Deep Learning and Remote Sensing Dataset For Turkey's Wildfire 2021 Multispectral Sentinel-2 Satellite Imagery

Published: 12 August 2022| Version 1 | DOI: 10.17632/hgctmx9y6c.1
Ali Al-Dabbagh,
Muhammad Ilyas


Wildfire maps are useful in understanding the growth of burned areas, change detection in vegetation, and decision-making. Wildfire is a complex exercise and is hard to capture using traditional means. The dataset consists of two folders; Train Images and Train Masks. Each image in train images folder has a size of 128*128 pixels and consists of three multispectral bands. The image is saved using the Universal Transverse Mercator (UTM) as the coordinate system, and GeoTiff as a format file. The image has a spatial resolution of 0.00025 degrees. The values of each pixel are saved in a 16-bit unsigned integer with a range of 0 to 65,535. The dataset has 25,563 images containing the burned area's objects. In train masks folder each image is a binary annotation image that consists of two classes; burned area as the foreground and the non-burned area as the background. These binary images are saved in 8-bit unsigned integer where the burned area is indicated by the pixel value of 1, whereas the non-burned area is indicated by 0.



Altinbas Universitesi


Remote Sensing, Image Segmentation, Features Detection, Deep Learning