Digital Video Signal Modification TRIZ Cases

A video coding device copies palette indices from multiple rows above to reduce redundancy in discrete-tone screen content.

A video signal processing method derives motion vectors for partitions using skip or direct mode indication information to predict pixel values.

Selective in-loop filtering at discontinuous virtual boundaries reduces face seam artifacts while preserving compression efficiency within continuous regions.

Adjusting decoding sequence allows using lower right reconstructed pixels as references, improving accuracy while managing complexity.

Selective bilateral filtering of video block samples enables parallel processing of neighboring blocks by refraining from filtering boundary-dependent data.

A video decoding method determines transform types for target blocks using size parameters and sets zero-out regions for inverse transformation.

Template-based adaptive weighting adjusts reference block weights using luminance characteristics to improve prediction accuracy.

Template matching determines motion vector sign pairs to reduce bitstream size and computational complexity in HEVC.

Deriving chrominance quantization parameters from luminance and neighboring block values reduces signaling overhead while maintaining picture quality.

A constrained inter-image prediction mode reduces candidate block positions and reference images to lower computational complexity.

Scattering pixels via geometric transforms recovers lost data from adjacent pixels, reducing network resource waste compared to full stream replication.

Dynamic deblocking filter adjustments inject comfort noise to prevent artificial stillness while maintaining standard decoder compatibility.

Enhancement encoder overwrites older temporal predictions with base stream data to optimize memory usage.

A variable localized bit-depth increase adjusts transform input and output signals to mitigate rounding errors in fixed-point video coding operations.

Position-dependent weighting in bi-prediction reduces computational complexity while improving prediction accuracy.

Dynamic mode selection adapts prediction strategies for mixed content, improving rate-distortion tradeoffs.

Partitioning 360 video into regions allows the computing device to allocate bitrate based on pixel count and distortion, resolving storage constraints.

Wavelet coefficients packed in 2D blocks encode mesh displacements relative to a base mesh.

A video decoding apparatus determines split flag context models using reference coding tree unit data.

Constraining spatial upsampling model parameters reduces coding bits while maintaining machine vision task accuracy.

A video quality assessment network splits images into blocks to determine sensitivity and quality scores.

Infers Local Illumination Compensation flags from neighbors to cut signaling costs while preserving prediction accuracy for affine coded blocks.

Embeds random access picture identifiers in non-random access frames to detect lost synchronization markers and trigger decoder cache clearing.

Scaling parameters in the encoding loop adjust luminance dynamic range, aligning quantization noise with human visual perception.

Applying a bilateral filter with distance-based coefficients to non-adjacent reference samples improves compression efficiency for non-square blocks.

Parsing transform coefficients identifies non-zero values to select a single matrix index, minimizing storage requirements and boosting coding throughput.

A video streaming system selects quality levels using pre-calculated bit rate data to maximize delivery speed.

A video codec selects reference frames by matching motion vectors to intra-coded areas.

A video interface transfers data using a lower chroma subsampling ratio to free bandwidth for non-video information.

A bit estimation process predicts macroblock sizes during video encoding to determine prediction modes.

Segmenting encoded streams into truncated data and signaling information reduces transmission congestion for high dynamic range video content.

Applying combined neural network filters to video units improves coding efficiency while managing device complexity and computational requirements.

Determining quantization parameters based on projected frame distortions to resolve visual discontinuities at seams in virtual reality environments.

Sharing partial inverse matrix results across clipping settings reduces multiplication operations in adaptive loop filter coefficient derivation.

A flexible transform tree structure partitions coding units into transform units using a quadtree with two-child nodes.

Multivariate multiple regression generates chroma mappings to avoid letterbox brightening and block vanishing artifacts in backward compatible video encoding.

A neural network video encoder uses a supervisory signal to control encoding operations for task-specific feature selection.

A convolutional neural network updates image features and superposes them with primary data to generate prediction images.

A gradual decoder refresh frame partitions video data into instantaneous decoder refresh and inter-coded regions to synchronize encoder and decoder operations.

Segmenting video frames into independent regions eliminates serial macroblock dependencies, reducing processing time and power consumption.

Quantizing pixel string lengths to multiples of a string length resolution improves decoding efficiency by enabling memory alignment.

A video decoder adapts prediction mode lists using distance measurements between predictor blocks to optimize signaling efficiency.

Weighted downsampling and transformations split video encoding into baseline and enhancement layers to resolve quality versus bitrate trade-offs.

Priority-based selection of spatial and temporal motion vector candidates reduces residual data transmission while maintaining high prediction accuracy.

Parallel GPU bit shifts and CPU packing resolve serial processing bottlenecks, achieving 10x throughput for variable length data.

A decoder derives a new luma intra prediction mode index to enable matrix-based intra prediction for luma signals.

Predictive coding of most significant bits for long-term reference picture order count values in video decoders.

Adjusting chroma quantization parameters at the coding unit level improves compression performance while managing bandwidth usage.

Deriving motion vector ranges from block size and center position prevents clipping near picture boundaries, preserving interpolation accuracy.