Artificial Reality Duo-Camera Super Resolution with DOE
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing artificial reality devices, such as AR glasses, face a challenge in achieving high-quality image capture while maintaining a small form-factor, as larger lenses and imaging sensors increase size and weight, and single-image super resolution methods fail to recover high-frequency information necessary for tasks like optical character recognition.
Innovation Solution
A duo-camera design with a guide camera and a detail camera, utilizing a diffractive optical element (DOE) and optional refractive lens, splits incoming light into multiple shifted copies, capturing high-quality images from different perspectives, which are then combined using a machine learning model to generate super resolution images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If larger lenses and imaging sensors are used to achieve high-quality image capture, then image quality is improved, but device size and weight increase
Solution Approach 1:
The patent divides the imaging function into two separate camera systems: a guide camera with a wide field of view for capturing overall scene information, and multiple detail cameras with narrower fields of view for capturing high-resolution details. This segmentation allows each camera to be optimized for its specific function, enabling high-quality imaging without requiring a single large camera system that would increase device weight and size.
2Measurement precision
If multiple cameras are used to capture high-quality images from different perspectives, then image quality is improved, but device complexity increases
Solution Approach 1:
The patent merges the functionality of multiple detail cameras into a single detail camera by using a beam splitter optical element. The beam splitter divides the light from the detail camera into multiple shifted copies that are captured on different regions of the sensor, effectively simulating multiple cameras. This merging approach maintains the image quality benefits of multiple cameras while significantly reducing device complexity.
Solution Approach 2:
The patent introduces a beam splitter as an intermediary optical element between the detail camera lens and the sensor. This beam splitter creates multiple shifted copies of the detail image, enabling the system to capture high-frequency information from different perspectives using a single camera. The beam splitter acts as a mediator that transforms a single camera into a multi-perspective imaging system.
3Volume of moving object
If single-image super resolution methods are used to upscale low-resolution images, then device size is reduced, but high-frequency information is lost
Solution Approach 1:
The patent performs preliminary capture of high-frequency information by using a detail camera with a narrower field of view to capture high-resolution details before the final image is constructed. The beam splitter creates multiple shifted copies that preserve gradient information and high-frequency details. This preliminary capture of detailed information allows the system to maintain high-frequency content even when using smaller camera components, avoiding the information loss associated with post-capture upscaling methods.
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
The duo-camera design enables high-quality image capture with a small form-factor, preserving image gradients and recovering high-frequency details, suitable for semantic tasks like OCR, while reducing the number of cameras and associated challenges.
Implementation Method 1
The optical elements of the duo-camera artificial reality device may include at least a diffractive optical element (DOE)
Implementation Method 2
In some embodiments, a refractive lens may also be included that may be used to minimize effects of chromatic aberration
Data Source
AI summary
In particular embodiments, a computing system may capture a first image of a scene using a first camera of an artificial reality device. The system may capture a second image of the scene using a second camera and one or more optical elements of the artificial reality device. The second image may include an overlapping portion of multiple shifted copies of the scene. The system may generate an upsampled first image by applying a particular sampling technique to the first image. The system may generate a tiled image comprising a plurality of repeated second images by applying a tiling process to the second image. The system may generate an initial output image by processing the upsampled first image and the tiled image using a machine learning model. The system may generate a final output image by normalizing the initial output image using the upsampled first image.