Farm monitoring and disease prediction by classification based on deep learning architectures in sustainable agriculture

Agriculture is necessary for all human activities to survive. Overpopulation and resource competitiveness are major challenges that threaten the planet's food security. Smart farming as well as precision agriculture ad-vancements provide critical tools for addressing agricultural sustainability concerns and addressing the ever-increasing complexity of difficulties in agricultural production systems. This research proposed novel tech-nique in agricultural farm monitoring and crop disease prediction using deep learning architectures. Here the monitored data has been collected based on IoT module along with the historical data of cultivation farm image data. This data has been processed for removal of noise removal and image resizing. The features of processed data has been extracted using deep attention layer based convolutional learning (DAL_CL) in which the features of data has been extracted. This extracted data has been classified using recursive architecture based on neural networks (RNN). The suggested system may use data categorization and deep learning to exploit obtained data and anticipate when a plant will (or will not) get a disease with a high degree of precision, with ultimate goal of making agriculture more sustainable.Experimental results shows the accuracy of 96%, precision of 89%, speci-ficity of 89%, F-1 score of 75% and AUC of 66%.

Automated summary

Agriculture is necessary for human survival, but overpopulation and resource competition pose major challenges to food security. This research proposes a new technique using deep learning for agricultural monitoring and crop disease prediction, which can accurately anticipate the impact of diseases on plants through classification and deep learning.