Stereo Disparity Computation via Vertex Space Embedding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining disparity values in stereoscopic imaging are computationally expensive and inefficient, particularly when dealing with large images, as they require calculating disparities for each pixel location, which can be time-consuming and resource-intensive.
Innovation Solution
The method involves receiving two images of a scene, determining a disparity loss function for each pixel location, associating these locations with vertices in a vertex space that includes spatial and color-space dimensions, and using vertex-disparity loss functions to calculate disparity values, thereby reducing the computational burden by focusing on a smaller set of vertices rather than individual pixels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If disparity values are calculated for each pixel location individually, then measurement precision is improved, but computing time increases significantly
Solution Approach 1:
The image is divided into superpixels or patches, and disparity is computed for each segment rather than every individual pixel. This segmentation reduces the total number of computations while maintaining accuracy through intra-segment assumptions.
Solution Approach 2:
Adjacent pixels with similar properties are grouped together and processed as a unit. The disparity values of grouped pixels are combined or averaged to produce a representative disparity value, reducing computational load while preserving measurement precision.
2Measurement precision
If disparity loss functions are computed for all pixel locations, then disparity determination accuracy is improved, but computational complexity increases
Solution Approach 1:
Only essential or representative pixel locations are selected for disparity loss function computation, such as key pixels within superpixels or pixels at boundaries of regions. This extraction approach maintains accuracy for critical areas while reducing overall computational complexity.
Solution Approach 2:
Disparity loss functions are computed for a subset of pixel locations rather than all pixels. The computation is performed partially on representative samples, which is sufficient to determine disparity accurately without the excessive computational burden of processing every pixel.
3Measurement precision
If high-resolution images are processed for stereo matching, then depth information precision is improved, but processing speed decreases
Solution Approach 1:
High-resolution images are segmented into smaller regions or superpixels, allowing efficient processing of large images while maintaining the precision benefits of high resolution. Each segment is processed independently or with reduced complexity algorithms.
Solution Approach 2:
Full disparity computation is performed only on a subset of high-resolution pixel locations, such as key points or representative samples. This partial computation maintains depth precision for critical areas while achieving practical processing speeds.
Data Source
AI summary
Example embodiments may allow for the efficient determination of disparity information for a stereo image pair by embedding pixels of the image pair in a multidimensional dimensional vertex space. Regularly-spaced vertices in the vertex space are associated with pixels of the stereo image pair and disparity loss functions are determined for each of the vertices based on disparity loss functions of the associated pixels. The determined vertex-disparity loss functions can be used to determine vertex disparity values for each of the vertices. Disparity values for pixels of the stereo image pair can be determined based on determined vertex disparity values for respective one or more vertices associated with each of the pixels. The determined pixel disparity values can be used to enable depth-selective image processing, determination of pixel depth maps, mapping and/or navigation of an environment, human-computer interfacing, biometrics, augmented reality, or other applications.