Scalable Encoding Adaptive Code Allocation Residual Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In scalable encoding, the fixed allocation of codes to each layer leads to significant encoding distortion and poor image quality, especially when data is concentrated in lower layers, as independent encoding of layers fails to restore quality even with additional data from higher layers.
Innovation Solution
Adaptive allocation of codes to each layer based on the difference in data amounts, where the residual signal of a lower layer is encoded using part of the codes allocated to a higher layer, minimizing encoding distortions and revising code allocation to improve image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the amount of codes allocated to each layer is fixed before encoding, then the encoding process is simple and efficient, but the image quality degrades significantly when data is concentrated in lower layers
Solution Approach 1:
The patent implements dynamic code allocation by introducing a feedback mechanism where the encoder first performs preliminary encoding with initial code allocation, evaluates the residual signals, and then redistributes codes adaptively. The code allocation amount for each layer is dynamically adjusted based on the actual residual energy and reconstruction error, transforming the static fixed-allocation system into a dynamic adaptive system that optimizes image quality while maintaining encoding efficiency.
Solution Approach 2:
The patent employs feedback control by encoding each layer sequentially, evaluating the residual signal quality after each layer's encoding, and using this evaluation to determine the optimal code allocation for subsequent layers. The feedback loop compares the actual reconstruction error against quality thresholds and adjusts code allocation accordingly, ensuring that layers with higher residual energy receive more codes to minimize overall distortion.
2Device complexity
If independent encoding is performed for each layer with fixed code allocation, then the encoding process is straightforward, but adding higher layer data cannot restore the quality of degraded lower layers
Solution Approach 1:
The patent applies preliminary action by performing preliminary encoding of all layers with initial code allocation before final code redistribution. This preliminary encoding phase allows the system to evaluate the residual signals and determine the actual information content of each layer, which then guides the optimal code allocation. By preparing the encoding structure in advance and then adapting it based on actual signal characteristics, the system ensures that lower layers receive sufficient codes while maintaining the straightforward sequential encoding structure.
3Manufacturing precision
If sufficient codes are allocated to lower layers, then encoding distortion is minimized, but the total code amount increases beyond fixed budget constraints
Solution Approach 1:
The patent changes the parameter of code allocation from fixed to adaptive by introducing a code redistribution mechanism. The system calculates the optimal code allocation amount for each layer based on the residual signal energy and reconstruction error, dynamically adjusting the code allocation parameters to minimize total distortion within the fixed code budget. This allows lower layers to receive more codes when needed while higher layers receive fewer codes, optimizing the overall distortion-rate performance.
Data Source
AI summary
A scalable encoding method for decomposing an image signal into signals assigned to layers so as to encode the image signal includes encoding an image signal of at least one target layer among the layers based on an amount of codes allocated to the target layer; outputting a residual signal of the target layer corresponding to a difference between the pre-encoded image signal and an image signal decoded from the encoded image signal; determining a target amount of codes allocated for encoding the residual signal based on the residual signal, an amount of codes allocated to an objective layer, and an image signal of the objective layer; encoding the residual signal based on the determined target amount of codes; revising the amount of codes allocated to the objective layer based on the target amount of codes; and encoding the image signal of the objective layer based on the revised amount.