一种桥梁拉索表面缺陷图像的背景分割方法研究

A method of background segmentation for surface defect of bridge cables

随着拉索在桥梁中的广泛应用,其表面缺陷检测成为桥梁年度检测的重要内容;然而桥梁拉索表面缺陷图像检测缺乏自动化处理技术,文中提出一种利用掩模对桥梁拉索表面缺陷图像背景进行分割的方法。首先选取一幅图像,将图像从RGB颜色空间转换到Lab颜色空间,在阈值分割后,利用数学形态学去噪获得背景掩模;然后基于该掩模对所有图像进行掩模处理,完成图像的背景分割。实验结果表明,该方法背景分割准确、分割效率高,解决桥梁拉索表面缺陷图像自动化处理中的图像分割问题。

With the wide application of cables in bridges,detecting the bridges＇ surface defect becomes an important content of bridge annual inspection.Because of the lack of automatic processing technology in the surface defect image detection,the paper proposes a method that segmented the image background of bridge cable surface by mask.Firstly,it selects an image,and converts it from the RGB color space to the Lab color space.Then,after threshold segmentation,it obtains the background maskby mathematical morphology denoising.Finally,the paper conducts the mask processing on all images based on the background mask,to complete the image background segmentation.The experimental result proves that the background segmentation in the proposed method is correct,as well as in high efficiency,which solves the image segmentation in the automatic processing of the surface defects of the bridge cables.