Multispectral Binary Coded Projection for Robust 3D Depth Measurement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional 3D scanning using structured light faces challenges with binary coded patterns, which require multiple projections and are not robust to depth discontinuities, while color coded patterns like rainbow projections are not suitable for dynamic scenes due to high capture times and sensitivity to abrupt depth changes.
Innovation Solution
The use of spectrally structured light with a multichannel projector and multispectral camera, allowing for real-time 3D measurement and material differentiation by capturing few multispectral images, leveraging Gray code patterns and photometric stereo methods to achieve robustness and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If binary coded patterns are used for 3D scanning, then depth measurement robustness is improved, but the number of projected patterns increases making high-speed scanning difficult
Solution Approach 1:
The patent combines multiple binary-coded patterns into a single composite pattern that encodes depth information across multiple wavelengths. Instead of sequentially projecting 10 separate Gray code patterns, the system projects one pattern containing multiple coded channels simultaneously, reducing capture time while maintaining depth measurement robustness through the coded structure.
Solution Approach 2:
The patent introduces a spectral dimension to the spatial coding problem. By encoding depth information across different wavelengths rather than only across multiple sequential patterns, the system achieves high-depth-resolution measurements in a single capture, transforming a temporal sequencing problem into a spectral encoding problem.
2Productivity
If rainbow projections are used for color coded patterns, then the number of projected patterns is reduced, but the system becomes sensitive to abrupt depth changes and unsuitable for dynamic scenes
Solution Approach 1:
The patent applies different coding strategies to different spatial locations within the projected pattern. By using binary-coded patterns that are robust to discontinuities in specific regions while maintaining overall spectral encoding efficiency, the system achieves both fast capture and reliability at depth boundaries.
3Measurement precision
If hyperspectral imaging is used for material classification, then measurement precision is improved, but capture time increases making real-time classification impossible
Solution Approach 1:
The patent extracts only the essential spectral information needed for material classification rather than capturing complete hyperspectral data cubes. By selecting specific wavelength bands that provide sufficient material discrimination capability, the system achieves accurate material classification with significantly reduced capture time.
Solution Approach 2:
The patent uses a limited set of strategically chosen wavelength bands rather than full spectral coverage. This partial spectral sampling provides sufficient information for material differentiation while avoiding the time penalty of complete hyperspectral imaging, achieving the minimum necessary action for effective classification.
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
This approach enables robust 3D measurement and material classification in real-time, reducing the number of projected patterns and capture time, while maintaining accuracy and efficiency, even in dynamic scenes.
Implementation Method 1
an object is illuminated with structured light that is also spectrally structured, and light reflected therefrom is measured spectrally
Data Source
AI summary
Illumination of an object with spectral structured light, and spectral measurement of light reflected therefrom, for purposes which include derivation of a three-dimensional (3D) measurement of the object, such as depth and/or contours of the object, and/or for purposes which include measurement of a material property of the object, such as by differentiating between different types of materials.