AsianVehicle: Image Dataset of Traditional Asian Vehicles for Computer Vision
Description
The AsianVehicle dataset presents 1600 RGB images of four traditional Bangladeshi vehicles — Auto Rickshaw, Rickshaw, Rickshaw Van, and Leguna — captured in various urban and semi-urban locations across Mirpur, Dhaka. Images were collected under natural daylight using smartphone cameras (Redmi Note 9, Vivo Y11) at ambient temperatures of 25–28 °C to preserve authentic color tones and textures. Each vehicle type includes 400 images taken from multiple angles and distances, enabling robust representation of real-world variations in shape, design, and background. The dataset is manually annotated and organized into raw dataset 400 images per class, making it suitable for benchmarking computer-vision models in fine-grained vehicle classification, cultural heritage preservation, and intelligent transport applications.
Files
Institutions
- Bangladesh University of Business and Technology