Spatial Prediction Compensation for Real-Time Image Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional spatial prediction in image data encoding and decoding is not suitable for real-time processing and leads to encoding and decoding delays, as well as error diffusion and display quality degradation, especially when dealing with images containing edges.
Innovation Solution
A system and method for spatial prediction that uses neighboring blocks in a row above the current block to predict pixel values, replacing pixel values every predetermined row unit with reference values to enable real-time encoding and decoding, and prevent error diffusion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional spatial prediction using pixel values of blocks on the left of the current block is used, then spatial prediction can be performed, but real-time encoding and decoding cannot be achieved due to encoding and decoding delays
Solution Approach 1:
The patent applies preliminary action by pre-defining the spatial prediction structure using pixel values from blocks in the row above the current block, rather than requiring sequential processing from left to right. This allows the spatial prediction to be prepared in advance and executed without waiting for previous block processing, thereby enabling real-time encoding and decoding while eliminating the delays associated with conventional sequential spatial prediction.
2Reliability
If conventional spatial prediction is used, then spatial prediction can be performed, but error diffusion and display quality degradation occur when images have edges
Solution Approach 1:
The patent applies segmentation by dividing the spatial prediction process into distinct components: using pixel values from blocks in the row above the current block for prediction, while separately handling edge cases. This segmentation allows the system to perform accurate spatial prediction in most regions while isolating edge regions that may cause error diffusion, thereby maintaining overall prediction accuracy without propagating errors across the entire image.
3Productivity
If pipeline processing is implemented with conventional spatial prediction, then processing can be continuous, but spatial prediction cannot be performed using pixel values of blocks on the left due to processing sequence constraints
Solution Approach 1:
The patent applies preliminary action by pre-establishing the spatial prediction mechanism using blocks from the row above the current block, which are already available in the pipeline before the current block is processed. This allows continuous pipeline processing to proceed without interruption, as the spatial prediction can be performed using pre-available data rather than requiring sequential left-to-right processing that would break the pipeline continuity.
Data Source
AI summary
An encoding and/or decoding system, medium, and method with spatial prediction and spatial prediction compensation. The system may include a spatial prediction unit that spatially predicts pixel values of a current block of an image using neighboring blocks in a row immediately above the current block among neighboring blocks that are spatially adjacent to the current block. The spatial prediction unit performs spatial prediction using replaced pixel values of neighboring blocks in a row immediately above the current block every predetermined row unit, with the pixel values of neighboring blocks in a row immediately above the current block every predetermined row unit being replaced with a predetermined reference value.