• Version
• Download 7
• File Size 339.75 KB
• File Count 1
• Create Date 2 May 2023
• Last Updated 2 May 2023

Smart Farming- Revolutionizing farming with hybrid technology for disease detection, crop protection and high yield production

DR. ANKIT GARG1 , ADITYA MITTAL2 , KRISHNA SAI SRI AASRITHA LINGAM2 , MADHAV AGARWAL2 , AND

UDIT VERMA2

1Assistant Professor, Electronics and Communications Engineering Division, NETAJI SUBHAS UNIVERSITY OF TECHNOLOGY, NEW DELHI-110078

2Student, Electronics and Communications Engineering Division, NETAJI SUBHAS UNIVERSITY OF TECHNOLOGY, NEW DELHI-110078 Compiled April 26, 2023

Agriculture is a critical sector of the economy and plays a significant role in the development of any country. Recently, the sector has witnessed a transformational shift driven by the need for food security, increased productivity, and sustainability. However, the ever-growing population, unpredictable weather patterns, and limited resources pose significant challenges to agriculture. With the advent of new technologies such as IoT, big data, and AI, there is an opportunity to address these challenges and enhance the efficiency of agricultural practices, leading to sustainable agriculture and food security. To this end, this project aims to develop an end-to- end web application that offers various services critical to the agriculture sector, including crop recommendation, crop price estimation, plant disease detection, crop health prediction, and crop yield prediction. The scope of the models is drastically improved, such that the model is trained to be used all over India, rather than working on one crop or region. The application leverages the use of pertinent, dense, and representative data, which is meticulously scoured and scraped from relevant sources such as the official Government of India websites. The proposed solution addresses the limitations of current agricultural practices by providing real-time information to farmers, enabling them to make informed decisions, leading to better yields and improved profits. Finally, algorithm optimization is performed, leading to significant accuracy improvements, and a minimum accuracy of 93% for each of the models is achieved.