Pipelined Video Encoding Slice Size Control Using Reserved Bit Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-stage pipelined video encoding, the calculation of slice sizes is inaccurate due to the exclusion of coding units in the pipeline, leading to inefficient network packet usage and potential congestion, as well as visual quality degradation from packet loss.
Innovation Solution
A system and method for predicting reserved bits per coding unit using a fitted function, such as linear, logarithmic, or polynomial, to accurately determine slice sizes, minimizing packet generation and ensuring seamless decoding even with packet loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If slice size control is implemented in multi-stage pipelined video encoding, then network packet loss recovery is improved and visual quality is maintained, but coding complexity increases and pipeline timing becomes difficult to manage
Solution Approach 1:
The patent applies preliminary action by pre-calculating reserved bits for coding units that are currently in the encoding pipeline before they complete encoding. This allows the slice size control to be performed in advance based on predicted bit counts, enabling accurate slice header insertion timing without waiting for actual encoding completion. The reserved bit calculation uses a formula based on quantization parameters and block sizes to estimate future bit consumption, thus resolving the timing conflict between pipeline processing and slice size control.
2Reliability
If reserved bits are overestimated in slice size calculation, then packet loss recovery is improved, but network bandwidth is wasted and transmission efficiency decreases
Solution Approach 1:
The patent applies parameter changes by dynamically adjusting the reserved bit calculation based on actual encoding parameters such as quantization step size, block dimensions, and encoding mode. Instead of using a fixed conservative estimate, the system calculates reserved bits using the formula: reserved_bits = f(QP, block_width, block_height, encoding_mode), where QP is quantization parameter. This adaptive approach ensures accurate reserved bit estimation that matches actual encoding consumption, preventing both overestimation (bandwidth waste) and underestimation (packet loss vulnerability).
3Productivity
If coding units in the pipeline are excluded from slice size calculation, then pipeline processing speed is maintained, but slice size accuracy deteriorates and packet generation becomes inefficient
Solution Approach 1:
The patent applies feedback by using actual encoding parameters (quantization step, block size, mode decision) from coding units currently in the pipeline to update and refine the reserved bit calculation. As coding units progress through the pipeline stages, their parameter information feeds back into the slice size control mechanism, allowing continuous refinement of the reserved bit estimate. This feedback loop ensures that even though coding units are processed in parallel through the pipeline, their contribution to slice size is accurately accounted for, maintaining both processing speed and measurement precision.
Data Source
AI summary
A technique to determine an optimum number of reserved bits for use in video encoding. The bits per coding unit is calculated as a function of bitrate, frame rate and frame size. Then for each coding unit, N reserved bits candidates are selected. For each test video the number of packets per frame is identified and the average number of packets per frame is calculated. The best reserved bits candidate out of the N reserved bits candidates is identified, and then a best fitted linear model between best reserved bits candidate and bits per coding unit is determined.