Drone-based crop monitoring for African banana farms

A new research has been published in which researchers has demonstrated the success of drone, mobile phone and machine learning technology to monitor the banana crops on East, Central and West Africa.

Published in the International Society for Photogrammetry and Remote Sensing (ISPRS) Journal of Photogrammetry and Remote Sensing, the research, which was was carried out in the Democratic Republic of Congo and Republic of Benin, proved that imaging and machine learning can be used to identify pests and other threats to banana trees for all kinds of farmers in the region. The images are fed into a platform that machine learning has taught to identify banana crops and analyse threats with 97 percent.

According to the research, 90 million people in East, Central and West Africa rely on the banana fruit and plantain for their livelihoods, whether as a source of food or income. Most of the farmers are small scale, and their trees are usually grown among other plants in a kind of mixed farming situation.

But climate change and change evolving land use has exposed the crop to the ever-increasing danger of serious diseases, fungal infections and viruses, which also worsens the food security of the continent, rendering it forever at the mercy of foreign donations, which usually come with many strings attached.

The major diseases affecting banana trees include bunchy top disease (BBTD) and Xanthomonas wilt of banana (BXW).

“Threats (to the banana crop) are currently detected by experts in the field using cell phones,” said Michael Gomez Selvaraj, a crop physiologist at the International Centre for Tropical Agriculture (CIAT) in Colombia, who co-authored the research. “But to track and detect diseases across huge tracts of land at country, district or village level, you need a platform that quickly detects threats.”

Enter the DJI Phantom 4.

Well; they first used a SOLO multicopter drone by 3D Robotics, but then defaulted to the Phantom 4 for its better camera. They flew the drone over banana trees, which were planted among other crops. To detect all possible banana plants in the field (either healthy or diseased), the object detection model – which is applied though a mobile app called Tumaini – came into play.

Using a pixel-based classification system, the researchers trained their software platform to tell the difference between banana trees and other plants on mixed-farm systems, through teaching it the patterns and algorithms of banana plants. After that, the newly smart programme got smarter by gaining the ability to analyse the banana plants and tell which ones carried disease.

All this from just analysing an image provided by the drone above.

Like the notorious new addition to football, the Video Assistant Referee (VAR), the combination of drones and smart software only supplies information as a statistic, which will be turned over agricultural extension officers, farming organisations and governments to act.

One again, the fate of humanity as far as famine and starvation are concerned, will fall on the action – or lack thereof – of humans.

“Otherwise potential threats multiply quickly, for example, farmers may give infected crop stems to others, and, in the case of a virus, spread it around the country or district without knowing until it’s too late,” said Selvaraj.

“We can now detect six major banana threats with speed and accuracy with the Tumaini mobile phone app,” he added. The database is free for farmers, organizations and governments to use, and can been downloaded from the Google App store.

The research did not say whether its smart technology can be taught to detect infections on other plants like tobacco and corn, but seeing how smart it is, it might just work.

It just needs a little calibration. Right?


Leave a Comment


Welcome! Login in to your account

Remember me Lost your password?

Lost Password