This study explores the impact of video resolution on the performance of machine learning-based traffic monitoring systems. Using a combination of empirical analysis and theoretical modeling, we assess how different resolutions affect detection accuracy, resource consumption, and computational efficiency. While other factors such as noise level, or compression artifacts also influence performance, resolution was chosen as a primary variable due to its critical role in balancing detail capture and computational cost. Higher resolutions can enhance object detection accuracy but also significantly increase data processing demands, making resolution a key trade-off in designing efficient surveillance systems. Findings of this study show significant insights into these trade-offs, guiding transportation authorities and system developers in making informed decisions to design scalable traffic monitoring solutions that meet the demands of modern urban environments.

Nowadays, traffic congestion and air pollution in urban areas become huge challenges to transportation engineering. On the other hand, technologies have evolved, and simulation and artificial intelligence have become important tools to face these challenges. In this paper, a review of this research field is done, and a possible combination of simulation and machine learning to mitigate traffic emissions is proposed. The long-term goal of the research is to establish an effective emission mitigation model for some specific crossings.