DATOS holds 3-day Seasonal Crop Mapping Training for Department of Agriculture
The Remote Sensing and Data Science: DATOS Help Desk team of the DOST- Advanced Science and Technology Institute (DOST-ASTI) organized a three-day Seasonal Crop Mapping Training with researchers from the Field Programs Operational Division of the Department of Agriculture (DA-FPOPD) last 20-22 January 2021.
The training involved a lecture series for Synthetic Aperture Radar (SAR) image pre-processing techniques and crop detection using Dynamic Time Warping (DTW). The outputs from the pre-processed SAR images were used as inputs for DTW.
DTW is one of the algorithms used to measure the similarity between two temporal sequences (e.g., comparing temporal signatures between crops). It is time-flexible, and it presents the capability to achieve good classification results for mapping crops.
In this method, the temporal signature of an identified crop within its whole crop cycle will be obtained and compared to the temporal signature of another crop. Signatures with a minimal difference will be classified as belonging to the same crop type.
DTW is capable of considering prolonged seasonal changes and has provided promising results for crops such as rice.