic flows, causing congestion and delays and contributing to economic and environmental costs.

The main goal is to gain an understanding of traffic situations by extracting data on speed, vehicle type, density and other factors through analysis of videos from the cameras.

However, because this type of system has mostly been carried out in developed countries where cars are the main means of transport, they are not particularly suited to Việt Nam, where motorbikes dominate. 

“These solutions in developed countries also require modern equipment, large costs and big data," Hưng said. "They are unsuitable for complicated transport infrastructure situations in developing countries like Việt Nam. Through our project, I learned about other technologies that could replace these with available sources in Việt Nam.”  

His research team has been developing a way to handle recognition based on the movement of vehicles under crowded conditions. The solutions of the research team will be tested experimentally on actual data samples collected and extracted directly from the surveillance cameras.

The research will play an important role in improving the system installed on Võ Văn Kiệt Avenue, and, at the same time, is a premise for further studies on expansion of the system.

Functions that can be exploited from future research include the automation of peak-hour signals and warnings about potential congestion. This will help reduce traffic pressure at intersections during peak hours.

 

Phan Ngọc Hưng (right), who has won a full scholarship to pursue a doctoral degree in Norway, and a lecturer at International University in HCM City. Photo provided by Hưng

The system will also provide the most accurate statistics and information about traffic so that authorities can upgrade transport infrastructure.

Last year, Hưng and his teammates won second prize at the AI- Challenge Contest held by the city People’s Committee and Department of Science and Technology.

The contest’s theme was artificial intelligence and smart traffic. As part of the contest, teams counted the number of vehicles of all kinds moving in different directions in videos recorded from traffic cameras.

The city will use the contest’s results to design solutions to reduce traffic congestion.

When he was a student at the postgraduate programme at the International University, Hưng became interested in computer vision, which has emerged as a critical research area with a great number of successful applications.   

“Video-based traffic surveillance systems are remarkable exemplars in extracting necessary information about types, speed and other data from cameras for traffic monitoring in many metro cities. In these systems, vehicle detection plays a pivotal role in localising and classifying vehicles, which are captured within the field of view of a camera,” he added.

He is currently studying how to perform visual perception tasks such as object recognition and scene categorization at extremely high levels of speed and efficiency. These tasks could assist security observations as well as the detection of cancerous cells that would improve cancer treatments.

Hưng will continue this kind of research at the Tromoso University - The Arctic University of Norway.

“After finishing my doctoral programme in Norway, I will return to Việt Nam to contribute to the country’s development, especially in the IT field. The country has many problems and there are opportunities for IT engineers to study and develop their own solutions,” Hưng said.  

“This field has been my passion for years. I saw the huge potential of the IT sector when I was in the 12th grade. Students at universities should try to identify what they want to do and their future orientation and make plans for implementation. They should also have self-discipline to ensure that they have enough time for studying and researching." VNS

 

 

E-paper