What does the Land-Use Scene Classification Dataset contain?

The dataset contains satellite images from 21 land-use classes such as buildings, baseball fields, and freeways. Each image is 256×256 pixels, and augmentation expanded each class to 500 images.

What is the main application of this dataset?

The dataset is mainly used for land-use classification, remote sensing research, geospatial AI projects, and urban planning applications.

How many classes are included in the dataset?

The dataset includes 21 land-use classes covering diverse environments such as buildings, roads, sports fields, and natural regions.

Why is data augmentation applied in this dataset?

Data augmentation was applied to increase robustness. Each class originally had 100 samples, which were augmented 4 times to create 500 samples per class for more effective machine learning training.

Is the Land-Use Scene Classification Dataset free to use?

Yes, the dataset is released under the CC0 Public Domain Dedication, making it free for academic, research, and commercial use.

Which fields benefit from this dataset?

The dataset is useful in computer vision, geospatial AI, earth observation, environmental monitoring, and smart city urban planning.

Land-Use Scene Classification

Home » Dataset Download » Land-Use Scene Classification

Land-Use Scene Classification

Datasets

File

Land-Use Scene Classification

Use Case

Computer Vision

Description

The main aim of this project is the masking of regions where land is being used in satellite images obtained through Landsat which have a low spatial resolution.

About Dataset

Context

The main aim of this project is the masking of regions where land is being used in satellite images obtained through Landsat which have a low spatial resolution.

Content

This dataset contains satellite images of 21 classes such as buildings, baseball fields, freeways, etc. The original size of the images is 256×256 pixels. Originally there were 100 images per class. After augmenting each image 4 times the size of each class was brought up to 500 images. This allows for making a more robust model.

Acknowledgements

This dataset is sourced from Kaggle.

Credits –

Dataset: Yi Yang and Shawn Newsam, “Bag-Of-Visual-Words and Spatial Extensions for Land-Use Classification,” ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems (ACM GIS), 2010.

Contact Us

Let's Discuss your Data collection Requirement With Us

To get a detailed estimation of requirements please reach us.

Land-Use Scene Classification