Abstract: For the complicated noise information around the bridge crack，the traditional filtering method can get the denoising effect in some aspects when dealing with such noises，but they often have their own big defects. Therefore，based on the traditional morphological filter，an algorithm based on multi-angle multi-structure element morphological filter was proposed. By combining mathematical morphology，a variety of structural elements were combined to avoid the single effect of single structural element on image processing，which could not only analyze the whole region of the image，but also retain the detail information of the image. The comparison analysis on the experimental data shows that the proposed filtering algorithm has achieved remarkable results.

Key words: bridge engineering, morphology, filter, bridge image, image denoising

摘要： 针对桥梁裂缝周边复杂的噪声信息，传统的滤波方法在处理此类噪声时虽然能在某方面取得去噪效果，但也往往存在着各自较大的缺陷。对此，在 传统的形态学滤波器滤波的基础上，提出了一种基于多角度多结构元素形态学滤波器的算法。通过结合数学形态学，将多种结构元素结合使用，避免了单 一结构元素对图像处理时单一的作用，不仅能对图像全区域进行分析，而且保留了图像的细节信息。实验数据对比分析表明，此滤波算法取得了明显的效 果。

关键词: 桥梁工程, 形态学, 滤波器, 桥梁图像, 图像去噪