基于近景影像边缘信息点云的精细三维重建

Fine 3D Reconstruction Based on Edge Information Point Cloud of Close-range Image

针对三维激光扫描技术在获取三维信息存在扫描漏洞及边缘信息不足等问题,本文以文物对象的精细三维重建为目的,在近景影像与激光点云精细三维重建中提出一种新的方法。首先利用尺度不变SIFT算法匹配,并设置多重约束逐级剔除误匹配点获得最优匹配点集,采用基础矩阵加严密解解算定向参数,然后采用改进的CANNY算子提取影像边缘信息,利用仿射变换的粗值点和核线几何搜索匹配得到边缘同名点,最后根据定向参数重建边缘点云并与扫描点云进行配准,生成对象的精细点云模型。实验结果表明,采用此方法可生成边缘信息丰富、精细度高的三维信息。

In view of the problems of 3D scanning technology in obtaining three-dimensional information,there are some problems such as scanning loopholes and lack of edge information.In this paper,a new method is proposed for fine 3D reconstruction of close-range image and laser point cloud.Firstly,we use the scale-invariant SIFT algorithm to match and set multiple constraints to remove the mismatched points and then use the improved CANNY operator to extract the edge information of the image.The edge of the corresponding points is obtained by using the rough point of the affine transformation and the geometric search of the peripolar transformation.Finally,according to the orientation the parameters reconstruct the edge point cloud and register with the scanning point cloud to generate the fine point cloud model of the object.The experimental results show that this method can generate three-dimensional information with rich edge information and high fineness.