3D Data Acquisition Sub-Pixel Dequantization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing coded light methods for three-dimensional geometry acquisition face limitations due to finite camera resolution, leading to quantization of code patterns and reduced spatial resolution, which can be improved by dequantization techniques that achieve sub-pixel accuracy and edge preservation.
Innovation Solution
The implementation of dequantization filters, such as one-dimensional edge-preserving filters and bilateral filters, to interpolate code and range values at sub-pixel locations, ensuring smooth interpolation between edges while preserving discontinuities, thereby enhancing the resolution of three-dimensional object geometry reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coded light methods use finite camera resolution for three-dimensional geometry acquisition, then the system is practical and implementable, but the spatial resolution is reduced due to quantization of code patterns
Solution Approach 1:
The patent transitions from discrete pixel-level code detection to continuous sub-pixel level code detection by introducing fractional pixel coordinates. This dimensional change allows the system to detect code transitions at sub-pixel precision, effectively adding a fractional dimension to the pixel coordinate system and achieving super-resolution beyond the camera's native pixel grid.
Solution Approach 2:
The patent changes the parameter of measurement precision from integer pixel coordinates to fractional sub-pixel coordinates. By detecting code transitions at sub-pixel levels and using these fractional positions to calculate three-dimensional points, the system achieves higher spatial resolution than the camera's pixel resolution would normally allow.
2Measurement precision
If dequantization techniques are applied to achieve sub-pixel accuracy, then the resolution of three-dimensional reconstruction is improved, but the complexity of processing increases
Solution Approach 1:
The patent performs preliminary detection of code transition locations at sub-pixel accuracy before final three-dimensional reconstruction. By detecting transitions at fractional pixel positions and calculating sub-pixel code values in advance, the system prepares high-precision data that simplifies the subsequent reconstruction process rather than complicating it.
Solution Approach 2:
The patent replaces traditional integer-based pixel sampling with continuous sub-pixel detection methods. Instead of relying on discrete pixel grids, the system uses continuous coordinate representation and mathematical interpolation to achieve sub-pixel precision, substituting mechanical pixel sampling with computational precision.
3Reliability
If traditional quantization methods are used, then the processing is simpler, but edge preservation and discontinuity detection are degraded
Solution Approach 1:
The patent applies different processing qualities to different regions of the image based on local characteristics. By detecting edge locations and applying sub-pixel detection specifically at transition regions while maintaining simplicity in uniform regions, the system achieves high edge preservation where needed without unnecessarily complicating processing in all areas.
Solution Approach 2:
The patent applies sub-pixel detection and dequantization selectively at code transition regions rather than uniformly across the entire image. This partial application of complex processing only where transitions occur achieves edge preservation benefits while limiting the overall processing complexity to only the necessary regions.
Data Source
AI summary
A projector illuminates an object, within the field of view of a camera, with a sequence of code patterns. The camera captures the illuminated object and provides object images to a decoder to convert the code patterns into code. A transition locator locates discontinuities in the code pattern images. A dequantizer reconstructs a range image from those discontinuities and said code.