Edge-Preserving Vertical Interpolation for De-interlaced Video
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vertical interpolation techniques in video processing fail to effectively preserve edge information, leading to poor picture quality, especially in the presence of noisy data, and often produce false edges or miss edges in scenes with diagonal or round objects.
Innovation Solution
A method that computes horizontal and vertical intensity gradients for a pixel window to determine edge direction, using a confidence factor to limit interpolation when noise is detected, allowing for more accurate edge detection and improved picture quality by interpolating along the edge direction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If simple vertical interpolation is used to construct de-interlaced frames, then processing complexity is reduced, but edge preservation capability deteriorates causing stair-step artifacts
Solution Approach 1:
The patent changes the interpolation parameters by computing intensity gradients (horizontal and vertical) to determine edge directions, then adjusts interpolation behavior based on detected edge orientations. This allows the system to switch between simple vertical interpolation and edge-directed interpolation dynamically, resolving the contradiction between processing complexity and edge preservation capability.
2Manufacturing precision
If edge detection techniques are applied to preserve edges during interpolation, then picture quality is improved, but sensitivity to noisy data increases causing false edge detection
Solution Approach 1:
The patent implements dynamic edge detection that adapts to local image characteristics. By computing intensity gradients and determining edge directions dynamically for each pixel region, the system can distinguish between real edges and noise-induced artifacts. The method dynamically adjusts interpolation behavior based on detected edge characteristics, improving reliability in noisy conditions while maintaining edge preservation capability.
3Manufacturing precision
If correlation-based edge detection is used within pixel windows, then edge information can be preserved, but false edges are generated due to sensitivity to noisy video data
Solution Approach 1:
The patent changes the detection parameters by using intensity gradient computations instead of simple pixel correlation. By calculating horizontal and vertical intensity gradients and deriving edge directions from these gradients, the method becomes more robust to noise while still preserving edge information. This parameter change reduces false edge detection caused by noisy video data.
Data Source
AI summary
One embodiment of an edge-preserving vertical interpolation system constructs a de-interlaced video frame on a pixel-by-pixel basis using an edge-preserving vertical interpolation technique. Pixels within a pixel window centered about a selected pixel determine the direction of an intensity gradient associated with the selected pixel. A first pixel is constructed by interpolating between pixels that are perpendicular to the direction of the intensity gradient and a confidence factor is computed that indicates the likelihood that there is only one edge within the pixel window. A second pixel is constructed using a non-edge-preserving vertical interpolation technique. Interpolating between the first pixel and the second pixel based on the confidence factor generates a pixel in the de-interlaced video frame corresponding to the selected pixel. Unlike prior art techniques, the present invention does not attempt to correlate between pixel features and is therefore more robust to noisy pixel data than prior art techniques.