IIIT-Naya Raipur has developed a drone-based crop health forecasting technology that can be used to detect insects and diseases in crops and offer more accurate solutions to address them.
The forecasting solution, designed by faculty from the departments of Electronics and Communication Engineering (ECE) and Computer Engineering (CE), according to the institute, can assist farmers in dealing with crop disease before it causes crop damage.
It will also limit pesticide overuse, which is widespread owing to farmers' lack of reliable information.
The forecasting solution was created in collaboration with the Indira Gandhi Agriculture University in Raipur, which also allows IIIT Raipur to use their agriculture fields for testing purposes.
The project was sponsored by the SEED Division of the Department of Science and Technology of the Government of India. The Department of Science and Technology approved the project in December 2020, and it is expected to be completed this year.
A drone can be deployed in the field and capture images if it detects insects or diseases in the crops as part of the project. The images are then sent directly to the institute's servers, where a convolutional neural network-based image classification model is utilized to detect the disease and insects. A convolutional neural network is a deep learning algorithm that can analyze an image and give weightage to distinct objects in it in order to distinguish one from the other.
A predictive analysis tool is used to analyze the issue and alert the farmers about it along with recommendations how much pesticide should be sprayed and where in the field it should be sprayed. This would reduce excessive use of pesticides among farmers.
According to Anurag Singh, assistant professor of ECE at IIIT-Naya Raipur, "we used images captured by drones and smartphones of crops infected with various sorts of diseases and insects that often impact crops in India to train the AI model."
The drones utilized for the project, according to Shrivishal Tripathi, assistant professor of ECE at IIIT-Naya Raipur, were designed internally and have in-built Internet of Things (IoT) sensors to capture time and position data. The idea is to record information on when the image was captured and the GPS location. “This would allow us to identify the owner of the land and suggest the remedy accordingly," he added.
Monitoring large tracts of land manually, according to Singh, may be challenging. Farmers may conduct inspections more frequently and in less time by using drones. Furthermore, it is difficult to distinguish between insects that look to be the same but are not.
Finance Minister Nirmala Sitharaman urged for a wider application of drones in farming and other sectors through drones-as-a-service in the Union Budget 2022.
