With the upcoming Sentinel-2 missions dedicated to land monitoring an unprecedented data stream will become available. The requirements for such a large data stream involve processing techniques enabling the spatio-temporally explicit quantification of vegetation properties. The spatial resolution of Sentinel-2 will make it also applicable to precision agriculture. Biophysical parameters such as leaf chlorophyll content (LCC) and leaf area index (LAI) are standard vegetation products that can be retrieved from Earth observation imagery.
This study explores the use of a new machine learning regression algorithms (MLRAs) toolbox as part of the scientific Automated Radiative Transfer Models Operator (ARTMO) software package. ARTMO facilitates retrieval of biophysical parameters from remote observations in a MATLAB graphical user interface (GUI) environment. The MLRA toolbox enables analyzing the predictive power of various MLRAs in a semiautomatic and systematic manner, and applying a selected MLRA to multispectral or hyperspectral imagery for mapping applications. It contains both linear and nonlinear state-of-the-art regression algorithms, in particular linear feature extraction via principal component regression (PCR), partial least squares regression (PLSR), decision trees (DTs), neural networks (NNs), kernel ridge regression (KRR), and Gaussian processes regression (GPR). MLRAs have the advantage of making use of the full optical spectrum as well as flexible, non-linear fitting.
From 2010 till 2014 several field experiments were conducted on potato fields in the South of the Netherlands. Within a field, a number of plots with different levels of nitrogen fertilization were applied. The nitrogen status and leaf chlorophyll content of the crop were measured (bi-)weekly using the Minolta Spad instrument in the field. Also fresh and dry aboveground weight, dry matter content, leaf area index (LAI) and total nitrogen concentration in the aboveground parts were determined. For all plots, spectral measurements were made during the growing seasons on a bi-weekly basis using a 16-band Cropscan field radiometer.
- Explore the implementation of MLRAs in the ARTMO toolbox
- Study the applicability of multiple implemented regression strategies using the available experimental dataset
- Evaluate the performance of the various regression strategies in terms of accuracy, bias, and robustness
- GRS-32306 – Advanced Earth Observation
- Affinity with scripting (e.g. MatLab) is a preference
Theme: Sensing & measuring