Abstract: In order to deal with the problem of false information and edge breakpoint about binary image of asphalt pavement crack image segmentation, a new method of discontinuity crack block linking was developed based on autonomous edge growing. This method removed false information according to the characteristics of circular noise and linearly cracks, and filled interstice by using method of region filling, thereby accomplished linking of discontinuity crack block based on autonomous edge growing. The experimental results show that the algorithm has excellent performance in edge linking of discontinuity crack block in different cases, and the noises can be removed at the same time, which is in favor of subsequent image processing such as image measurement and evaluating.

Key words: isolated noise, region filling, autonomous growing, crack block linking, asphalt pavement

摘要： 针对沥青路面裂缝图像分割成二值图像后，存在虚假信息和边缘断点的问题，提出一种基于主动生长的断裂裂缝块的连接方法。该方法依据噪声呈孤立圈状，裂缝呈线性的特征去除虚假噪声信息；并采用区域填充法将裂缝轮廓内部空隙填充，使裂缝呈现块状；然后采用主动生长法进行断裂裂缝块的连接。实验表明，该算法在去除噪声的情况下，较好实现了断裂裂缝块的连接，有利于后续的图像测量、评估等处理。

关键词: 孤立噪声, 区域填充, 主动生长, 裂缝块连接, 沥青路面