Detection of Coffee Leaf Diseases using DenseNet-264 and SLIC Segmentation

Coffee leaf disease (CLD) identification and management is critical to the coffee industry since it can lead to reduced yields and poor-quality coffee. We offer a new method for recognizing CLD in this work that makes use of the Simple Linear Iterative Clustering (SLIC) segmentation algorithm and the Densenet-264 deep learning model. Our method involves SLIC segmentation to extract diseased areas from a collection of coffee leaf images, which are then used to train a Densenet-264 model to classify images as infected or healthy. Experiments findings demonstrate that our suggested strategy outperforms existing methods for CLD identification with an accuracy of 97.17% in correlation with Resnet-50 with accuracy 96.54%. This innovation has significant ramifications for the coffee industry because it gives a more accurate and efficient method of measuring.