Multi-Crop Leaf Disease Dataset for Deep Learning-Based Classification and Explainable AI

This dataset contains a total of 7,255 high-resolution leaf images collected from four economically important crop species: Cashew, Jackfruit, Mango, and Tomato. All images were captured at a resolution of 480 × 360 pixels with 24-bit color depth, ensuring sufficient visual detail for computer vision and deep learning applications. The dataset is designed to support research in plant disease detection, image classification, transfer learning, explainable artificial intelligence, and precision agriculture. It was developed as part of the thesis entitled "Unified Deep Learning Framework for Multi-Crop Leaf Disease Detection Using Transfer Learning and Explainable AI". The dataset consists of 21 classes, including healthy and diseased leaf categories. The disease classes cover fungal, anthracnose, phyllosticta, insect-related, and other leaf disorders across multiple crop species.

Class Labels:
[
"cashew fungal infection",
"cashew healthy",
"jackfruit algal",
"jackfruit anthracnose",
"jackfruit healthy",
"jackfruit leaf spot",
"jackfruit phyllosticta",
"jackfruit sooty mold",
"mango anthracnose",
"mango die back",
"mango gall midge",
"mango healthy",
"mango powdery mildew",
"mango sooty mould",
"tomato bacterial spot",
"tomato blight",
"tomato curl virus",
"tomato healthy",
"tomato insect damage",
"tomato leaf mold",
"tomato spider mites"
]

Dataset Characteristics:
Total Images: 7,255
Number of Classes: 21
Crop Types: 4 (Cashew, Jackfruit, Mango, Tomato)
Image Resolution: 480 × 360 pixels
Color Depth: 24-bit RGB
Data Type: Leaf Images
Image Format: JPG/JPEG