多像位姿估计的全景纹理映射算法

Panoramic Texture Mapping Algorithm Based on Multi-image Pose Estimation

纹理映射技术作为获取具有丰富纹理信息的真彩色点云的有效手段,正以其独特的优势广泛地应用于众多行业领域。研究了一种利用三维激光扫描仪与外置数码相机联合标定解算多张影像位姿并获取全景真彩色点云的方法。其基本思想是利用摄像机与激光扫描仪固有的相对位置姿态,通过对首张影像进行标定得到其位置姿态后,利用摄像机空间旋转的几何特性,根据首张影像的位姿获取其余影像的位姿,继而完成多张影像的纹理映射,获取全景彩色点云。对比目前主流的全景影像纹理映射算法,该算法在精度与效率上均有一定提高。对多种点云数据进行纹理映射实验,结果表明,该方法能够快速准确地获取真三维全景彩色点云,为三维精细化建模提供了数据基础。

Texture mapping technology, which is an effective means to obtain true color point clouds with rich texture information, has been widely used in many fields with its unique advantages. It has a very broad application prospect. We study a method for solving multiple image poses based on joint calibration of 3D laser scanners and external digital cameras, and obtain a method for obtaining panoramic true color point clouds. The basic idea is to use the inherent relative position and posture of the camera and the laser scanner to calibrate the first image obtained by rotating the camera. After obtaining its position and posture, the geometrical characteristics of the cameras space rotation are used to determine the position and orientation of the rest of the images in terms of the pose of first image. Then the texture mapping of multiple images can be accomplished to obtain a panoramic color point cloud. Compared with the current mainstream panoramic image texture mapping algorithm, this algorithm has certain improvement in accuracy and efficiency. By performing texture mapping experiments on a variety of point cloud data, the results show that the method can quickly and accurately obtain true three-dimensional panoramic color point clouds, providing a data foundation for three-dimensional refined modeling.