A Labeled Dataset of Healthy and Diseased Maize from Tanzania

The maize images dataset was developed to support research in diagnosing major maize diseases and improving crop yields. Sponsored by GrowFurther and conducted by researchers at the Nelson Mandela African Institution of Science and Technology (NM-AIST) in collaboration with the Tanzania Agriculture Research Institute (TARI), the project produced a labeled dataset featuring healthy maize, Maize Lethal Necrosis and Maize Streak Virus. The dataset is designed for image classification and object detection tasks.

All images are in the .zip files; healthy maize leaf image files are “healthy1.zip – healthy8.zip” files each has 11,000 images, “healthy9.zip” has 5,164 images and “healthy10.zip” has 5,163 images. Maize Lethal Necrosis (mln) leaf images are in “mln1.zip – mln7.zip” files each with 11,628 images, “mln8.zip” and “mln9.zip” files each has 5,821 images. Maize Streak Virus (msv) leaf images are in “msv1.zip – msv4.zip” files each with 10,300 images, “msv5.zip” has 2,744 images, “msv6.zip” has 5,487 images and “msv7.zip” has 2,743 images. A total of 243,539 leaf image files are labelled.