In the heart of Burkina Faso, a groundbreaking initiative is underway that could revolutionize how we approach pest and disease management in agriculture. Led by Obed Appiah, a researcher at the Competence Centre, WASCAL, in Ouagadougou, Burkina Faso, and the Department of Computer Science and Informatics at UENR, Sunyani, Ghana, the TOM2024 dataset is set to become a cornerstone in the development of AI-based classification models for identifying pests and diseases in maize, tomato, and onion crops.
The TOM2024 dataset, published in ‘Data in Brief’, is more than just a collection of images; it’s a comprehensive resource designed to bridge critical gaps in existing datasets, particularly in developing regions. With over 25,844 raw images and more than 12,000 labeled images, the dataset covers 30 classes, including healthy crops, infested crops, and various pests. This extensive collection was meticulously gathered through fieldwork, ensuring that it captures diverse environmental conditions and crop stages.
“Our goal was to create a dataset that not only provides high-resolution imagery but also reflects the real-world challenges faced by farmers,” Appiah explains. “By including a wide range of pests and diseases, we aim to develop AI models that can accurately identify and classify these issues, ultimately aiding in precision agriculture and sustainable farming practices.”
The dataset is segmented into three categories: processed images (Category A), selected images with augmentation (Category B), and an online repository with over 25,000 raw images (Category C). Category A and B feature images of crops affected by 21 distinct pests and diseases, providing a robust foundation for training and validating AI models.
The implications of this research are vast. For the energy sector, which often relies on agricultural byproducts for biofuels and other renewable energy sources, the ability to quickly and accurately identify pests and diseases can lead to more efficient and sustainable crop management. This, in turn, can enhance the quality and quantity of agricultural waste used for energy production, reducing reliance on fossil fuels.
Moreover, the TOM2024 dataset’s versatility extends beyond research and educational purposes. It can be a practical tool for farmers, enabling them to use digital technologies to monitor their crops and take timely action against pests and diseases. This not only improves crop yields but also contributes to global food security.
As we look to the future, the TOM2024 dataset could shape the development of more advanced AI models, leading to automated pest and disease detection systems. These systems could be integrated into smart farming technologies, providing real-time insights and recommendations to farmers. This could be a game-changer for the agricultural sector, especially in regions where labor and resources are limited.
The TOM2024 dataset, published in ‘Data in Brief’, is a testament to the power of digital technologies in transforming agriculture. By providing a comprehensive and high-resolution collection of images, it paves the way for innovative solutions that can enhance crop management, improve food security, and contribute to sustainable agricultural practices. As Appiah puts it, “This dataset is just the beginning. We hope it will inspire further research and development in the field, leading to more efficient and sustainable farming practices globally.”