基于图像处理的桥梁缆索检测系统研制

Development of Cable Inspection System Based on Image Processing

研制了一套基于图像采集和处理的桥梁缆索健康检测系统,包括检测机器人、上位计算机、相机及图像采集系统。检测机器人采用单电机驱动,能自动适应不同直径的缆索。机器人通过对缆索进行全方位照相,并对缆索图像进行自动识别,来判断缆索的病害。对采集的缆索图像,经过图像去噪、图像增强、图像分割、特征提取,最后以图像的灰度直方图中的灰度分布和能量分布等为特征进行病害识别。

A stayed cable health inspection system based on the image acquisition and processing was developed, which con- sisted of robot, host computer, cameras and image acquisition system. The robot was driven with single motor and could automatically climb cables of various diameters. Omni-directional pictures of the cables were taken by the robot, and the defects and mars were identified through the automatical recognition of the cable image. The steps of image recognition including image de-noising, image enhancement, image segmentation and feature extraction, and the defects and mars were identified from the features of the grayscale distributions and energy distributions in grey level histogram.