Structured Light Planar Surface Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D imaging systems require high computational costs to detect planar surfaces from dense 3D point clouds, which can be noisy and resource-intensive, especially in applications like autonomous navigation and augmented reality where dense point clouds are not necessary.
Innovation Solution
A system using structured light that projects a two-dimensional light pattern with intersecting line segments on an epipolar line, allowing for direct recovery of planar surfaces by estimating plane hypotheses based on pattern and image feature properties, reducing the need for dense 3D point cloud generation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If dense 3D point cloud data is captured and processed to detect planar surfaces, then measurement precision is improved, but computational cost and processing time increase significantly
Solution Approach 1:
The patent extracts only the essential information needed for plane detection by projecting structured light patterns with specific features (intersecting lines on epipolar lines) rather than capturing complete dense point clouds. This extraction approach obtains sufficient geometric data for plane detection while eliminating unnecessary data that would increase computational burden.
Solution Approach 2:
The patent segments the problem of plane detection by using structured light patterns divided into multiple epipolar lines with specific features. Each epipolar line contains intersecting line segments that independently encode depth and plane information, allowing the system to process information in structured segments rather than handling entire dense point clouds.
2Measurement precision
If dense 3D point cloud data is generated, then planar surface detection accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary encoding of depth and geometric information directly into the projected light pattern structure before the actual measurement occurs. The intersecting line segments on epipolar lines are pre-configured to contain depth encoding information, so that when these patterns are captured, the plane detection can proceed directly without requiring time-consuming dense point cloud generation and processing.
3Loss of information
If complete 3D point cloud processing is performed, then comprehensive scene understanding is achieved, but computational resources are excessively consumed
Solution Approach 1:
The patent applies local quality by designing structured light patterns with different feature distributions across different regions of the scene. Each epipolar line and its intersecting line segments are optimized to provide locally sufficient information for plane detection in that specific region, rather than uniformly processing the entire scene with dense point clouds. This localized information encoding reduces overall computational energy while maintaining scene understanding quality.
Data Source
AI summary
In accordance with some embodiments, systems, methods and media for directly recovering planar surfaces in a scene using structured light are provided. In some embodiments, a system comprises: a light source; an image sensor; a processor programmed to: cause the light source to emit a pattern comprising a pattern feature with two line segments that intersect on an epipolar line; cause the image sensor to capture an image including the pattern; identify an image feature in the image, the image feature comprising two intersecting line segments that intersect at a point in the image that corresponds to the first epipolar line; estimate a plane hypothesis associated with the pattern feature based on properties of the pattern feature and properties of the image feature, the plane hypothesis associated with a set of parameters characterizing a plane; and identify a planar surface in the scene based on the plane hypothesis.