Smart Diagnosis of Rose Leaf Diseases using Advanced Image Processing and Machine Learning

Abstract

Roses, as one of the most widely cultivated ornamental flowers, hold significant aesthetic and economic value globally. However, their susceptibility to various foliar diseases presents major challenges to sustainable floriculture, impacting both crop quality and growers’ livelihoods. Conventional disease detection methods are often labor-intensive, time-consuming, and dependent on expert intervention. To address these limitations, this study introduces a smart and automated framework for the early detection and classification of rose leaf diseases using advanced image processing and machine learning techniques. A comprehensive dataset comprising images of both healthy and diseased rose leaves was curated, covering various disease types and severity levels. Image preprocessing steps—such as contrast enhancement, segmentation, and noise reduction—were implemented to optimize feature extraction. The system utilizes a hybrid classification approach by integrating Convolutional Neural Networks (CNN), Support Vector Machines (SVM), and K-Nearest Neighbors (KNN) to enhance accuracy, generalizability, and robustness across diverse visual patterns.

Experimental results demonstrate high classification performance, with the hybrid model outperforming individual classifiers in both accuracy and computational efficiency. The proposed system enables real-time, scalable, and precise rose disease identification, making it a valuable tool for floriculturists, researchers, and precision agriculture practitioners. In future research, the model can be extended to support a wider variety of ornamental plants and additional plant parts such as stems and petals