3D Sensor System Multi-Line Pattern Projection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional three-dimensional scanning using the triangulation method is limited by the need to project a single linear pattern, which restricts point-determining speed and efficiency, making it unsuitable for high-speed scanning applications.
Innovation Solution
A three-dimensional sensor system that projects multiple linear patterns simultaneously, using multiple cameras to capture images and generate three-dimensional point-cloud data, with a method to verify and select authentic points based on triangulation and epipolar constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple linear patterns are projected simultaneously, then scanning speed and point-determining efficiency are improved, but image matching complexity increases
Solution Approach 1:
The patent divides the image matching process into two stages: first matching a sparse set of reference points (keypoints) to establish correspondence between images, then using these correspondences to determine 3D points for all projected patterns. This segmentation reduces the computational complexity from matching all points to matching only critical reference points.
Solution Approach 2:
The patent performs preliminary image matching by identifying and matching reference points (keypoints) before determining the 3D coordinates of all points. This preliminary action establishes the correspondence between multiple images in advance, enabling efficient subsequent processing of multiple projected patterns without repeating full image matching.
2Ease of operation
If traditional single linear pattern projection is used, then image matching is simple, but point-determining speed is restricted
Solution Approach 1:
The patent merges multiple linear pattern projections into a single simultaneous operation. By projecting multiple patterns at once and capturing them in a single image, the system combines the advantages of multiple measurements while maintaining the simplicity of single-shot acquisition, eliminating the need for sequential projection and multiple image acquisitions.
3Productivity
If multiple linear patterns are projected, then scanning efficiency increases, but the risk of image mismatch increases
Solution Approach 1:
The patent implements a feedback mechanism where reference points are matched between images to verify correspondence and establish transformation parameters. This feedback loop ensures that even when multiple patterns are projected, the system can detect and correct for image mismatches by continuously verifying the consistency of matched reference points across all images.
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
Significantly improves point-determining efficiency and scanning speed by enabling the recognition and calculation of multiple linear patterns, resulting in several times higher efficiency compared to traditional single-line scanning.
Implementation Method 1
at least two linear patterns are projected
Implementation Method 2
the photosensitive element of camera captures a projection linear pattern on the image
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a three-dimensional sensor system comprising at least one pattern projector configured to project at least two linear patterns simultaneously, at least one first camera and at least one second camera configured to capture a two-dimensional image of a scanned object simultaneously, a two-dimensional image feature extractor configured to extract a two-dimensional line set of the at least two linear patterns on a surface of a scanned object on each two-dimensional image, a three-dimensional point-cloud generator, and a three-dimensional point-cloud verifier.