| description abstract | Quantifying the features of concrete cracking is of significance for assessing the bearing capacity and durability of concrete structures. However, there is a lack of efficient and reliable observation and description of the crack features, particularly for very early-age concrete. In this study, a high-precision, noncontact laser scanner is used to dynamically obtain the 3D point cloud data of the cracks in concrete starting from casting until the age of 24 h. A crack point extraction method based on the local density of the scanned points is proposed to recognize the crack points, which are then connected according to the nearest neighbor principle to form crack elements. By setting the angle and the distance thresholds, the crack elements are clustered and expanded into independent cracks, and further, the crack sets. The calculation methods for both the length and the width of the independent cracks as well as the long and the short axis of the crack set are proposed from the scanned 3D data, and thus, the cracks are quantified from one and two dimensions in terms of the independent crack and the crack set, respectively. The reliability of the crack extraction method based on the measured 3D point cloud data is compared to and verified by the traditional crack viewer method. The results of this study can provide a strong algorithm support for intelligently detecting concrete cracks and for predicting crack development. | |
