Cloud Image Positioning Using 3D Point Cloud Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image positioning methods relying on pre-deployed positioning tags are complex and inefficient, requiring high deployment complexity and reducing positioning convenience and efficiency.
Innovation Solution
A cloud technology-based positioning method that retrieves and registers to-be-positioned image data with point cloud data from a three-dimensional model database, utilizing point cloud data with location information to determine a pose without relying on pre-deployed tags, and divides the model into sub-regions for efficient data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If image positioning is performed using a positioning tag deployed in a specified region, then positioning can be achieved, but deployment complexity increases and positioning efficiency decreases
Solution Approach 1:
The patent extracts and removes the positioning tag from the system. Instead of requiring a positioning tag to be deployed in the specified region, the invention uses feature point extraction and matching from images alone to achieve positioning, thereby eliminating the complex deployment process while maintaining positioning capability
Solution Approach 2:
The patent creates a virtual three-dimensional model database that copies and stores pre-extracted feature points and their corresponding spatial information from multiple images. This virtual model serves as a reference for positioning without requiring physical positioning tags in the environment, reducing deployment complexity while preserving positioning accuracy
2Reliability
If image positioning is performed using a positioning tag deployed in a specified region, then positioning can be achieved, but positioning efficiency decreases
Solution Approach 1:
The patent performs preliminary actions by pre-extracting feature points from multiple images, establishing correspondences between feature points and spatial coordinates, and building a virtual three-dimensional model database in advance. This preparation work is done offline, so that during actual positioning operations, the system can quickly perform feature matching and retrieve results from the pre-built database, significantly improving positioning efficiency
Solution Approach 2:
The patent segments the positioning process into distinct phases: offline feature extraction and virtual model building, and online feature matching and positioning. By dividing the workflow, the computationally intensive tasks are performed in advance when time is not critical, while the online phase only requires lightweight matching operations, thereby improving real-time positioning efficiency
3Area of stationary object
If the entire three-dimensional model database is used for registration, then comprehensive coverage is achieved, but data processing complexity increases
Solution Approach 1:
The patent segments the three-dimensional model database into multiple local coordinate systems, each corresponding to a specific region or viewpoint. Instead of performing registration against the entire database, the system divides and conquers by processing smaller, manageable segments independently, thereby reducing computational complexity while maintaining comprehensive coverage through the aggregation of multiple local results
4Area of stationary object
If the entire three-dimensional model database is used for registration, then complete spatial coverage is achieved, but processing time increases
Solution Approach 1:
The patent segments the large three-dimensional model database into multiple smaller local coordinate systems. This segmentation allows the registration process to operate on smaller data subsets in parallel, reducing the time required for each individual registration operation while collectively covering the entire spatial area through the combination of multiple local coordinate systems
Solution Approach 2:
The patent applies partial action by performing registration operations on selected local coordinate systems rather than processing the entire three-dimensional model database at once. This selective processing approach reduces processing time by focusing computational resources on relevant portions of the data while still achieving comprehensive spatial coverage through the aggregation of results from multiple local registrations
Data Source
AI summary
A cloud technology-based positioning method and apparatus are disclosed, and relate to the field of computers. The positioning method includes: obtaining to-be-positioned image data; retrieving the to-be-positioned image data from a three-dimensional model database, to obtain first point cloud data having a matched similarity to the to-be-positioned image data; and then performing registration on the to-be-positioned image data based on a point having location information in the first point cloud data, to obtain a first pose corresponding to the to-be-positioned image data. The first point cloud data having the matched similarity to the to-be-positioned image data is retrieved from the three-dimensional model database, to determine, from the entire three-dimensional model database, point cloud data corresponding to a partial region matching the to-be-positioned image data.