Method and device for stitching point cloud data based on panoramic images

Abstract

The invention discloses a point cloud data splicing method based on panoramic images, comprising: synchronously collecting a sequence of group images and point cloud data of a scene, the images of adjacent sequences have a preset overlapping area; according to the preset Stitching a sequence of images in the overlapping area to obtain a panoramic image of the scene; projecting a sequence of point cloud data into the panoramic image according to a corresponding sequence; The set of point cloud data is spliced ​​to obtain a point cloud data set expressing the whole of the scene. The point cloud data splicing method and device based on the panoramic image of the present invention realizes splicing of the scene point cloud data based on the panoramic image of the scene by synchronously acquiring the images of the scene. Arrange marker points in the scene, and can reduce the difficulty of measurement operation, improve measurement efficiency and flexibility.

Description

technical field [0001] The invention relates to the technical field of visual image processing, in particular to a method and device for splicing point cloud data based on panoramic images. Background technique [0002] Using 3D measurement technology to obtain 3D spatial information and reconstruct the scene in 3D has broad application prospects in virtual reality, cultural relic protection, archaeological research, industrial design, deformation monitoring, etc. [0003] Three-dimensional laser scanning technology is a means of acquiring more three-dimensional spatial information in recent years. It is based on the principle of active measurement and has the characteristics of fast, accurate and non-contact acquisition of three-dimensional information of the measured object or scene, so it has been widely used. Due to the limitation of measurement equipment and environment, each scan can only obtain partial data of the scanned scene. To obtain the complete 3D point cloud d...

AI Technical Summary

Problems solved by technology

Among them, the accurate splicing of point cloud data requires points with the same name and obvious characteristics in adjacent frames. In practical applications, it is usually achieved by arranging feature points in the measured object or scene before measurement. This process not only greatly reduces the data Acquisition efficiency and flexibility, but also increases the difficulty of operation, it is difficult for untrained non-professionals to use it proficiently in a short period of time

[0004] Therefore, in order to realize the accurate splicing of multi-site point cloud data in the current 3D measurement method, it is necessary to arrange several marker points in the measured scene before scanning, which makes the measurement operation complicated, difficult and inefficient.

Images