Asian subcontinent dataset (ASCD)

The Asian Subcontinent Dataset (ASCD) is a multisensor dataset collected from different countries in the Asian subcontinent, specifically India, Bangladesh, and Sri Lanka. The dataset consists of remote sensing images from two primary sources: Google Earth Pro and IKONOS-2 satellite images. ASCD contains 14 land cover classes that represent various geographical features and man-made structures. The dataset includes 1674 images with a size of 227 × 227 × 3 (width × height × RGB channels).

This dataset is ideal for training and evaluating machine learning models in areas like land cover classification, environmental monitoring, and remote sensing.

Description of Each Land Cover Class:

1. Agricultural: Images representing agricultural fields, crops, and farmlands.
2. Desert: Images of dry, arid regions with sparse vegetation.
3. Port: Images of seaports or harbor areas.
4. Airport: Images of airports, runways, and airstrip areas.
5. Forest: Forested areas with a significant presence of trees and vegetation.
6. River: Images of rivers, streams, and freshwater bodies.
7. Beach: Coastal areas with sandy beaches.
8. Freshwater: Ponds, lakes, and other freshwater bodies.
9. Runway: Images of runways used for aviation.
10. Bridge: Images of bridges or overpasses.
11. Medium Residential: Residential areas with a medium population density.
12. Seawater: Coastal water bodies such as oceans and seas.
13. Denser Residential: Densely populated urban residential areas.
14. Playground: Parks, open spaces, and recreational areas.

Each class folder contains a number of images ranging from 100 to 204 images, providing a total of 1674 images.

Image Format:

• Each image is of size 227 × 227 × 3 (width × height × RGB channels).
• Images are stored in common formats like JPEG.