管道机器人基于折反射视觉的无损检测法

Nondestructive testing method for pipe robot based on the reflection of the reflection

管道在工业和生活中得到广泛应用,在役管道的在线检测对维护管道系统安全畅通运营有重要意义。基于视觉的管道机器人能够实现管道的在线无损检测,在此系统中成像系统承担了很重要的角色。本文建立了基于折反射传感器的圆柱空间分辨率检测系统,系统使用了一个激光环作为照明光源,检测系统包括一个曲面镜和一个摄像机作为折射反射传感器,管道瑕疵通过曲面镜在摄像机的CCD上成像,从而实现管道在线无损检测。为了在保证视场宽度的前提下得到最高的检测分辨率,文中针对针孔摄像机焦距的选择以及曲面镜形状的选择给出建议。

Pipes are widely used in industry and life. The online inspection for pipes in service is very important for to maintain the safety running of pipe systems. Pipe crawler based on catadioptric vision can be used to accomplish the online inspection for pipes. The imaging system plays a major role in the design of vision-based industrial inspection systems. This paper formulates the spatial resolution of a Catadioptric sensor that consists of a camera and a hyperbolic mirror for the pipe inspection. A laser ring is used as a structured light source for illumination. Also we provide a guideline for selecting the optimal configuration setup, subjected to the changes in the focal length of the perspective pin-hole camera and also the shape of the hyperbolic mirror, that yields highest inspection resolution.