Non-uniform pixel spacing for HMD rasterization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graphics processing technologies waste significant processing power by rendering uniformly spaced pixels across the entire window space, even though higher resolution is only necessary in the high-acuity region around the central gaze direction for head-mounted displays (HMDs) in AR and VR applications.
Innovation Solution
The method involves rendering pixels in window space with non-uniform spacing, achieving a constant pixel density per subtended angle across the user's field of view. This is achieved by using a distortion function that maps non-uniformly spaced pixels to uniformly spaced pixels in raster space, allowing for higher resolution in the fovea region and lower resolution in the visual periphery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If uniformly spaced pixels are rendered across the entire window space, then the image resolution is consistent across the display, but significant processing power is wasted in regions where high resolution is not necessary
Solution Approach 1:
The patent applies local quality by rendering pixels with non-uniform spacing such that higher resolution is concentrated in the high-acuity region (fovea) surrounding the central gaze direction, while lower resolution is used in the visual periphery. This matches the human visual system's characteristics where central vision requires high detail but peripheral vision does not, thereby reducing overall processing power requirements while maintaining perceived image quality.
2Use of energy by moving object
If non-uniform pixel spacing is used to reduce processing power, then resource usage is optimized, but the pixel density per subtended angle becomes non-uniform requiring distortion correction
Solution Approach 1:
The patent applies preliminary action by pre-defining distortion functions that map non-uniformly spaced pixels to uniformly spaced pixels in raster space. These distortion functions are prepared in advance based on the desired pixel density per subtended angle, allowing the rendering system to efficiently convert between coordinate systems without requiring complex real-time calculations during the rendering process.
3Measurement precision
If higher resolution is rendered in the fovea region, then perceived image quality is improved, but the overall rendering workload increases
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the pixel spacing parameter based on the spatial location within the field of view. The pixel density per subtended angle is modified as a function of eccentricity (angular distance from central gaze direction), with higher density values assigned to the fovea region and lower density values to peripheral regions. This optimized parameter distribution improves perceived image quality in critical areas while maintaining overall rendering efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus, such as a head mounted device (HMD)(300, 125), includes one or more processors (335) configured to implement a graphics pipeline (100) that renders pixels in window space (405, 510) with a nonuniform pixel spacing. The apparatus also includes a first distortion function (109, 525) that maps the non-uniformly spaced pixels in window space to uniformly spaced pixels in raster space (410, 520). The apparatus further includes a scan converter (110, 1225) configured to sample the pixels in window space through the first distortion function. The scan converter is configured to render display pixels used to generate an image for display to a user based on the uniformly spaced pixels in raster space. In some cases, the pixels in the window space are rendered such that a pixel density per subtended area is constant across the user's field of view.