A team of researchers at National Institute of Technology Rourkela (NIT Rourkela) has developed an innovative ‘multi-class vehicle detection’ (MCVD) model and a ‘light fusion bi-directional feature pyramid network’ (LFBFPN) tool using artificial intelligence. Led by Prof Santos Kumar Das, Associate Professor in the Department of Electronics and Communication Engineering, the team aimed to enhance traffic management in developing countries by leveraging an intelligent vehicle detection (IVD) system.
Traditional IVD systems typically excel in organised traffic environments in developed countries but face challenges in mixed traffic scenarios common in developing nations like India. Here, a wide array of vehicles, pedestrians, and other elements share the roads, making accurate vehicle detection a complex task.
The team recognized that sensor-based IVD systems like radar and LiDAR are effective in controlled environments but struggle in adverse weather conditions and are costly. Video-based systems show promise, but they are hindered by traditional video processing techniques that are not well-suited for fast-moving traffic and require high computational power.
To address these issues, the researchers developed the MCVD model that utilizes a video deinterlacing network (VDnet) to extract key features from traffic images, even when vehicles vary in size and shape. Additionally, they introduced the LFBFPN tool to further refine the extracted details.
Deep learning models, such as CNNs, prove to be efficient in detecting vehicles in video feeds by learning from existing data. However, they often fall short in accurately detecting vehicles of different sizes and angles in busy, mixed-traffic environments. Moreover, the lack of labelled datasets tailored for such complex conditions poses a significant challenge.
By developing the MCVD model and LFBFPN tool, the team at NIT Rourkela aims to optimize traffic flow, reduce congestion, and assist in future road planning in developing countries like India. Their innovative approach showcases the potential of artificial intelligence in revolutionizing traffic management and urban planning practices.
