The Mobile App That Uses AI To Spot Diseased Crops

Earlier this year I looked at an interesting project from researchers at Penn State.  The researchers had devised a ‘wearable’ leaf sensor to help capture the thickness and electrical capacity of leaves.

The system would be capable of continuously monitoring plant water stress, which is hugely important in arid regions.  It’s a process that has traditionally required laborious work in the field.

By contrast, the leaf sensor system provides the ability to measure leaf thickness and electrical capacitance in real-time, something which the team believe has never been achieved before.  Their work was tested in a lab environment on a tomato plant, with measurements taken at five-minute intervals.

They were able to detect noticeable changes in the thickness of the leaf as soil-moisture levels began to decline.

Disease detection

Another team from the university has recently teamed up with peers from the International Institute of Tropical Agriculture to develop an app that can identify East African crop diseases autonomously by using AI.

Suffice to say, the app is at an early stage of its development at the moment, and can only accurately identify cassava, but the team are hoping to rapidly expand its reach to diagnose root, tuber and banana diseases.

It’s being thoroughly tested in the field, with the early results very positive.  The app was capable of detecting disease even when the plant visually appears healthy.  Each diagnoses is capable of being distributed via text message, thus allowing smallholders to access expert advice and support much faster than they currently can.