| Abstrakt: |
One of the most important crops in the world, sugarcane, is constantly in danger from a number of diseases that can have a negative impact on the crop's productivity and quality. The cultivation of sugarcane is significantly hampered by one of these, Sugarcane Leaf Scald (SLSD). Effective SLSD management depends on early and precise illness identification. Due to their capacity to analyze enormous datasets and identify nuanced patterns, deep learning algorithms have recently come to be recognized as potent instruments for disease diagnosis in agricultural contexts. The diagnosis of diseases caused by Sugarcane Leaf Scald is presented in this research utilizing cutting-edge deep learning methods. We start by gathering a sizable dataset of photos of sugarcane leaves that spans several disease progression phases and includes both healthy and infected samples. Utilizing a Convolutional Neural Network (CNN) architecture, discriminative features are automatically discovered and extracted from these photos. Using methods like data augmentation and transfer learning, the deep learning model is trained on this dataset to improve its performance. The suggested deep learning-based approach's results show that it can correctly categorize sugarcane leaves as either healthy or SLSD-infected. The model may also determine the severity of the condition, which helps develop specialized disease management plans. High accuracy, precision, and recall rates are among the encouraging performance indicators the suggested system displays, indicating its potential as a dependable tool for early SLSD detection in sugarcane fields. The investigation shows how well deep learning algorithms work in diagnosing SLSD, which advances the field of agricultural disease management. The suggested method provides a non-invasive, affordable, and scalable answer for tracking and controlling sugarcane infections, thereby boosting crop output and the long-term viability of sugarcane farming. [ABSTRACT FROM AUTHOR] |
