This repository contains an inventory of traffic areas (called "traffic area map (TAM)") covering the city of Brunswick, Germany. The classification is based on image sequences acquired during flight campaigns at different times of the day and year in 2019 and 2020. Each aerial image is segmented by a neural network, and multi-temporal fusion is used to improve robustness.

The data set is available as 8-bit Cloud Optimized GeoTIFF (COG) in the WGS 84 / UTM zone 32 North projection (EPSG: 32632). Its dimensions are 56,000 pixels (X-axis) and 90,000 pixels (Y-axis), resulting in a pixel size of 10 cm.

The individual bit positions indicate the class to which they belong. The meaning of the bits is as follows:

Bit		Description
1		Class: Parking Area
2		Class: Road
3		Class: Access Way

A comprehensive publication with all technical details and examplary use cases in transportation research is currently being finalized.
For now, please cite: Hellekes, J., Merkle, N., López Díaz, M., Henry, C., Heinrichs, M., Azimi, S., Kurz, F., 2021. Assimilation of parking space information derived from remote sensing data into a transport demand model. ITS World Congress 2021, Book of Abstracts, pp. 2579-2590.

For further information please contact: jens.hellekes@dlr.de