OSD Image Processing Extended Block Weight Blending
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing methods for on-screen display (OSD) imaging areas suffer from halo phenomena and boundary breaks due to incorrect classification of blocks as OSD or non-OSD, leading to fly-out issues during motion estimation and interpolation.
Innovation Solution
The proposed OSD image processing method introduces an extended OSD block classification, which adjusts weight values for blending motion compensation and zero-motion data to reduce halo phenomena and boundary breaks, by defining specific thresholds for pixel counts and weight adjustments in interpolated blocks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If blocks at the boundary of the OSD imaging area are determined as OSD blocks, then the OSD imaging area is preserved, but serious halo phenomena are caused
Solution Approach 1:
The patent segments the block classification into three distinct types: OSD blocks, non-OSD blocks, and extended OSD blocks. This segmentation allows boundary blocks to be classified as extended OSD blocks (containing both OSD and non-OSD data) rather than forcing them into binary OSD/non-OSD categories, thereby preserving imaging area integrity while reducing halo phenomena through differentiated processing
Solution Approach 2:
The patent applies local quality by treating different block types with different processing rules. Extended OSD blocks at boundaries receive specialized handling with adjusted weight values in motion compensation, while pure OSD blocks maintain standard processing. This localized differentiation resolves the contradiction by applying appropriate processing quality to each block type based on its characteristics
2Object-affected harmful factors
If blocks at the boundary of the OSD imaging area are determined as non-OSD blocks, then halo phenomena are reduced, but broken OSD imaging area is caused
Solution Approach 1:
The introduction of extended OSD blocks as a separate category prevents boundary blocks from being misclassified as non-OSD blocks. By segmenting the classification system to explicitly recognize mixed-content blocks, the patent maintains imaging area integrity while avoiding the halo phenomena associated with treating boundary blocks as pure non-OSD blocks
3Manufacturing precision
If the third weight is increased for boundary blocks using two motion vectors, then the interpolated frame quality is improved, but serious halo phenomena are caused
Solution Approach 1:
The patent applies local quality by implementing weight adjustment rules specific to extended OSD blocks. When a block is classified as extended OSD, the third weight (zero-motion weight) is adjusted differently compared to pure OSD or non-OSD blocks. This localized weight control maintains interpolated frame quality while preventing halo phenomena at OSD boundaries
4Reliability
If the weight corresponding to the referred block of the current frame is increased, then the fly-out phenomenon is avoided, but the block of the interpolated frame flies out
Solution Approach 1:
The patent implements local quality through block-type-specific weight adjustment rules. For extended OSD blocks, the weights are adjusted according to the block's mixed characteristics rather than applying uniform weight rules. This ensures that weight adjustments stabilize blocks locally without causing fly-out artifacts, as each block type receives appropriate weight treatment based on its OSD content proportion
Data Source
AI summary
In an on-screen display (OSD) image processing method for generating an interpolated frame with interpolated blocks, motion compensation is performed to generate motion compensation (MC) data according to a first motion vector referring to a previous frame and a second motion vector referring to a current frame. Zero-motion data are generated according to a zero-motion vector referring to the previous frame and the current frame. The MC data of a first weight and the zero-motion data of a second weight are blended to generate each interpolated block as a processed block. When the number of pixels having OSD data in the processed block is less than a first positive threshold and larger than a second positive threshold, the processed block is determined as an extended OSD block and the second weight of the extended OSD block is less than that of an OSD block.