Research on Mango Ripeness Based on Yolo

Mangoes are a crucial agricultural crop in southern China, especially in the Guangxi region, where their production significantly contributes to the local economy. However, challenges in ripeness identification lead to quality control issues in storage, transportation, and sales. This research proposes a detection method for mango ripeness based on an improved YOLOv5 model. Traditional fruit ripeness detection methods have certain limitations in terms of accuracy and efficiency. To improve detection accuracy, this research replaces the backbone network of the YOLOv5 model with ResNet50 and introduces the Coordinate Attention (CA) mechanism to enhance the models feature extraction ability and the accuracy of ripeness judgment Experimental results show that the improved model demonstrates higher precision and robustness in detecting mangoes at different stages of ripeness. The research results indicate that the improved YOLOv5 model can effectively identify and distinguish mango ripeness, providing technical support for fruit quality management and harvesting in the agricultural field.