Automation in Agriculture and Smart Farming Techniques using Deep Learning

Agriculture is considered to be a field of great importance and with a serious economic impact in all successful countries. Due to the substantial increase in world population, it has become a relevant concern to be able to meet people's daily dietary needs. Henceforth, it has become inevitable to make a transition to smart agricultural techniques to achieve the set food security goals. In recent times, several deep learning methods, such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), have been vigorously studied, applied, and researched in different fields, including farming and agriculture. In this project, we aim at analyzing existing research on deep learning techniques in smart farming and agriculture and propose solutions for different aspects of farming using various deep learning architectures. Furthermore, we studied the farming parameters such as weather reports, plant irrigation information, pests that affect common crops, germination periods of the flowers/seeds, disease/anomaly detection in their leaves, etc., and proposed modular solutions for each of the respective areas of smart farming. Additionally, we also compared relevant studies regarding farming and focused agricultural methods, problems being faced, the method for collecting data being used, and the deep learning model suggested.