Video Foreground Contour Detection Using Pixel Border Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital video segmentation methods often generate false contours, leading to incorrect segmentation and visible artifacts in reconstructed video sequences, especially when high CPU processing power is not available or when the probability of failure is high.
Innovation Solution
The method processes video frames by generating a pixel mask indicating whether each pixel belongs to the foreground or background and determines contours by analyzing pixel borders between adjacent pixels, using techniques such as look-up tables and gradient operators to accurately identify foreground subject edges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional contour detection methods are used in video segmentation, then the segmentation process can be performed with simpler algorithms, but false contours are generated leading to incorrect segmentation and visible artifacts
Solution Approach 1:
The patent introduces an intermediary verification step between pixel classification and contour detection. A verification module checks whether detected contours are genuine by analyzing multiple factors including color consistency, spatial continuity, and temporal stability across frames. This intermediary layer filters out false contours before they propagate to the final segmentation result, thereby improving reliability without requiring completely complex algorithms.
Solution Approach 2:
The system implements feedback mechanisms where detected contours are validated against multiple criteria and previous frame information. If a contour fails verification checks (e.g., shows inconsistent color properties or discontinuous spatial patterns), feedback is sent to reject or correct the contour detection. This feedback loop continuously refines the segmentation accuracy by eliminating false contours while preserving genuine object boundaries.
2Reliability
If complex contour verification procedures are implemented to reduce false contours, then segmentation accuracy improves, but CPU processing power requirements increase significantly
Solution Approach 1:
The patent applies partial verification to contours based on their likelihood of being genuine. High-probability contours (those with strong color contrast and clear spatial patterns) receive minimal verification, while ambiguous contours undergo more rigorous checks. This selective approach maintains high accuracy for obvious cases while reducing computational overhead for uncertain cases, balancing reliability and processing power requirements.
Solution Approach 2:
The system performs preliminary filtering of potential contours using simple, low-cost criteria before applying complex verification procedures. Contours that fail basic checks (e.g., insufficient pixel count, extreme aspect ratios, or locations in uniform regions) are discarded early without requiring intensive CPU resources. This preliminary action eliminates obviously false contours before they consume significant processing power, maintaining accuracy while reducing overall computational burden.
3Device complexity
If traditional pixel-based contour definition is used, then the implementation is simpler, but the probability of generating false contours increases
Solution Approach 1:
The patent transitions from defining contours solely by pixel membership to defining them by border characteristics between adjacent pixels of different classes. This dimensional change from area-based to boundary-based definition, combined with analyzing border properties (orientation, continuity, color transition patterns), provides more discriminative information for distinguishing genuine from false contours while maintaining reasonable implementation complexity.
Data Source
AI summary
A method of image processing includes: receiving at least one video frame of a video sequence, the at least one video frame including at least one foreground subject and a background; and processing the at least one video frame so as to separate the at least one foreground subject from the background. The processing includes: generating a pixel mask containing information indicating, for each pixel of the at least one video frame, whether the pixel belongs to the foreground subject or to the background, and determining contours of the at least one foreground subject on the pixel mask. The determining of the contours includes for each pixel in the at least one video frame; based on the information included in the pixel mask, determining whether at least one pixel border belongs to a contour of the at least one foreground subject, the at least one pixel border separating the pixel from a respective at least one adjacent pixel.