Coffee Leaf Disease Detection Using Transfer Learning

Abstract

Recognizing disease in coffee leaves is an important aspect of providing a better quality of coffee across the world. Economies of many countries in the world depend upon the export of coffee and if we fail to recognize the disease in the coffee plant it will have a negative impact on them. The objective of this paper is to propose models for recognizing disease in coffee leaf plants. For achieving our objective we used various pre-trained models. We used transfer learning approach to identify coffee leaf detection. There were several models that achieved great results on both training and testing data. However, the best-achieving model was VGG19 due to less memory utilized and less time required for execution.