The accelerating shift towards remote work, virtual collaboration, and digital entertainment is intensifying the need for robust, high-performance video infrastructure. Companies face immense pressure to deliver seamless, high-quality visual experiences while managing soaring data costs and network congestion. This technology provides a strategic advantage by enabling more efficient content delivery and storage, allowing businesses to scale services, reduce operational expenses, and meet consumer expectations for superior visual fidelity across all platforms.
Reduces data volume by up to 30% by evaluating pixel-level similarity across multiple reference images, significantly improving prediction accuracy and lowering storage costs and network bandwidth load.
Maintains high-quality video reproduction by correcting synthesis targets based on prediction accuracy evaluation, minimizing quality degradation like noise and block artifacts even at high compression ratios, enabling high-definition visual experiences.
Enables real-time processing with optimized efficiency by limiting prediction accuracy evaluation to bi-directional prediction, reducing processing load. This supports applications in real-time streaming and video conferencing systems.
This patent protects an image encoding device, an image decoding device, and a program, with 7 claims covering the core components. Specifically, Claim 1 broadly and robustly protects the evaluation unit that calculates similarity between multiple reference images per image portion and the synthesis unit that corrects synthesis targets based on the evaluation results. The successful grant after two rounds of examiner rejections demonstrates its clear inventive step and robustness against invalidation.
This patent primarily covers core encoding/decoding logic. White space exists in developing adaptive streaming protocols, content-aware encoding strategies, or specialized hardware accelerators that leverage this improved compression.
Assuming an enterprise handles 10 petabytes (PB) of video data annually, and this technology reduces data volume by 30% compared to conventional compression. With cloud storage costs at ~$20/TB/month (AI est.) and network bandwidth costs at ~$33.5/TB/month (AI est.), the estimated annual savings would be (10,000 TB × 0.3) × ($20 + $33.5) × 12 months = ~$1.9M (AI est.). Accounting for capital investment and migration costs, the net annual impact is estimated at ~$1M (AI est.).
X: Data Compression Efficiency
Y: High Image Quality Retention