Date of Publication :20th March 2024

Abstract:Sugarcane, a member of the Poaceae family known for its high sucrose content, serves as a valuable resource for producing white sugar, jaggery, and by-products like molasses and bagasse. However, the presence of diseases in sugarcane crops can render them unproductive, necessitating prompt detection. This paper presents a novel deep learning framework designed to determine the health status of sugarcane plants by analyzing their leaves, stems, color, and other characteristics. Investigating sugarcane diseases has been a focal point for research due to the growing demand for this crop and the challenges posed by variable rainfall patterns, making disease identification a delicate task.To address this challenge, the authors propose a solution that involves evaluating the performance of two widely used pre-trained deep learning models, namely VGG19 and ResNet50, in the context of sugarcane disease detection. Sensitivity, error rates, and other performance metrics are computed to assess the effectiveness of these models. Additionally, a graphical comparison of the models based on their sensitivity is presented. These findings are promising, highlighting the potential of deep learning approaches for the early detection of sugarcane diseases. This has significant implications for agriculture, as it can enhance disease management and contribute to the protection and productivity of sugarcane crops.

Reference :