Tonemapping Acceleration via Mipmap Hardware Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for tonemapping high dynamic range (HDR) images are computationally expensive, making it difficult to process higher resolutions and frame rates on affordable and compact devices like smartphones and hand-held cameras.
Innovation Solution
The use of a mipmap representation of images, combined with hardware-accelerated sampling and Gaussian blurring, to accelerate tonemapping calculations, allowing for efficient use of memory and computational resources, and enabling real-time processing of HDR video.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional tonemapping methods are used, then image quality is maintained, but computational cost increases making real-time processing difficult
Solution Approach 1:
The patent divides the tonemapping computation into multiple independent stages: HDR to LDR conversion, contrast calculation at multiple scales, and final composition. Each stage can be processed separately and optimized independently, enabling parallel processing and reducing overall computational burden for real-time operation
Solution Approach 2:
The patent implements multi-scale contrast calculation where only the necessary number of scales are computed based on image content and display requirements. Rather than computing all possible scales, the method selectively processes only those scales that contribute meaningfully to the final output, reducing unnecessary computational overhead
2Productivity
If higher resolution and frame rate processing is implemented, then image quality and productivity improve, but device complexity and cost increase
Solution Approach 1:
The processing pipeline is segmented into discrete, independently optimizable stages that can be distributed across different hardware units. This allows affordable devices to implement tonemapping at reduced resolutions or frame rates by selectively enabling or disabling specific processing stages
Solution Approach 2:
The patent allows dynamic adjustment of processing parameters such as output resolution, frame rate, and contrast enhancement strength based on device capabilities. This enables the same algorithm to run efficiently on both high-end devices (producing high-resolution real-time output) and affordable devices (producing lower-resolution or lower-frame-rate output) without requiring fundamentally different hardware architectures
Data Source
AI summary
According to some embodiments, a camera captures video images at a high dynamic range. These images are then tonemapped into images of a lower dynamic range with enhanced contrast. The contrast enhancement for a given pixel depends on the image's local contrast at a variety of different scales. The tonemapped images are then shown on a display. Calculation of this contrast is accelerated by the camera creating a plurality of low-pass filtered versions of the original image at progressively stronger low-pass filtering; these images may be stored at increasingly lower resolutions in a mipmap. Calculations are enhanced by use of a massively parallel processor and a texture mapping unit for hardware-accelerated sampling of blended averages of several pixels. Other embodiments are shown and discussed.