IIT Mandi Researchers Develop AI based Model to Detect Disease in Potato Crops Using Photos of Leaves

Researchers at IIT Mandi together with the Central Potato Research Institute (CPRI) in Shimla, have developed a computational model for automated disease detection in potato crops using photographs of its leaves.

What is Potato Blight Disease?

Blight is a common disease of the potato plant, which develops as uneven light-green lesions near the tip and the margins of the leaf and then proliferates into large brown to purplish-black necrotic patches that eventually results in rotting of the plant. Blight could destroy the entire crop within a week under conducive conditions, if left undetected and unchecked.

How App detects Blight?

"This is a computational-model for automated disease detection in potato crops", said Srikant Srinivasan, an associate professor at the School of Computing and Electrical Engineering, and CPRI colleagues under a project funded by the Department of Biotechnology.

This model is built using an AI-tool called mask region-based convolutional neural network architecture and it can accurately identify the diseased portions of the leaf from the images, amid a complex background of plant and soil matter.

This entire process is complex and often impractical, especially for remote areas, as it needs the expertise of a horticultural specialist who may not be physically accessible.

“In India, as with most developing countries, the detection and identification of blight are performed manually by trained personnel who scout the field and visually inspect potato foliage,” explained Srinivasan.

“Automated disease detection can help in this regard and given the extensive proliferation of the mobile phones across the country, the smart-phone could be a useful tool in this regard,” said Joe Johnson, IIT Mandi research scholar.

The assimilation of high quality HD cameras, better computing power and communication avenues offered by today's smart-phones can deliver a promising platform for automated disease detection in crops.

To build a robust model, healthy & diseased leaf data were gathered from the fields across Punjab, Uttar Pradesh and Himachal Pradesh, the scientists added.

“Analysis of the detection performance indicates an overall precision of 98 per cent on leaf images in field environments,” said Srinivasan.

Following the outcomes of experiment, the team is sizing down the model to a few tens of megabytes so that it can be hosted on a smart-phone as an application so that when the farmer will take picture of the leaf which looks unhealthy, this application will confirm in real-time whether the leaf is infected or not.

With the real- time information, the farmer would know exactly about the good time to spray the field, saving his produce and minimizing costs associated with unnecessary use of fungicides.

Besides, Srinivasan and Johnson, IIT Mandi faculty Shyam K Masakapalli, research student Geetanjali Sharma and their CPRI counterparts Vijay Kumar Dua, Sanjeev Sharma and Jagdev Sharma participated in the research recently published in the journal ‘Plant Phenomics’.

Source- The Hindu Business Line