Denoising of Traffic Statistics in Multi-frame Video Sequences
Dawei Zhang, Dan Huang · Journal of Image Processing Theory and Applications · 2024
The advancements in technology have facilitated the interconnection of all things, with computer vision technology emerging as a prominent field in society. As cities continue to evolve towards greater levels of intelligence, the utilization of cameras as the primary means of collecting visual data for road traffic flow monitoring has become a ubiquitous sight. The current research endeavors to devise a sophisticated vehicle flow denoising algorithm that leverages the power of multi-frame video sequences. By processing images captured from road monitoring systems, this algorithm effectively filters out noise within vehicle flow data, enabling the acquisition of more precise vehicular statistics. Consequently, it facilitates a deeper analysis of vehicle counts and densities, ultimately providing vital data support for managing urban pressures and enhancing road decongestion efforts.