YOLOv8-DM: A Study on an Automatic Logbook Generation Method for Airborne Equipment

In the generation process of airborne equipment logbooks, data entry and QR code pasting are currently performed manually, resulting in a low level of automation. To address this issue, this study proposes an automatic logbook generation method based on YOLOv8-DM, utilizing a specific scenario for producing logbooks of airborne equipment as the research background. Firstly, representative sample images are collected to construct a dataset, and the YOLOv8-DM model is used for training to detect logbook QR codes and extract their embedded data. Secondly, an electronic logbook file is automatically generated based on the extracted QR code information and logbook index data, forming a complete logbook database. Finally, YOLOv8-DM is further employed to identify and remove defective QR codes on printed logbook pages, thereby improving the overall production quality. Experimental results show that the proposed model achieves a mean Average Precision (mAP₅₀) of 98.9% for QR code detection and 98.6% for QR code defect detection, exceeding other models by 0.6–8 percentage points. The logbook generation speed reaches 300 books per hour, representing a 14-fold increase compared to manual production. The defect rate is reduced by 98.7%, and the overall production cost decreases by approximately 5, 000 CNY per set (40 books per set). Therefore, the proposed method is well-suited for automatic generation of airborne equipment logbooks, offering significant improvements in both production speed and quality, along with substantial economic benefits. It can also be applied to other fields that require logbook generation.