 The Apple Target Recognition Algorithm is one of the core technologies of the Apple Picking Robot. However, most of the existing Apple Detection Algorithms cannot distinguish between the Apples that are occluded by tree branches and occluded by other Apples. To address this issue, a lightweight Apple Target's detection method was proposed for Picking Robot using improved YOLO v5s. This method uses a modified version of YOLO v5s, which has been optimized to improve its performance in detecting occluded Apples. Additionally, the model has been further enhanced by adding a visual attention module, SE, and a bonding fusion mode. The resultant model has shown significant improvements over the original YOLO v5s in terms of accuracy and speed. This article was authored by Bin Yen, Pan Fan, Xiaoyan Lei, and others.