Plant Disease Detection by Image Processing

Abstract

The growing menace of plant diseases to worldwide agriculture requires the creation of effective and precise detection techniques. This thesis investigates the utilization of artificial intelligence (AI) technologies in identifying and diagnosing plant diseases. The objective is to improve the early identification of diseases and decrease the dependence on conventional, labour-intensive approaches. We explore diverse AI models, encompassing machine learning and deep learning methods, to discern and categorize diseases of plants from image data. The research involves the construction of an extensive dataset, the formulation and refinement of models, and the establishment of a rigorous evaluation system to measure their effectiveness. This thesis examines the application of Convolutional Neural Networks (CNNs) in detecting plant diseases. The main focus is on automatically adjusting hyperparameters to enhance the performance of the model. Our study focuses on overcoming the difficulties of getting both high accuracy and efficiency in disease diagnosis. This is done by utilizing advanced CNN architectures and automated algorithms for optimizing hyperparameters.