Video Frame Type Classification via Fourier Spectrum Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to determine the type of video frames (2D or 3D) without supplementary data or when the display lacks decoding functionality, necessitating an alternative method for frame type recognition.
Innovation Solution
A computer-implemented method and system that analyzes video signals by extracting active regions, reducing data through a frame compactor, and using Fourier transforms and cepstrum analysis to distinguish between 2D and 3D frames, specifically identifying Left-Right (LR) and Top-Bottom (TB) 3D types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If supplementary data or signals are used to indicate video frame type, then the accuracy of frame type determination is improved, but the device complexity increases due to requirements for decoding and recognizing supplementary data
Solution Approach 1:
The invention extracts only the essential visual characteristics from video frames (active regions, edge maps, structural patterns) that are sufficient to determine frame type, discarding the need for supplementary data. This extraction approach maintains determination accuracy while eliminating complex decoding requirements.
Solution Approach 2:
The invention introduces intermediate processing steps (frame compaction, edge detection, structural analysis) that transform raw video data into simplified representations suitable for frame type determination, serving as mediators between the video signal and the classification decision without requiring supplementary data.
2Measurement precision
If manual user specification of signal type is used, then the frame type determination accuracy is improved, but the ease of operation deteriorates due to the troublesome process for inexperienced users
Solution Approach 1:
The system performs automatic frame type determination by analyzing video signal characteristics without requiring user intervention. The algorithm independently identifies frame types through computational analysis of visual patterns, making the system self-sufficient and eliminating the need for manual user specification.
Solution Approach 2:
The invention replaces the manual mechanical process of user specification with an automated computational system that uses image processing algorithms (Fourier transforms, edge detection, structural analysis) to automatically determine frame types based on signal characteristics.
3Device complexity
If basic image contents without type indication are used, then the device complexity is reduced, but the loss of information increases as the video signal lacks frame type information
Solution Approach 1:
The invention performs preliminary analysis of video frame characteristics (computing edge maps, analyzing structural patterns, detecting periodicities) before frame type determination is needed. This advance processing extracts implicit frame type information from basic image contents, preventing information loss without adding complex decoding requirements.
Solution Approach 2:
The invention transforms the analysis from examining raw pixel data to analyzing derived parameters (edge map characteristics, structural similarity metrics, spectral periodicities). This parameter transformation reveals frame type information that is implicit in basic image contents, recovering information that would otherwise be lost.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Accurately determines video frame types, even without explicit indication, enabling appropriate signal processing and display settings, and can be integrated into video display units or decoders to enhance video processing capabilities.
Implementation Method 1
calculating a two-dimensional Fourier Transform spectrum (Xmn) of contents of the video frame matrix
Implementation Method 2
calculating a cepstrum vector (H1...μ, V1...μ) for respectively the horizontal and the vertical direction of the two-dimensional Fourier Transform spectrum (Xmn)
Data Source
Figure 1A~1D
Figure 2A~2H
Figure 3A~3H
AI summary
A computer-implemented method for determining a video frame type, comprising the steps of receiving a video frame matrix describing the contents of the video frame, calculating a two-dimensional Fourier Transform spectrum (Xmn) of the contents of the video frame matrix, determining whether the spectrum (Xmn) is periodic with a period equal to 2 in a horizontal and/or vertical direction, classifying the video frame as a horizontal LR-type stereoscopic frame if the spectrum (Xmn) is determined as periodic in the horizontal direction and/or classifying the video frame as a vertical TB-type stereoscopic frame if the spectrum (Xmn) is determined as periodic in the vertical direction and classifying the video frame as a mono frame otherwise.