Omnidirectional 3D Point Cloud Acquisition via Stereo Phase Unwrapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current 3D surface measurement technologies using fringe projection struggle to achieve high-precision, real-time, and 360° omnidirectional point cloud acquisition, especially for dynamic objects, due to limitations in hardware and algorithms, and require complex and expensive auxiliary instruments for registration.
Innovation Solution
A quad-camera fringe projection system employing stereo phase unwrapping and simultaneous localization and mapping (SLAM) technology for real-time high-precision 3D data acquisition, using scale-invariant feature transformation (SIFT) for coarse registration and iterative closest point (ICP) for fine registration, without auxiliary instruments, enabling dual-thread parallel processing for efficient point cloud registration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional FPP system is used to obtain complete 3D model, then multiple viewpoint scanning is required, but measurement time and complexity increase
Solution Approach 1:
The patent transitions from single-viewpoint 3D measurement to multi-viewpoint omnidirectional measurement by adding dimensional coverage. The measurement system captures 3D data from multiple angles simultaneously, transforming the measurement space from a limited viewing angle to a complete 360-degree spherical coordinate system, thereby obtaining the complete 3D model in one measurement.
Solution Approach 2:
The patent creates a measurement system that performs multiple functions: it can measure objects from any viewpoint, handle both static and dynamic objects, and provides complete omnidirectional coverage. The system integrates multiple cameras and projectors to achieve universal measurement capability across all spatial directions, eliminating the need for manual repositioning.
2Measurement precision
If auxiliary instruments are used for point cloud registration, then registration accuracy improves, but device complexity and cost increase
Solution Approach 1:
The patent enables the measurement system to perform self-registration without external auxiliary instruments. The system uses its own multi-viewpoint 3D data to automatically register point clouds from different angles through feature matching and coordinate transformation algorithms, making the registration process intrinsic to the measurement system rather than dependent on external devices.
Solution Approach 2:
The patent replaces mechanical auxiliary registration instruments (such as rotating platforms, robotic arms, and flat mirrors) with computational methods. Instead of using physical devices to establish coordinate relationships, the system uses image processing algorithms and mathematical transformations to achieve automatic point cloud registration, thereby simplifying the hardware structure.
3Device complexity
If real-time 3D registration is performed without auxiliary instruments, then system simplicity improves, but registration precision deteriorates
Solution Approach 1:
The patent performs preliminary feature extraction and matching before final registration. The system pre-processes the 3D data by identifying key features and establishing initial correspondences between point clouds from different viewpoints. This preliminary action provides a solid foundation for subsequent precise registration, ensuring high accuracy without requiring complex auxiliary instruments.
Solution Approach 2:
The patent implements a feedback mechanism in the registration process. The system continuously evaluates the registration quality by comparing feature correspondences and adjusts the transformation parameters iteratively to minimize registration errors. This feedback loop ensures that the automatic registration achieves high precision comparable to instrument-aided methods.
4Productivity
If stripe boundary coding strategy is used for real-time acquisition, then acquisition speed improves, but point cloud sparsity increases
Solution Approach 1:
The patent merges multiple fringe projection patterns and combines data from multiple cameras simultaneously. By projecting multiple phase-shifted fringe patterns and capturing them with multiple cameras from different viewpoints, the system accumulates denser point cloud data while maintaining real-time acquisition capability through parallel processing.
Solution Approach 2:
The patent implements continuous omnidirectional measurement by maintaining uninterrupted fringe projection and data acquisition from all viewpoints. The system continuously captures 3D data as the object moves or is rotated, ensuring that no valuable measurement information is lost and achieving both high density and real-time performance.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables real-time, high-precision, and unconstrained 360° omnidirectional point cloud acquisition, eliminating the need for complex hardware and time-consuming post-processing, allowing arbitrary object rotation for continuous 3D data capture with improved accuracy and efficiency.
Implementation Method 1
fringe projection technology (FPP) has become one of the most popular 3D imaging technologies
Implementation Method 2
stereo phase unwrapping method SPU
Implementation Method 3
four cameras and one projector are kept at a long distance
Data Source
AI summary
Disclosed is a high-precision dynamic real-time 360-degree omnidirectional point cloud acquisition method based on fringe projection. The method comprises: firstly, by means of the fringe projection technology based on a stereoscopic phase unwrapping method, and with the assistance of an adaptive dynamic depth constraint mechanism, acquiring high-precision three-dimensional (3D) data of an object in real time without any additional auxiliary fringe pattern; and then, after a two-dimensional (2D) matching points optimized by the means of corresponding 3D information is rapidly acquired, by means of a two-thread parallel mechanism, carrying out coarse registration based on Simultaneous Localization and Mapping (SLAM) technology and fine registration based on Iterative Closest Point (ICP) technology. By means of the invention, low-cost, high-speed, high-precision, unconstrained and rapid-feedback omnidirectional 3D real-time molding becomes possible, and a new gate is opened into the fields of 360-degree workpiece 3D surface defect detection, rapid reverse forming, etc.