Application of a Depth Camera for Constructing Complex Three-Dimensional Models in Multiple Scanning Complexes

This work provides a solution to the problem of determining camera displacement in space using 3D data analysis, specifically by searching for reference landmarks that can “stitch together” point clouds, resulting in the construction of complex objects. The solution is achieved geometrically. The reference landmarks are local maxima and convex points identified in different planes of the point cloud. The algorithm, implemented in MATLAB, includes stages of data capture, point cloud segmentation, and search for reference landmarks. By solving this problem, camera displacement can be determined through geometric analysis of optical flow, enabling the localization of stationary objects or serving as an auxiliary algorithm for other orientation determination methods.

I. Introduction

In the modern world, computer vision has become a key component in the field of technical vision, finding applications in various areas ranging from robotics to medical diagnosis. One important aspect of this field is the processing of point clouds obtained using depth cameras, which allows for the reconstruction of a three-dimensional scene structure.