Stereo Matching Hardware Disparity Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for stereo matching in depth mapping face challenges due to differences between cameras and variations in feature presentation, such as lighting and acquisition noise, which complicate feature matching and are not well-suited for hardware solutions.
Innovation Solution
A method involving binarization of images, calculation of matching costs, and creation of a probability density function to produce disparity and confidence maps, implemented using a special-purpose digital signal processor, allowing for real-time processing and improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional stereo matching methods are used, then feature matching can be performed, but accuracy is reduced due to camera differences and feature presentation variations
Solution Approach 1:
The patent transforms image pixel values into binary representations (0 or 1) based on thresholding operations. This parameter transformation simplifies the complex continuous pixel values into discrete states, making the feature matching process more robust to lighting variations and camera differences. The binary transformation preserves essential structural information while eliminating sensitivity to illumination changes.
Solution Approach 2:
The patent divides the image processing into distinct sequential stages: binarization of master and slave images, calculation of matching costs for corresponding pixel pairs, generation of probability density functions, and extraction of disparity maps. This segmentation allows each stage to be optimized independently and enables hardware implementation through dedicated processing units for each operation.
2Measurement precision
If complex feature matching algorithms are applied, then matching accuracy improves, but processing time increases and real-time capability is lost
Solution Approach 1:
The patent replaces complex software-based algorithms with a specialized hardware architecture designed for parallel processing. The hardware system includes dedicated units for binarization, matching cost calculation, and probability density function generation, enabling simultaneous processing of multiple pixel pairs. This hardware acceleration maintains high accuracy while achieving real-time performance.
Solution Approach 2:
The patent focuses computational resources on calculating matching costs for corresponding pixel pairs rather than exhaustively analyzing all possible feature matches. By concentrating on the most promising candidate pixels and using probability density functions to filter results, the system achieves sufficient accuracy with reduced computational effort, enabling real-time operation.
3Productivity
If hardware implementation is pursued, then real-time processing is enabled, but handling of image variations becomes more difficult
Solution Approach 1:
The patent employs binary thresholding that transforms continuous image data into discrete representations. This parameter transformation creates a simplified data structure that is inherently more robust to variations in lighting, noise, and camera characteristics. The binary representations maintain essential geometric information while being insensitive to illumination changes, enabling reliable hardware processing under diverse conditions.
Data Source
AI summary
A method for calculating disparity in a pair of images includes receiving a first image of a scene and designating the first image as a master image. A second image of the scene is received, and the second image is designated as a slave image. The master image is binarized to produce a binarized master image. The slave image is binarized to produce a binarized slave image. A matching cost associated with matching each pixel within the binarized master image with a corresponding set of candidate pixels within the binarized slave image is calculated. A probability density function is created based on the calculated matching costs associated with each pixel within the binarized master image. The created probability density function is used to produce a disparity for the master image and the slave image and to produce a confidence for the produced disparity.