A Tea Buds Detection Algorithm Based on Improved YOLO11

The detection and localization of tea buds are a critical part of the intelligent tea harvesting process, and using deep learning-based object detection technology to detect tea buds is of great significance for intelligent tea harvesting. This paper proposes an improved tea bud detection algorithm based on YOLO11, which uses the CBAM (Convolutional Block Attention Module) attention mechanism to automatically adjust the focus on different areas, enhancing the model's ability to capture targets against complex backgrounds and improving the model's accuracy. Experimental results on the tea bud dataset show that the improved algorithm has an average precision mAP50 that is 1.4% higher than the original YOLO11, and the mAP50:95 increased by 2.2%, and there is also a significant improvement in detection accuracy compared to other YOLO detection algorithms, providing a technical foundation for the application of intelligent tea bud harvesting equipment.