Decoding Disease Intensity: Predicting Pythium Severity with CNN-SVM Hybrid Models

A precise and prompt evaluation is required to establish efficient disease control measures because Pythium disease poses a serious hazard to agricultural products. The severity levels of the Pythium disease in crops are divided into five classes in this study report, with 1-20%, 20-40%, 40-60%, 60-80%, and 80-100% being the most severe. To extract useful information from high-resolution photos of infected crop samples, the CNN model consists of three convolutional layers, three max-pooling layers, and one fully connected layer. Precision ranges from 95.22% to 98.31%, recall ranges from 95.49% to 96.79%, and F1-Score ranges from 95.74% to 97.48% are excellent results for each severity level class, according to their evaluation of the model's performance using these measures. The model's capacity for producing accurate predictions is further confirmed by its total accuracy of 96.99%. By allowing fast disease severity assessment and directing educated decision-making for efficient disease control, their CNN-based strategy has significant potential to revolutionize agricultural practices. Even though their work represents a substantial improvement in plant pathology, the model still has to be validated using a variety of datasets and real-world situations to determine its robustness and applicability. Advanced deep-learning methods for disease monitoring can help to improve global food security and enable sustainable agricultural production.