Structured Light Projection for 3D Depth Mapping on Low-Texture Scenes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic systems face challenges in accurately determining 3D scene geometry, particularly when objects have little or no visual texture, making it difficult to identify corresponding features in images and resulting in reduced spatial resolution and smearing of depth information.
Innovation Solution
Projecting multiple different patterns of light onto a scene using multiple projectors with varying wavelengths, and using optical sensors to distinguish between these patterns, followed by triangulation to determine corresponding features and generate a virtual representation of the environment, including depth measurements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional stereo vision is used to determine 3D scene geometry, then the system structure is simple, but measurement precision deteriorates when objects have little or no visual texture
Solution Approach 1:
The patent introduces projected light patterns as an intermediary to provide artificial visual texture on objects. These patterns serve as mediators between the stereo vision system and the target objects, enabling feature correspondence identification even when objects lack inherent visual texture. The projected patterns are captured by the stereo camera system and used to compute 3D geometry through triangulation.
2Measurement precision
If multiple different patterns of light are projected onto a scene, then spatial resolution and measurement precision are improved, but device complexity increases
Solution Approach 1:
The patent segments the lighting function by using multiple projectors, each emitting light at distinct wavelengths. This segmentation allows the system to project multiple different patterns simultaneously without interference, as each pattern can be distinguished by its wavelength. The segmentation of the optical sensing function into multiple sensors tuned to different wavelengths enables simultaneous capture of multiple patterns, improving spatial resolution while managing system complexity through functional division.
3Measurement precision
If multiple projectors with varying wavelengths are used, then corresponding feature identification is improved, but ease of operation deteriorates due to increased system complexity
Solution Approach 1:
The system employs feedback mechanisms where the computing device processes sensor data from multiple optical sensors, identifies corresponding features across different wavelength channels, and uses this information to reconstruct 3D scene geometry. The feedback loop involves continuously adjusting and optimizing the identification of corresponding features based on the patterns projected and captured, improving measurement accuracy while managing operational complexity through automated computational processes.
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
Enhances the accuracy and detail of 3D reconstruction by reducing smearing and improving spatial resolution, enabling more precise robotic manipulation and automation of tasks such as loading and unloading objects.
Implementation Method 1
projecting a plurality of different patterns of light using a plurality of projectors
Implementation Method 2
receiving sensor data by a computing device and from a plurality of optical sensors
Implementation Method 3
triangulating information observed from at least two known viewpoints to determine a representation of three-dimensional (3D) scene geometry
Data Source
AI summary
Example methods and systems for determining 3D scene geometry by projecting patterns of light onto a scene are provided. In an example method, a first projector may project a first random texture pattern having a first wavelength and a second projector may project a second random texture pattern having a second wavelength. A computing device may receive sensor data that is indicative of an environment as perceived from a first viewpoint of a first optical sensor and a second viewpoint of a second optical sensor. Based on the received sensor data, the computing device may determine corresponding features between sensor data associated with the first viewpoint and sensor data associated with the second viewpoint. And based on the determined corresponding features, the computing device may determine an output including a virtual representation of the environment that includes depth measurements indicative of distances to at least one object.