IIIT Allahabad Develops Ai For Crop Disease Detection

Researchers at the Indian Institute of Information Technology, Allahabad (IIIT-A) have introduced a new Artificial Intelligence (AI) system that can detect crop diseases in real time. This innovation allows farmers to monitor their crops without needing to consult agricultural experts.

The project is led by research scholar Pramod Kumar Singh and guided by Prof. Manish Kumar, who heads the Information Technology department at IIIT-A. The system combines AI, Internet of Things (IoT), deep learning, and federated learning to assess crop health accurately. This technological advancement, named CVGG-16, has been peer-reviewed and published in the international Internet of Things journal by Elsevier.

What sets CVGG-16 apart is its approach to data fusion. Instead of focusing only on leaf images, it incorporates data from environmental sensors. These sensors measure soil moisture, temperature, humidity, and weather conditions to provide a fuller understanding of plant health. This allows the AI to identify disease patterns effectively, even in challenging settings like dusty or low-light environments.

In field trials in the agricultural area around Prayagraj, the system showed impressive results, achieving a 97.25% overall disease detection rate. For specific crops, it hit 96.75% accuracy in maize and 93.55% in potatoes, proving its effectiveness across different types of crops.

A key feature of the system is its use of federated learning, thanks to a new Extreme Client Aggregation algorithm. This approach protects data privacy by ensuring that learning occurs on each farm without sending sensitive data to a central server. At the same time, it allows for shared learning across various regions.

Prof. Manish Kumar highlighted that the modular design of the system makes it easy to scale and adapt to India’s many agro-climatic zones. Whether it's in northern India’s maize fields or southern potato farms, the model can be adjusted to fit local growing conditions and crop types.

Understanding that accessibility is vital for widespread use, the team is developing mobile applications that support regional languages. This effort aims to help small and marginal farmers, who often have limited access to agricultural experts, get quick and reliable diagnostics. This could help reduce delays in managing crop diseases.

The CVGG-16 system could transform how farmers manage crop health by enabling nearly instant disease detection on the farm. With early intervention, it has the potential to minimize yield losses, lessen reliance on outside experts, and ultimately enhance food security and farmers’ livelihoods nationwide.