In order to solve the problems that the narrow space of aircraft structure is hard to be entered to check the tiny cracks quickly, leading to some major safety risks, a bionic robot platform suitable for narrow space movement and the matching image acquisition and crack recognition technology is carried out. The research of bionic robot platform involves bionic robot mechanism, movement behavior and gait planning, environment perception and obstacle avoidance. Image crack recognition research specifically includes a large number of image samples digital enhancement, image preprocessing, image crack algorithm based on deep learning and other studies. Finally, aircraft box section test sample with narrow space characteristics is adopted to verify this technology by presetting cracks, debugging of image acquisition environment, combined with robot path planning, crack algorithm optimization research, etc. The crack length and location in a narrow space are identified successfully.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Research on Crack Image Detection Technology of Aircraft Structure in Narrow Space

  • Wenjuan Wang,
  • Mengjie Zhang,
  • Yuanbo Liu,
  • Jingfeng Xue,
  • Jie Sun,
  • Haosheng Pang

摘要

In order to solve the problems that the narrow space of aircraft structure is hard to be entered to check the tiny cracks quickly, leading to some major safety risks, a bionic robot platform suitable for narrow space movement and the matching image acquisition and crack recognition technology is carried out. The research of bionic robot platform involves bionic robot mechanism, movement behavior and gait planning, environment perception and obstacle avoidance. Image crack recognition research specifically includes a large number of image samples digital enhancement, image preprocessing, image crack algorithm based on deep learning and other studies. Finally, aircraft box section test sample with narrow space characteristics is adopted to verify this technology by presetting cracks, debugging of image acquisition environment, combined with robot path planning, crack algorithm optimization research, etc. The crack length and location in a narrow space are identified successfully.