Temporal Video Edge Embedding for Display Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video recognition and digital watermark detection technologies face challenges with spatial distortion when imaging a display screen from varying angles, leading to reduced accuracy due to the conspicuous nature of static two-dimensional patterns, which are difficult to detect amidst video content.
Innovation Solution
An information embedding device and method that superimposes temporally varying embedded information on the edges of a video, using a combination of frequencies to create a pattern that is hard to perceive, allowing for accurate detection of the display screen's borders and improving resistance to noise and distortion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static two-dimensional patterns are used for information embedding, then the information can be embedded in video content, but the patterns become conspicuous and difficult to detect amidst video content
Solution Approach 1:
The patent applies dynamics by transforming static two-dimensional patterns into dynamic temporal patterns that vary over time. The embedded information is represented as temporal variations in video signal characteristics (such as luminance or chrominance) rather than static spatial patterns. This temporal dynamics makes the embedded information imperceptible to human vision while maintaining detectability through temporal analysis.
Solution Approach 2:
The patent employs periodic action by encoding information through periodic temporal variations in video signal parameters. The embedded data is represented as periodic modulations superimposed on the video signal, where the frequency and phase of these periodic variations carry the information. This periodic modulation approach allows reliable detection while remaining inconspicuous within the video content.
2Measurement precision
If video recognition is performed without spatial distortion correction, then processing is simpler, but accuracy is reduced due to spatial distortion from varying imaging angles
Solution Approach 1:
The patent applies preliminary action by performing spatial distortion correction as a preprocessing step before video recognition. The system detects the display screen geometry and applies perspective transformation to correct spatial distortion caused by varying imaging angles. This preliminary correction ensures that subsequent recognition operations work with geometrically accurate data, improving overall accuracy without adding complexity to the core recognition algorithms.
3Measurement precision
If more feature amounts are used in video recognition, then recognition accuracy improves, but processing time increases
Solution Approach 1:
The patent applies the extraction principle by selectively extracting and utilizing specific video signal characteristics (luminance, chrominance, or their combinations) that are most suitable for the embedded information detection task. Rather than processing all possible video features, the system focuses on the temporal variations in selected signal components, reducing processing complexity while maintaining high detection accuracy for the embedded information.
Data Source
AI summary
A memory stores first video information displayed on a display screen. A processor generates embedded information varying temporally and superimposes the embedded information on an image part at least corresponding to an edge of the display screen in the first video information so as to generate second video information on which the embedded information is superimposed. An output interface outputs the second video information.