Digital Video Signal Modification TRIZ Cases

Variable IBC predictor list sizing improves block vector prediction, balancing decoding efficiency with compression performance.

Tile-group extraction and picture-type parameter setting enable faster random access to high-attention video regions with lower coding overhead.

Maps interpolation filter types to video block transform kernels to improve decoding accuracy and reduce separate codeword signaling.

A segmented motion line buffer uses CTU-aware access and division factors to cut storage needs without corrupting motion prediction.

A mixed NAL type flag lets decoders handle IRAP and non-IRAP sub-pictures in one picture, supporting VR resolution changes with lower resource use.

Deriving chroma scaling from luma mapping codewords and an offset removes luma averaging delay while preserving decoding efficiency.

Parallel merge regions let decoders build spatial merge candidates from fixed neighbors, cutting complexity while preserving motion prediction efficiency.

Adaptive coefficient ranges based on bit depth preserve arithmetic precision in LFNST-based image encoding without unnecessary hardware cost.

Joint Cb/Cr residual coding uses mode-specific chroma QP offsets to improve compression while limiting picture-quality loss and decoder overhead.

A unified DCT matrix handles multiple transform block sizes to cut video coding complexity, memory use, and hardware implementation cost.

Combining DIMD with intra prediction subpartitions cuts mode signaling overhead while improving video coding efficiency.

Multiple scaling factors in block adaptive weighted prediction compensate local illumination changes to improve video coding efficiency.

Switching between spatial and phase prediction in 2×2 OCL RAW images suppresses MPC encoding loss caused by light-direction pixel deviations.

Stores image quality improvement metadata in the content file so clients can avoid VVC parsing and reproduce high-quality video at lower bit rates.

Applies deblocking to implicit TU boundaries in VVC to reduce block discontinuities and improve subjective video quality.

Position-based quantization in LFNST coding preserves key low-frequency coefficients to improve video compression while maintaining image quality.

When neighboring samples are limited, preset values replace model fitting, reducing VVC color prediction complexity without hurting performance.

Residual transform blocks are coded with ordered parity flags and fewer context-coded flags to improve image compression efficiency.

Non-power-of-two border blocks are coded as zero-units, skipping transforms to cut decoding overhead while preserving coding completeness.

Changing subblock coding order improves reference sample use for intra prediction, raising video coding efficiency and quality.

Reducing context-coded CABAC bins and switching later bins to bypass mode raises video codec throughput while preserving coding efficiency.

Adaptive color space transforms and bounded quantization improve residual coding efficiency for high-quality image transmission and storage.

Ordered scan-path context selection improves lossless image entropy coding by using prior non-adjacent symbols to build a more compact bitstream.

Bitstream-signaled normalization lets decoders apply the right feature scaling for neural network input, improving compression efficiency and accuracy.

Root blocks share neighboring reconstructed samples across small blocks, boosting intra prediction throughput and reducing repeated criteria checks.

Boundary-pixel selection based on adjacent block prediction modes improves intra prediction accuracy and encoding efficiency.

Adjacent PU motion vector correlation cuts HEVC merge MV rate-distortion checks, speeding inter-frame coding with lower computation.

Weighted combined inter/intra prediction improves video compression efficiency and image quality for larger, higher-resolution video data.

Adaptive reference line selection uses one line at largest coding unit boundaries and N lines elsewhere to cut video codec buffer size.

Stores reference pictures at different native resolutions so video coding can adapt to bandwidth changes without upscaling complexity or quality loss.

Recursive splitting of boundary blocks lets block-based encoding handle non-multiple image sizes while improving compression and neighbor use.

Reordering IBC prediction subblocks by split direction and ratio improves video coding efficiency and image quality with manageable complexity.

A guided filter applied to reconstructed video samples reduces ringing artifacts while smoothing transitions and preserving edge details.

Boundary-strength adaptation filters inter-predicted transform block edges to reduce visible artifacts and preserve coding efficiency.

Syntax flags for text description purpose, ID, and persistence improve video decoding accuracy while reducing redundant signaling.

Selective zeroing of transform coefficients and region flags cuts bit usage and improves video compression efficiency.

Adaptive switching between in-loop LMCS and out-loop reshaping improves HDR encoding efficiency and decoded picture quality at the same code rate.

Multiple candidate templates and block vectors improve Intra TMP prediction accuracy while keeping decoder complexity low.

Spatial resampling flags and filters cut video data volume while preserving usable quality for human viewing and machine vision tasks.

Using the NAL unit as the layer identifier in single-layer bitstreams cuts signaling overhead and improves compression efficiency.

Unit horizontal and vertical microlens intervals correct blocked light-ray motion vectors, improving plenoptic video encoding efficiency and image quality.

Recursive sub-block partitioning handles image edges more precisely while preserving neighboring block use for efficient encoding and decoding.

Asymmetric in-loop filtering uses padding or fixed values at virtual boundaries to prevent decoder leaks while preserving filtering accuracy.

Classifying adaptive loop filters by image dynamic range and intermediate data improves video quality without excessive encoding or decoding complexity.

Skipping redundant non-VCL NAL count data cuts media file overhead while preserving correct sample reconstruction and adaptive streaming.

Recursive sub-block splitting targets boundary blocks to handle non-multiple image sizes while improving compression and neighboring block use.

Boundary pixel prediction is corrected with adjacent block motion differences, improving inter-frame video coding accuracy and efficiency.

Recursive sub-block splitting handles image boundaries without padding, improving compression efficiency and neighboring block utilization.

Weighted refinement from a filtered dominant chroma component improves image quality while limiting bandwidth and processing overhead.

Candidate block vectors and a filtering flag improve predicted block accuracy in high-definition video decoding and compression.