Advanced Techniques for Spinach Disease Identification: CNN and Random Forest Fusion

Spinach, an essential constituent of the agricultural domain, confronts substantial dangers emanating from several illnesses that possess the potential to profoundly affect both crop productivity and quality. The present research aims to tackle this issue by introducing an innovative methodology that integrates Convolutional Neural Networks (CNN) and Random Forests to achieve precise identification of diseases in spinach leaves. This paper introduces a unique methodology for the precise identification of diseases in spinach leaves, using a fusion of Convolutional Neural Networks (CNN) and Random Forest. Preprocessing approaches serve to improve the process of feature extraction, while the convolutional neural network (CNN) component is responsible for extracting complex visual information. The classification process is enhanced by using a Random Forest classifier consisting of 20 decision trees, each having a depth of 18. The evaluation conducted on a dataset of 2543 spinach leaf pictures reveals a notable level of accuracy, as shown by F1-scores ranging from 76.82% to 84.53%. The F1-score, which is calculated as the weighted average of accuracy and recall, shows a balanced performance with a value of 81.52%. This study provides agricultural practitioners with a tool that enables rapid and accurate disease diagnosis, hence improving spinach production practices.