基于OpenVINO的飞机外表面检查技术研究

Research on Aircraft Surface Inspection Technology Based on OpenVINO

飞机的绕机检查是航线安全检查中最为重要的一个环节,传统的绕机检查经常遇到维修人员经验不足、光线较暗、检查位置过高、机体表面损伤不易发现等缺点;针对航线绕机检查这样的特点提出了一种基于OpenVINO平台的智能绕机检查的方法,将机器视觉的概念引入到航线绕机检查的工作中,针对飞机蒙皮损伤、感温探头堵塞、飞机机身表面有无油渍有较好的效果;在样本的训练中,对现有的模型进行了迁移学习,使得样本的训练量大大减小,能够快速准确的发现飞机外表面检查过程中出现相关问题。

Aircraft winding inspection is the most important link in route safety inspection.Traditional winding inspection often encounters shortcomings such as inexperience of maintenance personnel,low light,high inspection position,and difficulty in finding surface damage on the aircraft.This feature of winding inspection presents a method of intelligent winding inspection based on the OpenVINO platform,which introduces the concept of machine vision into the work of route winding inspection,has a better effect aiming at aircraft skin damage,temperature probe clogging,and the presence or absence of grease on the surface of fuselage.In the sample training,apply transfer learning for the existing model,which greatly reduces the amount of training,and can quickly and accurately find related problems during the inspection of the outer surface of the aircraft.