A Convolutional Neural Network-based Approach for Lettuce Leaf Disease Classification

Abstract

Digital image processing can automatically detect a plant's health condition, reducing the need for traditional manual methods like visual inspection and tactile assessment. This research used a classification model based on Convolutional Neural Network (CNN) using TensorFlow to detect the condition of lettuce leaves. The dataset of this model was divided into 80% training set, 10% testing set and 10% validation set. The health condition of the leaves was classified into four categories: Healthy, Septoria Blight, Powdery Mildew and Bacterial Spot. The results of this study showed that the proposed model achieved high accuracy, with a training accuracy of 98.50%, testing accuracy of 87.50% and validation accuracy of 97.50%. These findings indicate that the proposed approach has the potential to significantly improve the detection of lettuce leaf diseases and enhance the quality of plant monitoring and management practices. The integration of this technology in the agriculture industry could revolutionize the way plants are monitored and managed, leading to more efficient and sustainable farming practices.