A Combined Architecture of Image Processing Techniques and Deep Neural Network for the Classification of Corn Plant Diseases

Abstract

Agriculture is the economic backbone of a number of countries. The agriculture sector considerably contributes to the overall GDP of a growing nation like India. Corn (Zea Mays) is one of the principal crops farmed in the nation. It is a significant food source and a critical raw element for several businesses. Plant diseases are a severe setback that all farmers endure. These illnesses lead to a drop in yield, a serious concern since the gap between demand and supply keeps rising. This research describes an architecture that utilizes Image Processing Techniques and Deep Learning. The suggested architecture employs the Non-Local Means method for noise reduction, Unsupervised Wiener filter, and Entropy to accomplish picture pre-processing. It uses Otsu’s Morphology and Canny Edge detection Method for picture segmentation. A histogram of Oriented Gradients is utilized for feature extraction, and Deep Convolutional Neural Network categorizes the illness.