This is the first analysis we have carried out at a roundabout. Traffic analysis in roundabouts presents several challenges due to the complexity of the movements made by vehicles and the distance that the camera must have.

Deep Learning techniques were used to automatically identify and track vehicles through images captured by a drone-type aerial camera.

Data were collected on a 10-lane section in one direction over a 60-minute period (07:31-08:32 on September 1, 2023), with emphasis on the following parameters: Traffic Volume, Classification and Direction, and Origin-Destination Matrix.