Advanced Sunflower Leaf Disease Detection Using CNN-SVM Hybrid Models

An extensive investigation of effectiveness metrics derived from a classification system intended to identify several common disorders in sunflower leaves is presented in the abstract. The model's accuracy, recall, or F1-score are important performance metrics that indicate how well it can categorize various illness categories. Outstanding results, which span from 96.65% to 98.75% in all disease categories, demonstrate how robust and dependable the model is at misidentifying both damaged and healthy sunflower leaves. The abstract demonstrates how well the model performs, with a total precision of around 96.75% for a data set comprising 9,695 photos. The model's overall effectiveness and the degree to which it correctly recalls every positive instance and recognizes positive occurrences are both shown through the F1-score, which corresponds to the harmonic average of these two measures. Every type of sickness in the dataset has support values according to the number of instances in each category. The representative dataset was trained to remove all preexisting trends, and class distribution also plays a major factor in mitigating bias in learning new concepts. Deep learning model level d at the staying different illness, just various levels of nourishment health, or types of medicines than fighter planes, support rates of these diseases will make it a delightful close whistle and the quotient. As a disease management tool in sunflower farming, the classification model showed an overall precision of 99% in identifying illnesses like blight, with high accuracy. Timely, focused action has armed farmers to help prevent diseases and keep crops free from disease now. The model's excellent performance metrics in this evaluation point to its effectiveness at diagnosing infections; they're also an opportunity to better understand both agricultural output and sustainable methods for growing sunflowers.