Plant Disease Detection Over Multiple Datasets Using AlexNet

Abstract

Plants diseases are responsible for huge loss of crop yield. Manual inspection of plant disease is a time taken and inefficient process. Image processing and machine learning-based approaches have been offered as a solution for creating such automated plant disease detection systems. Plant diseases leads to change in color and texture of leave, this property is used for developing plant disease detection systems. Deep learning models such as Visual Geometry Group (VGG) and ResNET are extensively used in this field. However, most of these models are not scalable as they are either focused on disease classification on a particular crop or dataset. The focus of this study is to showcase a new method for identifying leaf diseases. AlexNet is used in the system's development, and it is trained and verified using data from many sources. Results indicate improved performance as compared to previously published works.