Current Projects

Agricultural Economics Unit

Figure: LULC map of Kegalle District of Sri Lanka developed using Random Forest (RF) machine learning algorithm

Using a machine learning framework for a regional scale application, spectral and topographic information are fused to map land usage.

The Agricultural Economics unit, the Soils and Plant Nutrition Department, and the Biometry Section are researching to investigate land use and land cover (LULC) information, focusing on rubber land change detection during the recent decade. The work reported here uses remotely sensed datasets and machine learning algorithms to examine the accuracy of three categorization approaches in mapping LULC categories throughout time in the study area, largely utilizing Google Earth Engine (GEE) and machine learning algorithms.