Boundary-Aware Image Encoding with Recursive Sub-Block Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image encoding/decoding technologies face challenges in efficiently processing image boundaries and adjusting image sizes on a per block basis, leading to difficulties in compression efficiency and effective utilization of neighboring block information.
Innovation Solution
The method involves partitioning current images into blocks and sub-blocks, using quad tree or binary tree partitioning based on boundary conditions, and determining partition directions and types to optimize image encoding/decoding processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If image encoding is performed on a per block basis with fixed block sizes, then the encoding process is simplified and standardized, but it becomes difficult to efficiently process images with sizes that are not multiples of the block size and leads to poor compression efficiency at boundaries
Solution Approach 1:
The patent divides the image into multiple blocks of different sizes using quadtree and binary tree partitioning methods. This segmentation allows the encoding process to adapt to different image regions and boundary conditions, improving compression efficiency while maintaining standardized encoding procedures for each block type.
Solution Approach 2:
The patent introduces dynamic block partitioning where the image is divided into blocks of varying sizes rather than fixed sizes. The partitioning structure adapts to the image content and boundary requirements, allowing efficient processing of images with sizes that are not multiples of a standard block size while maintaining encoding standardization.
2Ease of operation
If the image size is adjusted to be a multiple of block size through padding, then per block encoding can be performed, but the additional processing steps and memory usage increase
Solution Approach 1:
Instead of padding the entire image to a multiple of block size, the patent segments the image into blocks of appropriate sizes using quadtree and binary tree partitioning. This allows per-block encoding without requiring global image size adjustment, reducing padding processing complexity while maintaining encoding capability.
Solution Approach 2:
The patent applies different block sizes and partitioning strategies to different regions of the image based on local requirements. Boundary regions may use different partitioning than interior regions, optimizing the balance between ease of operation and device complexity locally rather than globally.
3Manufacturing precision
If recursive partitioning is performed until no sub-block includes the boundary, then boundary processing is thorough and accurate, but the number of partitioning steps and computational complexity increase
Solution Approach 1:
The patent uses quadtree and binary tree partitioning to systematically divide blocks containing boundaries into smaller sub-blocks. This segmentation continues recursively until boundary conditions are satisfied, ensuring thorough and accurate boundary processing while providing a structured approach that manages computational complexity.
Solution Approach 2:
The patent performs preliminary partitioning to identify blocks that contain boundaries before detailed encoding. This preliminary action allows the system to prepare appropriate partitioning strategies in advance, ensuring boundary processing accuracy while organizing the computational workload to manage complexity.
4Productivity
If neighboring block information is utilized for current block encoding, then compression efficiency is improved, but the complexity of encoding and decoding processes increases
Solution Approach 1:
The patent utilizes neighboring block information as feedback for encoding the current block. By referencing already-encoded neighboring blocks, the system improves compression efficiency through predictive coding and mode copying while maintaining a structured feedback mechanism that manages the complexity of the encoding and decoding processes.
Data Source
AI summary
The present invention provides an image encoding method and an image decoding method. The image encoding method of the present invention comprises: a first dividing step of dividing a current image into a plurality of blocks; and a second dividing step of dividing, into a plurality of sub blocks, a block, which is to be divided and includes a boundary of the current image, among the plurality of blocks, wherein the second dividing step is recursively performed by setting a sub block including the boundary of the current images as the block to be divided, until the sub block including the boundary of the current image does not exist among the sub blocks.