Compressed Image Noise Removal via High-Frequency Component Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital compressed images suffer from blurring and mosquito noise due to high-frequency component dropout during quantization, leading to degraded image quality, especially in edge portions where quantization accuracy is compromised.
Innovation Solution
The solution involves using peripheral block data and frequency domain analysis to predict and restore lost high-frequency components, ensuring that noise removal occurs within the quantization width domain to maintain image fidelity and quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If quantization width is set largely for edge portions to reduce data amount, then compression efficiency is improved, but image quality deteriorates due to loss of high-frequency components
Solution Approach 1:
The patent applies preliminary action by predicting the high-frequency components before they are completely lost during quantization. The approximate image is generated in advance using low-frequency components and edge information, then the predicted high-frequency components are added back to the decoded image, effectively reversing the harmful effect of quantization before it fully degrades image quality.
Solution Approach 2:
The patent uses an intermediary approach by introducing an approximate image as a mediator between the decoded image (with lost high-frequency components) and the original image. This approximate image, generated from low-frequency components and edge information, serves as a bridge to reconstruct the lost high-frequency components and improve image quality without requiring the original data.
2Device complexity
If approximate image is created using only target block image information, then processing complexity is reduced, but accuracy of high-frequency component prediction deteriorates
Solution Approach 1:
The patent merges multiple information sources to create the approximate image: it combines low-frequency components from the decoded image, edge information from the decoded image, and peripheral block data. This merging of multiple data sources enriches the approximate image with sufficient information to accurately predict high-frequency components while maintaining reasonable processing complexity.
Solution Approach 2:
The patent applies dimensionality change by utilizing peripheral block data from adjacent blocks to infer information about the current block's high-frequency components. This spatial dimension extension allows the system to predict high-frequency content that would otherwise be completely lost, improving prediction accuracy without significantly increasing computational burden.
3Quantity of substance
If high-frequency components are completely removed during quantization, then data amount is significantly reduced, but image quality becomes severely degraded with noticeable artifacts
Solution Approach 1:
The patent converts the harmful effect of high-frequency component removal into a beneficial process. Instead of simply reversing quantization, it uses the approximate image generation process to create a corrected image that actually improves upon the decoded image by adding back predicted high-frequency components, thereby converting the harm of data reduction into a benefit of quality enhancement.
Data Source
AI summary
A compressed-image noise removal device includes a decoder unit for decoding a digital-image-compressed stream, an information holding unit for holding sub information by the amount of a plurality of blocks, the sub information being decoded by a VLD unit, a noise judgment unit for making a judgment on noise removal of a display image generated by the decoder unit, and the information holding unit, a noise removal unit for executing the noise removal of a block whose noise removal has been judged to be executed by the noise judgment unit, using image data outputted from an inverse quantization unit, motion compensation data outputted from a motion compensation unit, and the sub information held in the information holding unit, and a display-image holding unit for holding, as a display image, an output image of the noise removal unit if the noise removal has been judged to be executed by the noise judgment unit, or the output of the decoder unit if the noise removal has been judged not to be executed thereby.