I spy on lethal crop diseases using cameras

Friday June 17 2016

Daniel Maitethia checks on the Electronic System for Automatic Crop Infection Monitoring system in a tomato farm. The technology tracks and notifies the farmer via a short message on an invasion of pests like white flies and diseases on the farm and the possible measures to be taken. PHOTO | CAROLINE WAMBUI | NATION MEDIA GROUP


On a small farm at the Meru University of Science and Technology, four students and a lecturer are looking at a laptop screen with great interest.

“As you can see from the screen, the disease has attacked several parts of the plants and if we do not control it as first as possible, then all the crops will die,” Daniel Maitethia, 37, a physics and electronics lecturer informs the students.

The lecturer and the students are looking at a computer application that helps detect the bacterial wilt disease on tomato farms.

Maitethia has worked on the app that he has named Electronic System for Automatic Crop Infection Monitoring for the past five years with agricultural expert at the university, ensuring it works to perfection.

The technology tracks and notifies the farmer via a short message invasion of pests like white flies and diseases on the farm and the possible measures to be taken.

However, once the disease is identified, an agronomist is informed for pesticide application.
“The technology comprises of a raspberry pi camera interlinked to a computer system programmed to capture images of the crops in the fields after every 15 minutes and process them using algorithms to determine the nature of the infection or infestation,” explains Maitethia, who holds a Masters degree in physics from the University of Nairobi and is currently pursuing a PhD in Electronic and Instrumentation at Kenyatta University.


Daniel Maiteithia demonstrates to the students how the system works. PHOTO | CAROLINE WAMBUI | NATION MEDIA GROUP

A Global System for Mobile communication unit is also attached to the computer to communicate with the farmer when an anomaly is detected.

On the tomato farm at the university, the system has been used since September last year and is programmed to diagnose bacterial wilt in tomatoes and is currently being ‘trained’ to diagnose other diseases like tomato blight and leaf rust.


Bacterial wilt is one of the biggest enemies of the tomato plant. The disease attacks the important crop causing it to start drying in less than five days.

Once a section of the land is infected, the soil-borne disease will attack crops of the solanacea family like potatoes, capsicum and eggplant planted on the portion.

“The camera is placed in the greenhouse while the computer system is placed in the laboratory though a portable computing system is being installed in the greenhouse.

The test has been found fool-proof when it comes to detection of bacteria wilt,” says Maitethia, who was motivated to work on the project after doing a similar one during his Masters studies, where he worked on training computers to understand malaria diagnosis without human intervention.


Maitethia with some of his students examining the crops at the farm. PHOTO | CAROLINE WAMBUI | NATION MEDIA GROUP

The camera picks and detects colour variations on the leaves, leaf shape and rate of growth as it takes images periodically before identifying the disease.

According to Maitethia, the technology saves the farmer huge costs incurred on various pesticides that would have been incurred before proper diagnosis.

Under the traditional way, a farmer collects samples and takes them to the laboratory for testing, which takes far too long.

“The real-time monitoring enables the infection to be detected early before it even becomes visible to the farmer’s eye.

“The farmer can, therefore, not guess on what to treat,” he says, noting that the system can also be adopted to alert a farmer when crops are ready for harvesting and estimating the quality of harvest as well as serving as CCTV surveillance.

The process of perfecting the gadget has been cumbersome and quite involving with the testing phase being the most demanding, according to Maitethia.

“We often lacked an ideal environment with specialised laboratories as the university is still growing,” says the lecturer, adding that with about Sh43,000, one can acquire the technology.


Away from the disease-monitoring app, Maitethia and his team have further developed a technology named Sensor-based Automatic Irrigation System that helps irrigate a farm automatically.

“It works using sensors, where the system automatically detects the moisture level content in the soil and commands itself to either open up or stop water for flowing to the plants,” says the lecturer.
“If the moisture content level is high, the system automatically shuts itself and opens up to water the crops when it detects otherwise.”


Maiteithia with some of his students inspecting a tomato farm. PHOTO | CAROLINE WAMBUI | NATION MEDIA GROUP

According to Maitethia, his technology which they have been using for five years at the university, aims at curbing water wastage while at the same time increasing the acreage under irrigation.

Besides the computer, other gadgets the app works with include electronically controlled valves, power supply modules (solar panels and battery), computerised control units and water tanks.

“The system can also be controlled remotely by a farmer by sending a short message text to instruct it to either shut or water,” James Karuri a fourth year student working on the project states, adding that in case of a technical hitch and water cannot be delivered to the crops, yet the soil is dry, the farmer receives the information via SMS.

The technology, according to the student, works well with drip lines, where the land is subdivided into different portions.

0n the tomato farm at the university where the technology has been installed, nothing is left to chance as the irrigation water in a tank undergoes ultraviolet purification to destroy any bacteria and other macro organisms while the soils are sterilised with very hot water heated with solar power.

Felix Muthamia, an agronomist at Meru University, says the project has proved effective as it has saved them water, time, labour and helped increase productivity.