结合路面深度影像梯度方向直方图和分水岭算法的裂缝检测

发布时间：2018-01-13 19:41

  本文关键词：结合路面深度影像梯度方向直方图和分水岭算法的裂缝检测　出处：《华中师范大学学报(自然科学版)》2017年05期 　论文类型：期刊论文

  更多相关文章： 裂缝检测 深度影像 梯度方向直方图 分水岭算法

【摘要】：裂缝检测对于道路维护和管理具有重要作用.由于深度影像对路面油污、阴影等因素不敏感,近些年来基于深度影像的检测方法已成为路面裂缝检测新的研究方向之一.传统的激光扫描线方法没有顾及裂缝在整个空间分布的变异性、各向异性和全局性特征,无法有效检测横向、块状、网状等裂缝.针对以往算法的不足,提出一种结合梯度方向直方图和分水岭算法的路面裂缝检测方法.首先,通过梯度方向直方图算法提取路面深度影像的裂缝边缘强度和方向;然后,利用裂缝边缘方向改进传统分水岭算法,最终提取裂缝目标.实验结果表明,该方法不仅能够准确检测多种类型的裂缝目标,而且能识别裂缝破损程度.
[Abstract]:Crack detection plays an important role in road maintenance and management, because depth image is not sensitive to road oil, shadow and other factors. In recent years, the detection method based on depth image has become one of the new research directions of pavement crack detection. The traditional laser scanning line method does not take into account the variability of crack distribution in the whole space. Anisotropic and global characteristics, can not effectively detect transverse, block, mesh and other cracks. In view of the shortcomings of previous algorithms. A method of pavement crack detection based on gradient direction histogram and watershed algorithm is proposed. Firstly, the edge strength and direction of pavement depth image are extracted by gradient direction histogram algorithm. The experimental results show that the proposed method can not only accurately detect various types of crack targets, but also identify the degree of fracture damage.
【作者单位】： 湖北工业大学计算机学院;
【基金】：湖北省教育厅资助基金项目(2014277)
【分类号】：TP391.41;U418.66
【正文快照】： 裂缝是路面最常见的病害之一,自动检测裂缝对于公路检测与养护管理具有重要意义.目前,裂缝检测算法主要以路面二维图像的裂缝灰度特征及形态特征作为判别裂缝的准则;由于受到图像采集系统硬件条件的限制及外界光照影响,伴随路面油污、阴影、轮胎痕迹、随机噪声等因素带来的干，

本文编号：1420294