Stereo Imaging Parallel Optical Axes Adjacent Sensor Readout
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Stereo cameras with non-identical fields of view (FOVs) fail to capture complete depth information due to overlapping image content, resulting in inaccurate or missing depth data for certain image portions, leading to compromised 3D image quality.
Innovation Solution
The arrangement of parallel optical axes for first and second image sensors, with readout areas and support data from adjacent light-sensitive areas, allows for the derivation of disparity information by reading pixel values from specific areas on each sensor, enabling the inclusion of previously uncaptured image content and enhancing depth map accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If two image sensors are arranged with separated positions to model human vision, then stereo imaging capability is achieved, but fields of view become different causing missing image content in certain portions
Solution Approach 1:
The patent extends the two-dimensional image capture by utilizing the adjacent light-sensitive area beyond the standard readout area. This additional spatial dimension provides extra image content that can be used to compensate for the FOV limitations of stereo camera setups, allowing disparity calculation for regions that would otherwise be missing.
Solution Approach 2:
The patent captures and stores support data from adjacent light-sensitive areas in advance, before the actual stereo image processing is needed. This preliminary capture of additional image content ensures that when disparity calculation is performed, all necessary pixel data is already available, eliminating the need for re-capture or estimation.
2Device complexity
If readout areas are limited to standard sensor regions, then data processing is simplified, but disparity information cannot be derived for portions depicted only in one image
Solution Approach 1:
The patent divides the light-sensitive area into distinct functional zones: the standard readout area for primary image capture and the adjacent light-sensitive area for support data collection. This segmentation allows each region to serve its specific purpose while maintaining overall system efficiency and enabling comprehensive disparity calculation.
3Measurement precision
If support data is read from adjacent light-sensitive areas, then disparity information accuracy is improved, but data reading complexity increases
Solution Approach 1:
The patent makes the image sensor serve multiple functions by using the same light-sensitive area for both standard image capture and support data collection. The adjacent regions, which would otherwise be unused or wasted, are repurposed to provide additional image content, maximizing the utility of the sensor hardware.
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 solution ensures accurate disparity information and improved 3D image quality by augmenting the field of view with support data, addressing the limitations of conventional stereo imaging systems.
Implementation Method 1
a first image sensor arranged to receive light via a first lens and a second imaging subsystem that comprises a second image sensor arranged to receive light via a second lens
Data Source
AI summary
A technique for digital stereo imaging is provided. According to an example embodiment, the technique comprises facilitating stereo image capturing means for capturing image data, said stereo image capturing means comprising a first imaging subsystem that comprises a first image sensor arranged to receive light via a first lens and a second imaging subsystem that comprises a second image sensor arranged to receive light via a second lens, wherein said first and second imaging subsystems are arranged such that the optical axes of said first and second lenses are in parallel to each other; reading pixel values from respective light-sensitive areas of said first and second image sensors, comprising reading pixel values for a first image from a first readout area on the first image sensor, reading pixel values for a second image from a second readout area on the second image sensor, and reading pixel values for support data from at least one portion of the light-sensitive area adjacent to the respective readout area on at least one of the first and second image sensors; and deriving disparity information pertaining to said first and second images on basis of said pixel values read for the first image, for the second image and for said support data.