Adaptive Image Filtering Based on Activity and Directivity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing demand for high-resolution and high-quality images in various fields leads to challenges in managing image data efficiently, as devices and network environments vary, necessitating a method to enhance image coding efficiency and subjective/objective image quality.
Innovation Solution
A method and device for adaptive filtering based on image properties, where the activity and directivity of a target region are calculated to select an appropriate filter from a set, allowing for variable region size, diagonal activity consideration, and flexible filter shapes, enhancing image quality and coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a single fixed filter is applied to all regions, then device complexity is reduced, but image quality deteriorates due to inability to adapt to different image properties
Solution Approach 1:
The patent implements dynamic filter selection by calculating image properties (activity and directivity) for each target region and adaptively choosing the most suitable filter from a set of candidate filters. This allows the filtering system to dynamically adjust to different image characteristics rather than using a static fixed filter, thereby improving image quality while maintaining reasonable system complexity through algorithmic adaptation.
Solution Approach 2:
The patent applies different filtering strategies to different regions of the image based on local image properties. By dividing the image into target regions and calculating activity and directivity for each region, the system applies appropriate filters locally rather than uniformly across the entire image, improving overall image quality while managing complexity through region-based processing.
2Manufacturing precision
If multiple filters are applied to different image regions, then image quality is improved, but processing time increases
Solution Approach 1:
The patent segments the image into target regions and processes each region independently based on its specific properties. By calculating image properties for each region and applying filters selectively, the system avoids unnecessary processing in regions that don't require filtering, thereby reducing overall processing time while maintaining image quality where needed.
Solution Approach 2:
The patent changes processing parameters (filter selection, filter strength) based on calculated image properties such as activity and directivity. This adaptive approach allows the system to apply stronger filtering only where necessary and use lighter or no filtering in other regions, optimizing the balance between image quality improvement and processing time reduction.
3Measurement precision
If activity and directivity calculation is performed for each sample, then filtering precision is improved, but computational complexity increases
Solution Approach 1:
The patent performs activity and directivity calculations for each sample within target regions, which represents a thorough (excessive) approach to property assessment. This enables precise filter selection for each sample, improving filtering precision. The computational complexity is managed by limiting calculations to only the necessary samples within defined target regions rather than the entire image.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A restoring picture filtering method performed by a decoding device, according to the present invention, comprises the steps of: calculating the activity of an image of a target region of a restoring picture; determining the directivity of the image of the target region; selecting a particular filter within a filter set on the basis of the activity and the directivity; and performing filtering on the restoring picture on the basis of the selected filter. According to the present invention, adaptive filtering may be applied on the basis of image properties of a target block of a restoring picture, and the objective/subjective image quality of the restoring picture may be enhanced therethrough.