Perceptual Texture Deduplication for Storage and Bandwidth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing size of textures in digital content, particularly in video games, poses challenges for storage and transmission, especially for users with limited bandwidth or storage capacity, as existing solutions like optional texture packs or compression are inadequate.
Innovation Solution
Identify perceptually equivalent textures using color and spatial properties to reuse existing textures, reducing the need for duplicate storage by using pointers or generating new textures that combine similarities, thereby optimizing storage and transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If texture size and detail are increased to improve graphical quality, then visual fidelity is improved, but storage space and data transmission requirements increase significantly
Solution Approach 1:
The patent creates perceptual copies of textures - simplified versions that replicate the visual appearance of high-detail textures at a fraction of the data size. These copies are generated by analyzing the visual importance of different texture regions and reconstructing only the essential features, allowing storage and transmission of texture data that appears identical to users but occupies minimal space.
Solution Approach 2:
The patent transforms texture data by changing its representation parameters - converting high-resolution pixel data into a compact set of visual features and parameters that can reconstruct the texture's appearance. This involves analyzing texture complexity, identifying dominant patterns, and storing only the essential parameters needed to recreate the visual effect, rather than storing all original pixel information.
2Reliability
If high-quality textures are provided for all content to maintain consistent user experience, then visual quality is improved, but data transmission bandwidth and storage capacity requirements increase
Solution Approach 1:
The patent applies different levels of texture quality locally based on visual importance - critical texture regions are preserved with high fidelity while less important regions are simplified or omitted. This allows the system to maintain perceived visual quality where it matters most while significantly reducing overall data transmission and storage requirements.
Solution Approach 2:
The patent transmits only the essential portion of texture data needed to achieve acceptable visual quality - sending a compressed representation that contains just enough information to reconstruct textures that appear high-quality to users, rather than transmitting complete high-resolution texture data for all scenarios.
3Quantity of substance
If texture compression is applied to reduce file size, then storage efficiency is improved, but visual quality and detail are degraded
Solution Approach 1:
The patent replaces traditional mechanical compression methods with a perceptual analysis system that understands human visual processing. Instead of applying generic compression algorithms that uniformly reduce quality, the system analyzes texture content, identifies visually critical features, and preserves those while allowing compression of non-essential elements, achieving both small file size and high visual quality.
Data Source
AI summary
Techniques for managing storing textures corresponding to respective content are described. In an example, a system identifies a texture to be stored at a device, the texture associated with first content. The system identifies a degree of perceptual similarity between texture associated with the first content and each of one or more candidate textures associated with content other than the first content, the one or more candidate textures stored at the device. The perceptual equivalence between textures is identified in response to an above-threshold degree of perceptual similarity identified for those textures. When a perceptual equivalence is identified between the texture associated with the first content and a candidate texture, the system generates and stores a pointer directed toward that candidate texture instead of texture associated with the first content. When no perceptual equivalence is identified, the system stores the texture associated with the first content.