| Abstract: | Using image processing techniques, several forms of study can be conducted in the domain of Image of rose leaf disease classification. However, Image of rose leaf disease detection is still a problem for people who do not know about rose leaf diseases. Image-based rose leaf disease detection using deep learning involves training a deep learning model to analyze images of rose leaf and identify signs of diseases such as bacteria, viruses, and fungi. This process can help in the early detection and management of plant diseases, ultimately contributing to improved agricultural productivity and the economy. Deep learning algorithms are trained using a large dataset of images showing healthy and diseased rose leaf. The model learns to recognize patterns and features associated with different diseases, enabling it to accurately classify new images. To classify whether the image is Fresh, Black spot, or Downy mildew we used three classifiers, such as Support Vector Machine (SVM), K-Nearest Neighbor Classifier (KNN), and convolutional neural network (CNN). The datasets are gathered from the Ethio Agri CEEFT PLC Holeta Flower Farm, which is in the Oromia region, Ethiopia. The data is preprocessed through data collection, cleaning, augmentation, image preprocessing, dataset splitting, and data normalization. Feature extraction is performed using the automatic feature extraction capabilities of convolutional layers in CNNs. The total data set used for the experiment is 4342 Rose Leaf Images. The data is split into train and test data sets such that, 20% of the data set is used for testing the model's performance, and 80% for training machine learning as well as deep learning and creating disease detection models from rose leaf images. Experimental results shows that the model created by SVM, KNN, and CNN registers an accuracy of 80.32%,71.23%, and 98% respectively. The model created by CNN therefore outperforms the other classification algorithms. To effectively train the deep learning model, this approach requires a vast and diverse dataset, which is one of its main weaknesses and limitations. Furthermore, it might be difficult to record all possible combinations of environmental variables and disease symptoms due to the reliance on image-based data and further research needs to be done to combine image based with text based so as to come up with a generic model for rose leaf disease detection. |
