The relentless demand for higher resolution video across all sectors, from entertainment to enterprise, is pushing existing infrastructure to its limits. Companies face escalating costs for data storage and transmission, alongside increasing pressure to meet ESG goals by reducing data center energy consumption. This technology offers a strategic advantage by enabling significant data compression, allowing businesses to deliver superior content experiences, optimize operational expenditures, and align with global sustainability mandates, all while staying ahead of competitors.
Reduces data volume by up to ~30% by optimizing prediction residuals through pixel-level similarity evaluation and multiple reference images.
Maintains high visual quality by selectively applying orthogonal transformation and quantization to prediction residual regions, minimizing degradation under high compression.
Reduces processing load by limiting transformation and quantization to partial prediction residual regions, improving computational efficiency over full-region methods.
This patent provides robust protection for an image encoding device, image decoding device, and program, covering multiple aspects of the technology. Its rapid grant within 14 months and successful overcoming of a single office action with precise arguments and amendments indicate clear novelty, inventiveness, and a strong, stable scope of claims with low invalidation risk.
This patent primarily covers the core encoding algorithm. White space exists in hardware-accelerated implementations, integration with specific content delivery network (CDN) architectures, or advanced AI-driven content analysis layers that leverage the compressed data.
Assuming an average 25% reduction in video data volume. For an enterprise processing 100 PB of video data annually, combining storage costs (average ~$0.013/GB/year (AI est.)) and network bandwidth costs (average ~$0.033/GB (AI est.)), the estimated reduction is (100 PB × 0.25) × (~$0.013/GB + ~$0.033/GB) = 25 PB × ~$0.046/GB = ~$1.15M (AI est.). The direct data-related cost reduction, excluding capital investment and operational improvements, is estimated at ~$1.5M per year (AI est.).
X: Video Data Compression Efficiency
Y: High Image Quality Retention