Stereo Depth Map Disparity Sampling for Far-Range Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining depth maps from stereo images, such as SGM, face challenges in achieving high accuracy in far-range depth measurements due to limitations in disparity sampling, leading to reduced depth resolution and increased resource demands.
Innovation Solution
A method that selectively samples disparities with a non-uniform distribution, providing finer intervals in the far range to enhance depth resolution without increasing resource usage across all ranges, by reducing sampling in near ranges and using a comparison operator like census to determine sub-pixel accurate disparities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If uniform disparity sampling is used across all ranges, then depth resolution in near range is maintained, but depth resolution in far range deteriorates
Solution Approach 1:
The patent applies local quality by using different disparity sampling intervals for different depth ranges. Specifically, a first sampling interval is used for a first depth range and a second, finer sampling interval is used for a second depth range. This allows the system to allocate computational resources locally where they are most needed - with finer sampling in the far range where depth resolution deteriorates, while maintaining coarser sampling in the near range where resolution is already sufficient.
2Measurement precision
If finer disparity sampling is used in far range, then depth resolution in far range is improved, but resource demands increase
Solution Approach 1:
The patent implements local quality by applying finer disparity sampling only to the far depth range where it is needed for improved depth resolution, while using coarser sampling in the near range. This localized approach to quality enhancement avoids the need to increase sampling density across the entire depth range, thereby improving far-range measurement precision without proportionally increasing overall device complexity and resource demands.
3Measurement precision
If disparity values are determined as integer shifts, then computing speed is maintained, but depth resolution deteriorates
Solution Approach 1:
The patent applies parameter changes by transitioning from integer disparity shifts to sub-integer disparity values in the far depth range. By allowing disparity values to take non-integer values (sub-integer shifts), the system achieves finer depth resolution in the far range where precision is most critical, while maintaining integer shifts in the near range where computing speed is prioritized. This selective parameter change optimizes the trade-off between precision and speed in different operational contexts.
Data Source
AI summary
In a method for determining a depth map from stereo images, the disparity for a pixel is selected from a predefined quantity or set of predefined discrete disparity values that are distributed over the entire predefined disparity value range, whereby the distribution is non-uniform or has at least two different distances or intervals between different adjacent disparity values. This method makes it possible to more precisely determine (with finer resolution) especially only those disparities for which a more precise determination is required.