Traffic Density Identification From UAV Images Using Deep Learning: A Preliminary Study

This study presents a preliminary investigation into the application of deep learning techniques for the identification of traffic density from unmanned aerial vehicle (UAV) images. The primary objective is to categorize the traffic flows into three classes: low, moderate, and high. The study proposes a VGG16-based framework with the aim of achieving high classification rates. The experimental results, obtained from a curated dataset with a 70:30 data-splitting ratio, demonstrate a good accuracy at 95.67%. The initial findings regarding traffic density identification are notably satisfactory, especially considering the challenges posed by the UA V -based images utilized in the experiments.