As a common dataset for change detection, its image can be divided into three parts: the image before the change, the image after the change, and the label image showing the changed area. This dataset is characterized by significant seasonal differences between bi-temporal image pairs, which makes up for some of the deficiencies in existing datasets. The labels for this dataset include some irregular changes, such as the appearance and disappearance of cars; but do not include seasonal changes, such as changes in the ground surface caused by snowfall.

Dataset Files

You must be an IEEE Dataport Subscriber to access these files. Subscribe now or login.

[1] M. A. Lebedev, "Change Detection Dataset", IEEE Dataport, 2023. [Online]. Available: Accessed: Jun. 23, 2024.
doi = {10.21227/gg9z-0802},
url = {},
author = {M. A. Lebedev },
publisher = {IEEE Dataport},
title = {Change Detection Dataset},
year = {2023} }
T1 - Change Detection Dataset
AU - M. A. Lebedev
PY - 2023
PB - IEEE Dataport
UR - 10.21227/gg9z-0802
ER -
M. A. Lebedev. (2023). Change Detection Dataset. IEEE Dataport.
M. A. Lebedev, 2023. Change Detection Dataset. Available at:
M. A. Lebedev. (2023). "Change Detection Dataset." Web.
1. M. A. Lebedev. Change Detection Dataset [Internet]. IEEE Dataport; 2023. Available from :
M. A. Lebedev. "Change Detection Dataset." doi: 10.21227/gg9z-0802