Image Processing for HMDs with Gaze-Adaptive Intensity Blending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques for generating low-intensity images for display apparatuses, such as HMDs, result in non-uniform and suboptimal immersive experiences due to challenges in combining multiple low-intensity images with different scotopic and photopic vision ranges, leading to prominent perceptual color differences.
Innovation Solution
A system and method that adjust the intensities of pixels within first and second images to achieve smooth blending, accommodating differences in color reproduction properties of various image renderers, and optically combining these images to emulate foveation characteristics, ensuring uniformity and immersion under both scotopic and photopic vision conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple low-intensity images are optically combined to produce a low-intensity scene, then the scene brightness is reduced, but the scene appears non-uniform to the user
Solution Approach 1:
The patent applies local quality by differentiating image processing between the foveal region (center of vision) and peripheral regions. The foveal region receives higher resolution, higher intensity images while peripheral regions receive lower resolution, lower intensity images. This local differentiation maintains uniformity in the critical viewing area while allowing overall scene brightness reduction.
Solution Approach 2:
The patent segments the visual field into foveal and peripheral regions, processing each region separately with appropriate intensity and resolution settings. This segmentation allows the system to maintain uniformity in the foveal region where uniformity is most critical, while reducing overall intensity across the entire scene.
2Adaptability or versatility
If intensities of multiple low-intensity images are adjusted using conventional techniques, then the images can be combined, but prominent perceptual color differences become visible
Solution Approach 1:
The patent changes intensity parameters differently for foveal and peripheral regions. The foveal region maintains higher intensity and accurate color reproduction, while peripheral regions use lower intensity. This parameter differentiation compensates for the different color reproduction ranges of multiple image renderers, preventing perceptual color differences from becoming prominent.
Solution Approach 2:
The patent dynamically adjusts image parameters based on the user's gaze direction. As the user moves their gaze, the system dynamically recalculates which region is foveal and applies appropriate intensity and color correction. This dynamic adaptation ensures consistent color uniformity across different viewing positions and conditions.
3Device complexity
If conventional image processing techniques are used, then the processing is simpler, but the immersive experience is suboptimal
Solution Approach 1:
The patent segments processing into two distinct pipelines: one for foveal region images and one for peripheral region images. This segmentation allows each pipeline to be optimized independently - the foveal pipeline maintains high quality for immersion while the peripheral pipeline can use simpler, more efficient processing. The segmented approach delivers superior immersive experience without requiring all processing to be equally complex.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
Disclosed is system (100, 200) for producing images for display apparatus (102, 202). The system comprises image source (104, 204) to obtain input image and processor(106, 206). The processor is configured to obtain information of gaze direction of user, determine region of interest of input image based on gaze direction, and process input image to generate first image (302) and second image(304).First image comprises first region (302A) that is blurred with respect to region of interest. Second image corresponds to region of interest. Processor adjusts intensity of pixels within first region of first image and intensity of pixels within second image. When intensity of given pixel within region of interest is lower than or equal to predefined intensity threshold, intensity of a corresponding pixel (306A) within first region of first image is lower than intensity of corresponding pixel (308A) within second image.