A review of fruit ripeness recognition methods based on deep learning
Xiaoyu Zhou, Xing Hu, Jihong Sun
Abstract
Computer vision holds immense potential in solving agricultural issues. Traditional fruit maturity assessment methods are labor-intensive and costly, relying on manual sampling and chemical analysis. Visual technologies offer accurate and rapid alternatives for agricultural automation. This paper systematically reviews 80 references on deep learning-based fruit maturity recognition, categorizing the latest methods. It examines traditional machine learning techniques and their limitations, highlighting the advantages of deep learning. Recent advancements in image feature extraction are reviewed, and the performance of various models is compared. The paper also explores current limitations and future trends, providing insights to guide future research in this field..