Researchers at Incheon National University in South Korea have developed a smart traffic monitoring system using adaptive cameras.

The system dynamically activates more cameras during busy times and fewer during quiet periods, optimising resource use and improving road safety.

Tested in diverse scenarios, it shows the potential to reduce accidents, ease congestion and conserve energy, making it promising for smarter urban traffic management.

Effective urban traffic management remains a cornerstone of smart city development. With the rise of autonomous vehicles and connected transportation, dynamic surveillance is critical to ensuring smooth traffic flow, reducing accidents and optimising efficiency. However, traditional static camera setups often fail to adapt to rapid changes in traffic conditions, resulting in inefficient monitoring and resource use.

To address this, researchers from Incheon National University (www.inu.ac.kr), led by associate professor Hyunbum Kim, have introduced an augmented fluid surveillance system designed to adapt in real time to varying traffic scenarios.

The system employs a network of single-lens cameras arranged in a dynamic grid. This model intelligently adjusts its surveillance coverage by activating or deactivating cameras based on real-time traffic conditions, ensuring efficient and flexible monitoring.

“Our motivation stems from the growing need for adaptive traffic monitoring systems that can handle diverse and unpredictable scenarios,” said Kim. “By creating an augmented fluid surveillance system, we aim to revolutionise traffic management and provide seamless intelligent transportation services.“

To achieve this, the study formalised the “augmented fluid surveillance efficiency maximisation problem”. This problem focuses on finding the best way to place and use cameras for maximum efficiency while still covering all necessary areas. The researchers came up with two options to address this.

The first approach, called the random-value-camera-level algorithm, organises cameras in a three-by-three grid. Some cameras are always on to ensure basic coverage, while others switch on or off depending on traffic levels. This way, during busy times, more cameras turn on to monitor the situation, and during quiet times, fewer cameras are active, saving energy.

The second approach, called the all-random-with-weight algorithm, works similarly but is even more flexible. It assigns a unique role to each camera based on its position in the grid. Cameras in key positions stay active all the time, while others adjust their activity to match traffic conditions. This method ensures a balance between thorough monitoring and efficient energy use.

Simulations showed these methods worked effectively under different conditions, such as varying traffic levels, slopes and angles. The system reduced energy use during low traffic and provided strong coverage during peak hours by predicting and adjusting to traffic patterns.

“Our approach optimises camera usage and saves energy while ensuring reliable surveillance,” said Kim. “It’s a step towards smarter and more eco-friendly traffic management.”

Beyond traffic control, this adaptive system could also be used for crowd monitoring, disaster response and industrial safety. Future efforts will focus on real-world tests and integrating technologies such as deep learning for better performance.

This paper can be found in the IEEE IoT Journal at ieeexplore.ieee.org/document/10571581.