From Above the Sky to Below the Earth: Crop type classification using satellite imagery and deep learning

Researcher:  Yusuf Mansoor, University of the Witwatersrand, Johannesburg
Supervisor: Prof Adam Elhadi, University of the Witwatersrand, Johannesburg

Crop type mapping and classification is necessary for optimal cropland management. Remote sensing with satellite imagery has gained popularity due to the ease of accessibility and availability.  For this study deep learning neural networks will be assessed due to their superior accuracy and robustness.  The aim of the study was to assess the performance of NN for crop classification. And to determine the optimal temporal window for classification.