Planar Intra-Prediction Coding for Video Blocks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video coding techniques, such as those in the H.264/AVC standard, face inefficiencies in encoding frames with smooth image content and gradual pixel value changes, leading to blocking artifacts at low to medium bitrates due to limitations in intra prediction modes like DC and directional modes.
Innovation Solution
A low complexity planar mode coding method is introduced, where the encoder calculates the value of the bottom-right pixel in a prediction block using boundary pixels and performs linear and bi-linear interpolations to derive values for remaining pixels, signaling residuals for decoding, and switches between primary and secondary transform kernels to optimize encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If DC mode and directional modes are used for intra prediction, then coding simplicity is maintained, but blocking artifacts increase and coding efficiency deteriorates for smooth image content
Solution Approach 1:
The patent introduces a planar prediction mode that changes the prediction parameters by using a gradient-based approach instead of constant DC or fixed directional modes. This allows the prediction to adapt to smooth variations in image content, reducing blocking artifacts while maintaining coding efficiency.
Solution Approach 2:
The patent makes the prediction mode dynamic by allowing the encoder to select between DC mode, directional modes, and planar mode based on the characteristics of the image block. This dynamic selection enables the system to adapt to different image content types, reducing blocking artifacts in smooth regions while maintaining simplicity in other regions.
2Productivity
If planar mode with gradient estimation is used, then coding efficiency improves for smooth images, but computational complexity increases
Solution Approach 1:
The patent simplifies the planar mode implementation by using a reduced gradient estimation approach that requires fewer computational operations while still capturing the essential smooth variations in image content. This reduces the computational complexity burden while maintaining the coding efficiency benefits.
Solution Approach 2:
The patent implements a simplified planar mode that performs partial gradient estimation rather than full estimation, applying the planar prediction only where it provides significant benefit. This partial application reduces computational complexity while maintaining coding efficiency for the most critical smooth image regions.
3Loss of substance
If AC coefficients are quantized to zero in blocks with gradual pixel changes, then compression ratio improves, but image quality deteriorates due to loss of detail
Solution Approach 1:
The patent applies planar prediction before transformation and quantization to pre-smooth the image content in a controlled manner. This preliminary action reduces the high-frequency variations that would otherwise be quantized to zero, allowing better preservation of important image details while achieving good compression ratios.
Solution Approach 2:
The patent changes the prediction parameters to use planar gradients that better match the smooth variations in image content. This results in smaller prediction errors that require fewer bits to encode, improving compression ratio without sacrificing image quality, as the important gradual changes are preserved rather than quantized to zero.
Data Source
AI summary
The present invention provides low complexity planar mode coding in which a value of a bottom-right pixel in a prediction block is calculated from a value of at least one pixel in at least one of an array of horizontal boundary pixels and an array of vertical boundary pixels. Linear and bi-linear interpolations on the value of the bottom-right pixel and values of at least some of the horizontal and vertical boundary pixels to derive values of remaining pixels in the prediction block. A residual between the prediction block and an original block is signaled to a decoder.