Analysis of Deep Learning Algorithms for Intelligent Plant Disease Identification

In developing countries like India, Rice is the fundamental and most important food crop. Also, rice is an essential food crop for low and lower middle income countries. The disease in the rice leaf affects rice production which results in severe poverty across the country. The proposed work aims to design and build a real-time system to detect healthy and diseased rice leaves. The reliability analysis of three distinct deep learning algorithms, namely Linear kernel Deep Belief Networks (LDBN) and Polynomial kernel Deep Belief Networks (PDBN), and Convolutional Neural Networks (CNN) was also performed. The findings revealed that the implementation of the Convolutional Neural Networks (CNN) has very high accuracy when compared to the Deep Belief Networks (DBN) with linear and polynomial kernels. As the proposed hardware can detect rice leaf diseases, this work seems to be of high social significance.