In December 2021, Swift Lab Limited successfully conducted crop health mapping for one of our large-scale avocado clients in Kenya. The main objective was to obtain crop health data (using NDVI data) and use the generated maps to identify areas (in this case trees) that had significant crop stress. As a second step, we would then do ground truthing and take phone images to verify and overlay on top of the NDVI map to generate a complete report.
Why validate through ground truthing?
– This is a new technology that we are introducing at scale in Kenya. Therefore, it was important for us to validate the accuracy of our drone-collected crop health data via ground truthing. In this case for our client, not only did our drones identify areas already known to the client, but also uncovered many more areas that were not known before
– This is a new technology that we are introducing at scale in Kenya. Therefore, it was important for us to validate the accuracy of our drone-collected crop health data via ground truthing. In this case for our client, not only did our drones identify areas already known to the client, but also uncovered many more areas that were not known before.
– We always recommend regular ground truthing exercises as a continuous avenue to validate our data, but also to help our data processing and algorithms to track farm performance and effects of response measures e.g., chemical spray that are done over time.
Example results obtained
– Pest Infestation
map showing a cell phone image taken after following the map to the highlighted red zone.