GRAPE BERRY DETECTION ALGORITHM BASED ON IMPROVED YOLOv5

To address the problem of low detection accuracy due to the dense distribution and complex background of grape clusters in modern farms, a fast and accurate grape cluster detection method based on improved YOLOv5 is proposed. YOLOv5 was used as the base target detection model, and the feature extraction network was improved by using the coordinate attention mechanism to enhance its feature representation capability, and Bi-FPN was used for efficient fusion of image features to enhance the overall prediction capability of the network. The experimental results show that the detection accuracy of the model can reach 83.1%, which can effectively detect grape clusters in complex environments.