A Study on Deep Learning Methods to Identify the Infected Regions from Papaya Fruit Images

Fruits account for a significant share of world production across many countries. Many countries desire to maximize fruit production, however, the quality and quantity of the fruits are affected by several diseases that occur during pre- harvesting and post-harvesting of the fruits. To increase the production of fruit with high quality, several types of research have been developed for early detection of fruit diseases with Deep Learning (DL), Machine Learning (ML), and Transfer Learning (TL) techniques from fruit images. However, there exist several limitations along with the advantages of those techniques. A comprehensive review of the existing ML and DL approaches for the classification of fruit disease at the early stage is presented in this study. This research primarily focuses on discussing various fruit diseases and presents an effective way to considerably detect the diseases from papaya fruit images. Various ML and DL models to detect diseases in plants and fruits are discussed in this research. A deep survey has been conducted on conventional methodologies, followed by their purpose, advantages, and limitations. Based on the findings, a suitable framework for future studies to identify disease-infected regions in images of fruits and plants is suggested.