Classification of Radish, Radish Leaf and Potato Leaf Disease Using Deep Learning Algorithm: Study and Accuracy Measurement

Radish and potato are important root vegetables that are extensively grown for their nutritional and economic importance.  However, foliar diseases drastically impair productivity and quality.  Early detection of these diseases is crucial for prompt intervention and successful crop management. This study uses Deep Learning based classification approach to automatically detect diseases in radishes, radish leaves and potato leaves using image data. The dataset consists of 9 distinct classes: healthy radish, healthy radish leaves, healthy potato leaves and several diseased categories. The radish leaf diseases include Alternaria brassicae and flea beetle damage, radish diseases include radish scab and white Mold and potato leaf diseases cover nutrient deficiencies and potato late blight fungus. Multiple convolutional neural network (CNN) models were evaluated for classification performance. The tested models and their respective followed with an accuracy of Xception 99.86%, VGG16 at 99.86%, VGG19 at 98.95%, ResNet50V2 at 99.86% and InceptionV3 at 99.51%. Among the models evaluated the Modified DenseNet121 model proposed demonstrated the highest accuracy achieving a score of 99.93 %. Deep Learning (DL) offer precise and well-timed solutions for disease detection, classification and eradication. The results yield significant potential for enhanced crop management methodologies, facilitating a considerable reduction in economic losses linked to radish and potato diseases. Farmers are able to make informed decisions, minimize crop losses and reduce pesticide use by detecting advancements in disease through the targeted application of agrochemicals. This results in a more sustainable agricultural environment, market stability and healthier crops.