Interlaced Video Lapped Transform With Vertical-Edge Overlap Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The application of spatial-domain lapped transforms to interlaced video content leads to issues such as overflow, clipping, and reduced coding efficiency due to poor temporal continuity between scan lines, particularly when objects in motion cause jagged edges and discontinuities.
Innovation Solution
The interlace frame lapped transform is implemented with limited overlap filtering applied only to vertical edges within and between intra-coded macroblocks, and a conditional lapped transform is used, allowing spatial flexibility in the application of pre- and post-filters, with signaling based on macroblock granularity to optimize overlap filtering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If spatial-domain lapped transforms are applied to interlaced video content, then cross-block correlation is improved and blocking artifacts are reduced, but overflow and clipping occur due to poor temporal continuity between scan lines
Solution Approach 1:
The patent applies lapped transform selectively only to specific block edges (horizontal edges within intra-coded macroblocks and vertical edges between macroblocks) rather than all block edges uniformly. This local application approach maintains the benefits of reduced blocking artifacts where temporal continuity exists while avoiding overflow and clipping issues in regions with poor temporal continuity between interlaced scan lines.
2Manufacturing precision
If overlap filtering is applied to all block edges, then blocking artifacts are minimized, but coding efficiency is reduced due to increased complexity and unnecessary processing
Solution Approach 1:
The patent implements conditional lapped transform that applies overlap filtering only at specific block edges based on local image characteristics and macroblock type, rather than uniformly across all edges. This selective application reduces computational complexity and improves coding efficiency while maintaining artifact reduction benefits where needed.
Solution Approach 2:
The patent applies lapped transform to a subset of block edges (partial action) rather than all edges, specifically targeting horizontal edges within intra-coded macroblocks and vertical edges between macroblocks. This partial application avoids unnecessary processing overhead while sufficient to reduce visible blocking artifacts in critical regions.
3Productivity
If lapped transform is applied to interlaced video, then temporal continuity issues are exacerbated causing jagged edges and discontinuities, but the transform is needed for compression efficiency
Solution Approach 1:
The patent applies lapped transform selectively to specific macroblock types and edges where temporal continuity is sufficient, avoiding regions where interlacing causes significant temporal displacement. This selective application maintains compression efficiency benefits while preventing exacerbation of jagged edge artifacts in problematic regions.
Data Source
AI summary
A video encoder/decoder implements a lapped transform by applying an overlap filter in the spatial or transform domains to transform blocks. For interlace frames whose alternating scan lines are temporally displaced, the encoder/decoder imposes a limitation on application of the overlap filter to exclude horizontal block edges between adjacent transform blocks. This limitation can be imposed in both implementations where the overlap transform is applied across all blocks of an image, as well as implementations in which the lapped transform is conditionally applied on a spatially varying basis across the image.