Image Encoder Block-Level Quantization for Compression Quality
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The challenge is to maintain image quality during lossy compression of high-resolution video images, as existing methods often result in increased quantization errors and quality deterioration, especially in image blocks with higher spatial complexity, leading to reduced processing speed and potential error propagation across multimedia IPs.
Innovation Solution
An image processing device and method that perform lossy compression based on a target compression ratio and a constraint quantization step size, adjusting the quantization step to prevent excessive error increase, thereby maintaining image quality by selectively applying quantization and compression processing to image blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If lossy compression with quantization is applied to increase compression ratio, then data size is reduced, but quantization error increases and image quality deteriorates
Solution Approach 1:
The patent applies different quantization strategies to different image blocks based on their spatial complexity. Blocks with high spatial complexity use a constrained quantization step size to preserve quality, while blocks with low spatial complexity can use larger quantization steps to achieve higher compression ratios. This local differentiation resolves the contradiction by allowing quality preservation where needed while maximizing compression where tolerable.
Solution Approach 2:
The patent dynamically adjusts the quantization step size parameter based on the spatial complexity of each image block. By changing this critical parameter locally rather than uniformly, the system achieves both high compression ratios in appropriate areas and quality preservation in areas where it matters most, thus resolving the contradiction between compression ratio and image quality.
2Productivity
If quantization step size is increased to achieve higher compression ratio, then data size is reduced, but quantization error increases and quality deteriorates
Solution Approach 1:
The patent implements local quality control by assessing spatial complexity of each block and applying appropriate quantization strength. This allows the system to achieve high overall compression ratios while preventing quality deterioration in blocks where it would be most noticeable, thus resolving the productivity-precision contradiction.
Solution Approach 2:
The patent applies partial compression action selectively - using strong quantization (excessive compression) only where spatial complexity allows, and gentle or no quantization where quality must be preserved. This partial application of compression resolves the contradiction by avoiding excessive action in inappropriate areas while achieving sufficient compression overall.
3Quantity of substance
If compression processing is applied to all blocks, then overall compression ratio increases, but processing time and complexity increase
Solution Approach 1:
The patent segments the image into blocks and applies different compression processing to different segments based on their spatial complexity. This segmentation allows the system to avoid unnecessary complex processing in blocks that don't require it, thus reducing overall processing complexity while still achieving good compression ratios where needed.
Solution Approach 2:
The patent applies compression processing partially rather than universally - skipping or using simplified processing for blocks with low spatial complexity while applying full processing only where necessary. This partial action approach reduces processing complexity and time while still achieving the desired overall compression ratio.
Data Source
AI summary
An image processing device and an operating method of the image processing device are disclosed. The image processing device includes a multimedia intellectual property (IP) configured to generate second image data by performing image processing on first image data and an encoder configured to compress the second image data, the encoder is further configured to generate first compressed data by performing quantization and at least one compression processing on a first block of the second image data based on a target compression ratio in a first compression condition, and generate second compressed data by performing the quantization on a second block of the second image data based on a first quantization step in a second compression condition, without performing the at least one compression processing.