Embedded Codec Circuitry for Multiple Reconstruction Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image compression techniques suffer from increased quantization error and inefficiency due to the use of a single reconstruction point for all image blocks, leading to visible artifacts and sub-optimal memory usage in on-chip codecs, especially when pixel values span different numeric ranges within a quantization bin.
Innovation Solution
An embedded codec circuitry that selects an optimal reconstruction point for each image block based on the quantization error between input and locally decoded pixel values, allowing for multiple reconstruction points to minimize overall quantization error and improve image quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single middle value reconstruction point is used for all image blocks, then the codec structure is simple and easy to implement, but the quantization error increases significantly when pixel values lie in different numeric ranges within a quantization bin
Solution Approach 1:
The patent divides the single reconstruction point approach into multiple reconstruction points (first reconstruction point and second reconstruction point) corresponding to different numeric ranges within a quantization bin. This segmentation allows each range to have its own optimized reconstruction value, reducing quantization error for image blocks with pixel values in specific ranges while maintaining overall system functionality.
Solution Approach 2:
The patent applies different reconstruction points based on the local characteristics of image blocks. Specifically, a first reconstruction point is used for image blocks where pixel values lie in a first numeric range, and a second reconstruction point is used for image blocks where pixel values lie in a second numeric range. This local adaptation optimizes reconstruction accuracy for each specific case without requiring complex global optimization.
2Measurement precision
If multiple reconstruction points are implemented with different numeric ranges, then quantization error is reduced and image quality improves, but the device complexity and memory usage increase
Solution Approach 1:
The patent changes the reconstruction point parameter based on the numeric range of pixel values in each image block. By selecting between a first reconstruction point and a second reconstruction point according to whether pixel values fall in a first or second numeric range, the system adapts the reconstruction parameter to match the local data characteristics, improving accuracy without requiring a completely new reconstruction mechanism.
Solution Approach 2:
The patent introduces dynamic selection of reconstruction points based on the input data characteristics. The system dynamically determines which reconstruction point to use by evaluating the numeric range of pixel values in each image block, allowing the reconstruction mechanism to adapt flexibly to different input scenarios while maintaining a relatively simple overall structure.
3Productivity
If multiple reconstruction points are used for different image blocks, then compression efficiency improves through better quantization, but on-chip memory usage increases
Solution Approach 1:
The patent implements a selective approach where only necessary reconstruction points are used based on the actual data distribution. Rather than implementing all possible reconstruction points for every scenario, the system uses a first reconstruction point for image blocks in a first numeric range and a second reconstruction point for image blocks in a second numeric range, applying the additional complexity only where it provides actual benefit.
Data Source
AI summary
An embedded codec (EBC) circuitry includes encoder circuitry to encode an image block by application of a sequential encoding scheme to generate an encoded image block. The encoder circuitry determines a local pixel value of each pixel of the plurality of pixels in the encoded image block, based on a quantization bin size of each pixel in the encoded image block. The encoder circuitry selects a reconstruction point from a plurality of reconstruction points based on an accumulated difference of the local pixel value and an input pixel value of each pixel in the encoded image block. The encoder circuitry allocates a set of signaling bits that indicates the selected reconstruction point in a bit-stream of the encoded image block. The EBC circuitry further includes decoder circuitry to reconstruct the input pixel value of each pixel in the encoded image block, based on the set of signaling bits.