Image Sensor Inclination for Asymmetric Pixel Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional light field cameras capture view information in equal proportions in all directions, which does not meet the differential requirements for applications like multiview image generation and 3D reconstruction, as human eyes are more sensitive to horizontal views, leading to inefficient use of pixel resources.
Innovation Solution
Inclining image sensing units relative to their initial normal line to change the ratio of pixels distributed along two directions in the effective region, allowing for differential capture of view information in different directions by adjusting the shape and size of the effective region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If image sensing units are arranged with equal pixel distribution in all directions, then the camera can capture view information uniformly in all directions, but this does not meet the differential requirements for applications like multiview image generation and 3D reconstruction where human eyes are more sensitive to horizontal views
Solution Approach 1:
The patent applies asymmetry by inclining the image sensing unit relative to the optical axis, which transforms the originally symmetric circular effective region into an asymmetric elliptical effective region. This asymmetric arrangement causes pixels to be distributed unevenly in different directions, with greater pixel density in the horizontal direction to match human visual sensitivity characteristics, thereby resolving the contradiction between uniform capture and differential application requirements
Solution Approach 2:
The patent implements local quality by creating non-uniform pixel distribution within the effective region through inclination. Different regions of the image sensing unit have different pixel densities - the horizontal direction has higher pixel density while the vertical direction has lower pixel density. This local differentiation optimizes pixel resource allocation according to the specific requirements of multiview and 3D reconstruction applications
2Productivity
If the effective region is circular with uniform pixel distribution, then all pixels are used equally for capturing view information in all directions, but this results in equal proportions of parallax information in all directions which does not meet diverse application requirements
Solution Approach 1:
By inclining the image sensing unit, the patent transforms the symmetric circular effective region into an asymmetric elliptical effective region. This asymmetric geometry naturally creates differential pixel distribution where more pixels fall within the effective region along the horizontal axis compared to the vertical axis, enabling higher productivity in horizontally-sensitive applications like multiview generation while maintaining reasonable coverage in other directions
3Adaptability or versatility
If image sensing units are inclined relative to the initial normal line, then the ratio of pixels distributed along two directions in the effective region changes to enable differential capture, but this requires adjustment of the imaging system configuration
Solution Approach 1:
The patent applies dynamics by making the image sensing unit's orientation adjustable rather than fixed. The inclination angle can be dynamically changed to optimize pixel distribution for different application scenarios. This dynamic adjustability allows the system to adapt between different effective region shapes (from circular to various elliptical configurations) without requiring multiple fixed cameras, thereby achieving differential capture capability while controlling overall system complexity
Data Source
AI summary
Various image capture control methods and apparatuses are disclosed. An image sensing unit can be inclined relative to an initial normal line, to change a ratio of pixels distributed along two directions in an effective region of the image sensing unit in which light through a lens that is in an imaging system and corresponds to the image sensing unit is imaged on the image sensing unit. The two directions are perpendicular to the initial normal line and orthogonal to each other, the image sensing unit comprises pixels distributed in array, and the pixels are used for recording multidirectional view information of a same object of a scene. An image of the scene is captured using the imaging system. Proportions of parallax information actually recorded by the image sensing unit in the two directions can be changed, and differential capture of view information in different directions can be implemented.