Abstract: ObjectivesThe standardisation of ship's engine room operation is a key part of the ship's safety control, so the crew practical test takes the dismantling of ship's equipment as an important part. In order to improve the electronic and intelligent level of the practical examination, this paper proposes a computer vision-based operation standardization identification method for ship equipment disassembly and assembly, which uses YOLOv8n as the backbonenetwork, and introduces the SA attention mechanism, the GFPN structure and the WIoU loss function to realize operation standardization identification. Methods Firstly, the SA attention mechanism is introduced within the last three C2f modules in the backbone network to improve the model feature extraction capability and training efficiency. Then, the GFPN feature fusion structure is used in the neck network to enhance the multi-scale feature fusion capability of the model. Finally, the original CIoU of Yolo is replaced by the WIoU loss function to improve the model accuracy. Results By testing on the self-built ship equipment disassembly and assembly dataset, it is shown that the proposed method achieves 15% improvement in mAP@0.5, 0.6 improvement in FPS, compared to YOLOv8n, and the disassembly operation of the worker can be accurately identified. Conclusions The improved algorithm has a stronger recognition capability and can be better applied to the task of operation standardization identification tasks for ship equipment disassembly and assembly.