Image Decoding Apparatus Adaptive Wiener Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image coding and decoding techniques, such as hybrid coding, introduce quantization noise and block distortions, leading to inferior image quality in decoded images, which existing adaptive filtering methods cannot adequately address without increasing bit rates and storage or transmission requirements.
Innovation Solution
An image coding and decoding apparatus that generates and inserts cross-correlation data into the coded stream to compute filter parameters for a Wiener filter, allowing for adaptive filtering to restore the image quality of decoded images by minimizing differences between the decoded and original images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If hybrid coding technique is used to compress video data, then data volume is reduced, but image quality is impaired due to quantization noise and block distortions
Solution Approach 1:
The system performs preliminary filtering of reference images before they are used in motion compensation. By filtering reference images in advance and storing the filtered versions, the system prevents quantization noise and block distortions from propagating through subsequent frames, thereby maintaining image quality while preserving compression efficiency
Solution Approach 2:
The system implements a feedback mechanism where decoded images are fed back into the filtering process. The filtered decoded images are then used as reference images for future motion compensation, creating a closed-loop system that continuously improves image quality while maintaining compression benefits
2Manufacturing precision
If adaptive filtering is applied to decoded images, then image quality is improved, but bit rate and storage requirements increase
Solution Approach 1:
The system performs filtering operations on reference images before they are used in motion compensation, rather than filtering decoded images after the fact. This preliminary filtering approach reduces the bit rate requirement because the filtering is performed on compressed reference images rather than on full-resolution decoded images
Solution Approach 2:
The system applies different filtering strategies to different regions of the image based on their characteristics. By analyzing the local properties of reference images and applying appropriate filters selectively, the system maintains high image quality in important regions while minimizing the overall bit rate consumption
3Object-affected harmful factors
If de-blocking filter is used to remove block distortions, then block boundaries are smoothed, but other types of noise and image quality issues remain
Solution Approach 1:
The system changes the filtering parameters and methods based on the specific characteristics of the reference images and the type of distortion present. By adapting filter coefficients and filtering strategies to the local image properties, the system effectively removes block distortions while also addressing other types of noise and quality issues
Solution Approach 2:
The system combines multiple filtering techniques and processing steps into a composite filtering system. By integrating de-blocking filtering with other filtering operations and processing stages, the system achieves comprehensive image quality improvement that addresses multiple types of distortions and noise simultaneously
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provide is an image decoding apparatus which reliably prevents deterioration of the image quality of decoded images which have been previously coded. An image decoding apparatus (200) includes: an inverse quantization and inverse orthogonal transform unit (220) and an adder (230) which decode a coded image included in a coded stream (Str) to generate a decoded image (Rc); an entropy decoding unit (210) which extracts cross-correlation data (p) which indicates a cross-correlation between the decoded image (Rc) and an image which corresponds to the decoded image and has not yet been coded; and an adaptive filter (240) which computes a filter parameter (w) based on the extracted cross-correlation data (p), and performs a filtering operation on the decoded image (Rc) according to the filter parameter (w).