Augmented reality material management system based on post-processing of aero-engine blade code recognition

The efficient management of aero-engine blades plays a vital role in improving material turnover and produc-tivity, while the automatic recognition of blade code provides support for digital production. However, there is a lack of discussion about intelligent recognition, result correction, and effective solutions to material management methods for aero-engine blades. There is still a lot of manual work. Therefore, we developed an augmented reality (AR) material management system based on the post-processing of aero-engine blade code recognition, which not only supports the semi-automatic recognition of blade codes but also provides users with guidance on efficient storage and selection. Firstly, we proposed a line classification method of code Optical Character Recognition (OCR) recognition results to determine the structural relationship of discrete characters. Secondly, we constructed a Bayes error correction model which fuses the noise channel model to achieve the post -processing correction of wrong codes. Lastly, we developed the AR auxiliary guidance module to guide users on warehousing. The results show that the proposed line classification algorithm outperforms the baseline method, achieving a high success rate in terms of line classification. Post-processing error correction is effective in improving the accuracy of OCR results. In addition, AR visualization guidance can lead to a significant improvement in efficiency in material entry and out for aero-engine blades. In conclusion, the system proposed in this paper provides an effective solution to material management for the aero-engine blade.

Automated summary

This study aims to address the issues of intelligent recognition, result correction, and effective solutions in material management for aero-engine blades. By using augmented reality technology and post-processing, a material management system that supports semi-automatic recognition and provides guidance has been proposed. The experimental results demonstrate the system's good performance in line classification and error correction.