In the strawberry harvesting process, to solve the problem of the strawberry picking robot quickly and accurately identifying strawberry fruits, an improved model based on YOLOv8n is proposed for detecting strawberries at different ripeness stages: immature, transition, and mature. First, the DualConv convolutional network structure is used to improve the C2f module of YOLOv8n, enhancing the model's feature extraction ability and computational efficiency. Secondly, the CBAM attention mechanism is added to the output layer of the object detection model to improve its ability to handle occlusions and complex backgrounds. The performance of the improved model (YOLOv8n-DC) achieved a mAP that is 1.2% higher compared to the original YOLOv8n model.