Video Signal Processing Device Using Inverted Filter Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video signal processing methods fail to accurately align the positions of brightness and chroma pixels with standard formats and match the frequency characteristics between the top and bottom fields during sub-band division of interlace images, particularly in 4:2:0 format, leading to incomplete scanning line position structures and mismatched frequency characteristics.
Innovation Solution
A video signal processing device that uses vertically inverted filter coefficients for low-frequency and high-frequency decomposition and composition filters to ensure perfect reconstruction filter bank conditions, aligning pixel positions and matching frequency characteristics between the top and bottom fields by sampling down and up-sampling signals in a way that maintains the 1:3 and 3:1 ratios for pixel division.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional sub-band filtering with linear phase filters (odd or even taps) is applied to interlace images, then frequency division is achieved, but pixel position alignment with standard formats and frequency characteristic matching between top and bottom fields deteriorates
Solution Approach 1:
The patent applies vertically inverted filter coefficients to the second field processing. Specifically, if the first field uses filter coefficients [a0, a1, a2, ..., an], the second field uses [an, an-1, ..., a0]. This inversion approach resolves the pixel position alignment issue and frequency characteristic mismatch between top and bottom fields while maintaining standard format compliance.
Solution Approach 2:
The patent applies different filter coefficient configurations to different fields: the first field uses conventional linear phase filters, while the second field uses vertically inverted filter coefficients. This local differentiation allows each field to be processed optimally for its specific position in the interlace structure, achieving both pixel alignment and frequency matching.
2Measurement precision
If sub-band division is performed on interlace images, then frequency band separation is achieved, but frequency characteristic matching between top and bottom fields deteriorates
Solution Approach 1:
By vertically inverting the filter coefficients for the second field, the patent ensures that the frequency characteristics of the decomposed signals from both fields are properly aligned. This inversion compensates for the phase differences introduced by the interlace structure, enabling accurate frequency characteristic matching while maintaining signal reconstruction fidelity through perfect reconstruction filter bank conditions.
3Manufacturing precision
If conventional filtering is applied to align pixel positions, then position alignment is improved, but frequency characteristic matching between fields deteriorates
Solution Approach 1:
The patent implements field-specific filter coefficient configurations where the first field uses standard linear phase filters for position alignment, while the second field uses vertically inverted coefficients. This local quality approach ensures that both position alignment and frequency characteristic matching are optimized for their respective fields simultaneously, resolving the contradiction between these two requirements.
Data Source
Figure 1
Figure 2A~2B
Figure 3A~3B
AI summary
An decomposition filter(670) includes a top-field low-frequency decomposition filter (610) and a bottom-field low-frequency decomposition filter (611) whose coefficient is obtained by vertically inverting the coefficient of the top-field low-frequency decomposition filter. It also includes a top-field high- frequency decomposition filter (620) and a bottom-field high-frequency decomposition filter (621) whose coefficient is obtained by vertically inverting the coefficient of the top-field high-frequency decomposition filter. The above-described filters are switched and used according to which an input signal is for, a top-field or a bottom-field.