Parallel Depth Fragment Processing for Mobile Fidelity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing algorithms for estimating depth information values in digital images face a trade-off between fidelity and computational efficiency, with high-fidelity methods being computationally complex and unsuitable for real-time processing, while simple methods provide limited fidelity and are suitable for real-time processing but not on mobile devices.
Innovation Solution
An image processing apparatus and method that processes fragments of a digital image in parallel, computing similarity measures for depth information value candidates and applying an updating function to determine the best depth information value by comparing candidates to previously selected values, using weighting functions to adjust similarity measures based on differences or similarities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sophisticated cost functions with smoothing terms and multiple iterations are used for depth estimation, then measurement precision is improved, but device complexity and computational requirements increase making real-time processing infeasible
Solution Approach 1:
The patent divides the image into multiple layers or subsets of fragments, processing them in a hierarchical manner. This segmentation allows the system to apply different processing strategies to different parts of the image, reducing overall computational complexity while maintaining depth estimation precision through multi-layer processing.
Solution Approach 2:
The patent performs preliminary processing by computing depth information values for all fragments first, then uses these preliminary results to guide subsequent refinement steps. This preliminary action allows the system to avoid computationally expensive operations on all fragments while still achieving high precision where needed.
2Productivity
If simple depth estimation algorithms are used, then productivity is improved for real-time processing, but measurement precision deteriorates providing limited fidelity
Solution Approach 1:
The patent implements a dynamic processing approach where the level of processing applied to each fragment or layer can be adjusted based on local image characteristics, computational resources available, and desired output quality. This allows the system to adapt between speed and precision modes dynamically.
Solution Approach 2:
The patent applies different processing qualities to different regions or layers of the image. Some regions may receive full high-precision processing while others use simplified algorithms, allowing the system to maintain high fidelity where needed while improving overall processing throughput.
3Measurement precision
If high-fidelity depth estimation methods are used, then measurement precision is improved, but use of energy increases making mobile device processing unsuitable
Solution Approach 1:
By segmenting the processing into multiple layers, the system can distribute computational workload over time and energy budget, allowing high-precision processing to be applied selectively rather than uniformly across the entire image, thus reducing total energy consumption.
Solution Approach 2:
The patent applies high-precision processing only where necessary (partial action) rather than to all fragments uniformly. This selective application of computationally intensive algorithms reduces overall energy consumption while maintaining depth estimation fidelity in critical regions.
Data Source
AI summary
The disclosure relates to an image processing apparatus for selecting a plurality of depth information values for a subset of currently processed fragments of a set of currently processed fragments of a currently processed digital image. The image processing apparatus includes a non-transitory memory storage comprising instructions and one or more hardware processors in communication with the memory storage. The one or more hardware processors execute the instructions to: process the currently processed fragments of the set of currently processed fragments in parallel; and process the currently processed fragments of the subset of currently processed fragments in parallel.