IMonitor – AI for Monitoring Changes and Food Supply from Space
In recent years, Earth observation (EO) has entered the era of Big Data. Satellite imaging, combined with other data sources (e.g. financial trends, soil information, weather patterns), currently enables us to monitor the Earth almost on a daily basis. Frequent revisits of the same spot on the ground have opened up the possibility for fine-level, task-specific change detection monitoring and understanding applications.
However, due to the complex nature of Earth observation data and the significant number of possible scenarios, current research is still only beginning to understand the opportunities created by analyzing fine-grained multi-temporal information. In the new research group we will develop change detection (CD) methods that can be generalized to different scenarios. To be specific, we focus on:
- A supervised Land Use / Land Cover (LULC) change detection method that can be easily adapted to new applications, using diverse sensors and employing only a few training examples;
- An application-specific model that uses multi-temporal EO data for food supply forecasting.