Science

Under-leaf sensors could provide live reports on the well-being of crops

Under-leaf sensors could provide live reports on the well-being of crops
The sensors are attached to the underside of the leaves (right) in order to ensure that the plant still receives enough sunlight
The sensors are attached to the underside of the leaves (right) in order to ensure that the plant still receives enough sunlight
View 2 Images
The system's algorithms were trained on approximately 90 variously colored leaves collected from 30 plant species
1/2
The system's algorithms were trained on approximately 90 variously colored leaves collected from 30 plant species
The sensors are attached to the underside of the leaves (right) in order to ensure that the plant still receives enough sunlight
2/2
The sensors are attached to the underside of the leaves (right) in order to ensure that the plant still receives enough sunlight

When it comes to monitoring the health of crops, aerial images captured by aircraft flying high overhead are only going to tell you so much. That's where an experimental new spectral sensor comes in, as it's mounted directly on the underside of individual plants' leaves.

Developed by Ko-ichiro Miyamoto, Kaori Kohzuma and colleagues at Japan's Tohoku University, the device is intended to be used on a number of "sentinel plants" located in different areas of each field. It consists of two parts.

There's the sensor head itself, which is taped to the underside of the leaf where it won't block the plant's exposure to sunlight. That head is cable-connected to an adjacent control unit which incorporates a lithium battery, Arduino microprocessor, antenna, and other electronics.

At regular intervals, two LEDs in the sensor head briefly illuminate the underside of the leaf. A photodiode located between those LEDs analyzes the spectrum of the light that is reflected back. The diode also takes a reading with the LEDs shut off, to measure the sunlight that is passing through the leaf from above.

By subtracting the second measurement from the first, it's possible to determine the spectrum of LED light that is being absorbed by the leaf. This reading is in turn used to determine the present color of the leaf, which is an indicator of plant health (and of season).

The sensor data is processed and presented to users by a cloud server, which the sensor accesses via a cellular network.

The system's algorithms were trained on approximately 90 variously colored leaves collected from 30 plant species
The system's algorithms were trained on approximately 90 variously colored leaves collected from 30 plant species

In a test of the technology, a number of the sensor units were attached to the leaves of golden birch trees which were growing in the experimental garden of the University of Tokyo. The sensors automatically took readings once every two hours over a period of two weeks, as the leaves naturally turned from green to yellow or brown in color.

When the sensor readings were subsequently analyzed, they were found to closely match readings which were obtained over the same period by people using handheld spectrometers. In large-scale agricultural settings, however, manually taking readings in the latter manner would be considerably more time- and labor-intensive than using the sensors.

The fabrication cost of each of the prototype units was "a few tens of dollars," although that figure would likely drop drastically if they were being commercially manufactured in a factory.

"This affordable sensor is a promising tool for accurately monitoring plant health and stress through leaf color and light reflection data," says Kohzuma. "Its low cost makes it possible to place multiple sensors across various locations, creating a network for simultaneous monitoring in many spots."

A paper on the research was recently published in the journal Sensing and Bio-Sensing Research. Scientists from MIT, North Carolina State University, and Iowa State University are also working on plant-worn stress-detecting sensors of their own.

Source: Tohoku University

No comments
0 comments
There are no comments. Be the first!