deforestation-detector | Detect deforestation from satellite data timeseries

Deforestation Detector is a Python library. Deforestation Detector has no bugs, it has no vulnerabilities, it has a Permissive License and it has low support. However Deforestation Detector build file is not available. You can download it from GitLab.

Detect deforestation from satellite data timeseries by comparing observations with a predefined historical mean and variation. Observations are translated to deforestation probabilities, given the historical statistics. Observations that pass a probability threshold are flagged, and subsequent observations are added in a bayesian manner ('bayesian updating'). If the posterior probability of the bayesian updating passes the confirmation threshold, a detection is made and the flagging and confirmation date for that pixel are saved. If on the other hand the posterior probability after updating goes down again and crosses the unflagging threshold, the flagging event was probably an anomaly, and the flagging is canceled. Implementation is based on this paper and was done in the scope of the Radar Alerts for Detecting Deforestation. Deforestationdetector implements the methods developed by Wageningen University in a scalable way, making it possible to apply it to arbitrary large datasets with good computational performance.