Point Cloud Fusion Using Image Movement Matrices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional point cloud fusion methods fail in unstructured environments with poor position or inertial navigation signal quality, leading to inaccurate position and orientation determination and unsuccessful three-dimensional reconstruction.
Innovation Solution
A method and apparatus that acquire images associated with point cloud data frames to determine a point cloud movement matrix, enabling fusion even in unstructured scenes by using image characteristics for point cloud data alignment, and switching to inertial navigation or point cloud matching when signal quality improves.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional point cloud fusion methods are used, then position and orientation determination can be performed in structured environments, but fusion fails in unstructured scenes with poor position or inertial navigation signals
Solution Approach 1:
The patent introduces images as an intermediary element to bridge the gap in unstructured environments. When position or inertial navigation signals are poor, the system uses images to extract characteristics and determine movement matrices, which then serve to guide point cloud fusion. This intermediary approach allows fusion to succeed in environments where traditional direct methods fail.
Solution Approach 2:
The system dynamically changes the parameters and methods used for point cloud fusion based on environmental conditions. In structured environments with good signals, it uses traditional position and orientation determination. In unstructured environments with poor signals, it switches to using image characteristics and movement matrices, effectively changing the operational parameters to adapt to different conditions.
2Reliability
If image-based movement matrix determination is used, then point cloud fusion can be achieved in unstructured environments, but system complexity increases
Solution Approach 1:
The system dynamically selects between different fusion approaches based on environmental assessment. It monitors position signal quality, inertial navigation signal quality, and scene structure characteristics, then adaptively switches between traditional signal-based fusion and image-based movement matrix fusion. This dynamic adaptation allows the system to handle complex unstructured environments while maintaining operational simplicity in structured environments.
3Measurement precision
If traditional point cloud characteristic extraction is used, then fusion can be performed in structured scenes, but insufficient characteristics exist in unstructured scenes like tunnels
Solution Approach 1:
The patent uses images as an intermediary source of characteristic information. In unstructured scenes where point clouds lack sufficient features for matching, the system extracts characteristics from associated images instead. These image characteristics are then used to determine movement matrices that guide the point cloud fusion process, effectively compensating for the loss of point cloud characteristic information.
Data Source
AI summary
A method and apparatus for fusing point cloud data, and a computer readable storage medium are provided. Some embodiments of the method can include: acquiring a first image and a second image, the first image and the second image being respectively associated with a first frame of point cloud data and a second frame of point cloud data acquired for a given scene; determining a point cloud movement matrix between the first frame of point cloud data and the second frame of point cloud data on the basis of the first image and the second image; and fusing the first frame of point cloud data with the second frame of point cloud data on the basis of the point cloud movement matrix.