Multi-Source Agricultural Image Dataset for Robust Classification and Semantic Segmentation
Description
This dataset includes images of various agricultural items, grouped into four main categories: crops and grains, flowers, fruits, and vegetables. Each folder contains subcategories named using standard, widely understood English terms to support easy use in machine learning tasks. In total, there are over 4,900 .jpg images, each resized to 512×512 pixels. Every subcategory includes 70 images, making the dataset consistent and balanced. Highlights: --4 primary categories --70+ subcategories --All images in .jpg, sized to 512×512 --Labels use consistent English naming Use Cases: --Training image classifiers for fruits, vegetables, and flowers --Supporting agricultural and botanical research --Educational projects or resource-efficient AI systems This dataset is suitable for both research and practical applications in agriculture and Arttificial Intelligence.
Files
Steps to reproduce
1. Download the dataset and unzip it. 2. All images are organized into 4 main folders: Crops/Grains, Flowers, Fruits, and Vegetables. 3. Each subfolder represents a single class label with standardized English names (e.g., Corn, Mango, Rose). 4. All images are .jpg format, resized to 512x512 pixels. 5. Use any standard image loading and preprocessing library to load and preprocess the images. The input size must be 512x512 dimension 6. Class labels are automatically inferred from folder names. 7. This datset can be used directly in any standard image classification or object detection process.