Pest & Disease Database
TOM2024 is an open collection of field images of tomato, onion and maize, healthy and affected, labelled for building pest and disease recognition models.
- 25,844raw images from the field
- 12,227labelled images
- 30classes across 3 crops
- 200+drone surveys with multispectral imagery
Real photos from farms in Burkina Faso
Trained extension officers collected the images on mobile phones during two field campaigns, in the 2022 rainy season and across the 2023 dry and rainy seasons. Each image was labelled by trained technicians and checked by the project's entomologist and plant pathologist.
The images show healthy and affected crops in many lighting conditions and backgrounds, which helps models work outside the lab. The raw images were cropped to produce the labelled set used to train PlanteSaine.
Dataset at a glance
Name: TOM2024
Crops: tomato, onion, maize
Published: Mendeley Data, 5 September 2024
Image size: high-resolution, for example 4128 x 3096 pixels
What the dataset covers
Class names in the data are in French. English equivalents are shown here.
Tomato (14 classes)
Pests: tomato fruitworm (Helicoverpa armigera), tomato leaf miner (Tuta absoluta)
Diseases and conditions: viral disease, mites, Alternaria, bacterial wilt, blossom-end rot, late blight, sunscald, nitrogen excess, Fusarium, healthy leaf, healthy fruit
Onion (6 classes)
Pests: caterpillars
Diseases and conditions: Fusarium, Alternaria, viral disease, bulb rot, healthy leaf
Maize (10 classes)
Pests: fall armyworm (and its feeding damage), aphids
Diseases and conditions: maize streak, Curvularia leaf spot, Helminthosporium leaf blight, viral disease, rust, abiotic disorder, healthy leaf
Drone imagery
Processed images from more than 200 drone surveys of maize, onion and tomato farms are published on a map with the exact farm locations. Each survey provides six bands: RGB, blue, green, red, near-infrared and red edge. Because the files are large, researchers request them through a form.