Background Information
Welcome to the technical documentation of the Bird’s-Eye project, a collaboration between Pula and Tetra Tech under the Enabling Crop Analytics at Scale Project. This work, funded by the Bill & Melinda Gates Foundation, is meant to foster the collection of Ground truth georeferenced Agriculture labelled datasets that would improve predictive analytics in smallholder agriculture, through the use of Artificial intelligence and Machine Learning (AI/ML) and satellite imagery.
The twenty-first century has presented the most devastating effects of climate change, especially in sub-Saharan Africa. These effects are packaged with climate variability that adversely affects small-holder farmers.
Towards resilience building, agricultural statistics and farm evidence is proving a key variable in mitigating the effects of climate change. However, these activities have been faced with challenges in data gaps that can be used to train and validate machine learning and artificial intelligence models.
Agricultural statistics (ag-stats) entails the generation of geo-referenced farm information including basic information such as land ownership and size. Ag-stats information will enable the consequent support in generating scientific information, including crop-type mapping, and the generation of crop masks in different agro-ecological zones.
With accurate crop masks data, yield forecasting will then be possible, of which a ripple effect on mitigation measures and preparedness on food security policy can be achieved.
Project Implementation
The project was Implementated in two countries namely;
Kenya
Zambia
Crops of Interest
Various Crops were targeted this includes;
Maize
Potatoes
Millet
Sorghum
Beans
Aknowledgement
Special Appreciation to Tetra Tech, on the support granted during the project implementation phase. We also acknowledge the support from our partners in Kenya, and Zambia
License
Bird’s- Eye Project is an open-source project, licensed under the GNU General Public License, Version 2.0 or later.
Note!
The products generated with boundary names and extends do not imply official endorsement by Pula.