German Traffic Intersection Images Dataset

Overview This dataset contains street-level imagery of various intersections in Germany. It was created as part of a university semester project for the course Applied Artificial Intelligence at HTW Dresden. The dataset is intended for research, educational purposes, and computer vision tasks, such as image classification or object detection.

Dataset Structure & Contents The dataset consists of a total of 250 images, organized into six specific categories based on the type of intersection or traffic situation:

Railway Crossing: Images of intersections with railway tracks.

Roundabout: Images of circular intersections.

Signed Intersection: Intersections where the right-of-way is controlled by traffic signs.

Street with Right-Hand Rule: Uncontrolled intersections following the default "right-before-left" priority rule.

Traffic Light Controlled: Intersections regulated by traffic lights.

Unclear Situation: Complex or ambiguous traffic situations and intersections.

Annotations (CSV) Alongside the images, the dataset includes a CSV file containing annotations and metadata. This file lists each image, detailing its specific properties and clearly indicating which classification category it belongs to.

The overall collection includes both originally captured photos and images sourced from Mapillary.