Category: Time series and analysis
Change Detection
Change detection identifies meaningful differences between dates, such as harvest, flooding, deforestation, or construction.
Also known as: change analysis, alerting
Expanded definition
Change detection compares observations across time to identify where something changed on the ground. Methods range from simple thresholds on indices to sophisticated time-series models.
The hard part is separating real change from noise and artifacts. Clouds, shadows, haze, BRDF effects, and inconsistent preprocessing can all look like change.
Robust change detection usually includes quality masks, normalization, and a definition of what counts as change for the application. For monitoring, the goal is often stable alerts with known false positive behavior, not just high sensitivity in one-off tests.
Related terms
Time Series
A time series is a sequence of observations over time for the same location, used for monitoring and change detection.
Cloud Shadow
Cloud shadow is the darkening of the surface caused by clouds blocking sunlight, often mistaken for real change.
Harmonization
Harmonization reduces differences between scenes or sensors so values are more comparable across time.
NDVI
NDVI is a vegetation index that uses red and near-infrared reflectance to approximate vegetation greenness and vigor.