Detecting diseased crops in Africa using Artificial Intelligence (AI) has been difficult due to lack of datasets.
As a result, African scientists are forced to rely on data from western countries which can produce inaccurate results.
Fortunately, scientists at the Responsible Artificial Intelligence Lab at the Kwame Nkrumah University of Science and Technology (RAIL-KNUST), have been able to produce datasets which makes it possible for African scientists to use them to create accurate systems to detect African plant diseases.
This dataset's collection of annotated photos of leaves from diverse crops, showing both healthy specimens and those affected by illnesses, is one of its main components.
These pictures show the subtle symptoms of crop diseases at various phases of crop development.
The dataset's emphasis on inclusion, which makes sure that it captures the agricultural diversity found in Africa, is essential to understanding its value.
The dataset offers a thorough understanding of disease patterns and manifestations by including annotated pictures of leaves from a range of crops.
Researchers, data scientists, and innovators who want to create specialized and efficient solutions for the agricultural problems the African continent faces will find this wealth of knowledge very useful.
“What we have done at RAIL is to work with local farmers in creating datasets that can be used by developers within the African space.
“This is one of the first steps in our attempts at creating several of such datasets so that AI practitioners within the African space can rely on such data in creating solutions for the African environment,” said principal Investigator of RAIL-KNUST, Prof. Jerry John Kponyo.
Latest Stories
-
Ghana validates first National E-commerce Strategy
4 minutes -
Dagbon Overlord applauds Mahama’s Code of Conduct for his appointees
5 minutes -
You’ve my undivided support in transforming Ghana – Yaa-Na tells Mahama
7 minutes -
Here’s the list of 12 suspects to be charged over GH₵548m NSS ghost names scandal
21 minutes -
Airwave Apocalypse or Political Performance? A Frequency of Suspicion
25 minutes -
Abstinence education alone can’t reduce alarming teenage pregnancy rate – CSOs
29 minutes -
A/R: Duase drivers block road in protest arrest of drivers charging unapproved fares
36 minutes -
Standard Chartered champions business growth through digital banking innovation
42 minutes -
A step toward energy security: Understanding the new Energy Sector Levy
45 minutes -
UEW inaugurates 2026 Public Lecture Series Organising Committee
48 minutes -
Dagbon Overlord commends Mahama’s lean gov’t as symbol of efficient leadership
52 minutes -
Attorney General denies political bias in ORAL prosecutions
54 minutes -
Trump warns Iran of ‘more brutal’ Israeli attacks after major strikes on nuclear and military sites
58 minutes -
Israel disappointed Ghana abstained from IAEA voting on Iran’s non-compliance with nuclear obligations
58 minutes -
10 Invigilators arrested over exam malpractices
1 hour