Free-View Watermark Detection via Light Field Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In free-view televisions, existing technologies fail to detect embedded invisible watermarks in virtual views generated for arbitrary camera angles, making it difficult to identify copyright ownership and prevent illegal transmission of content.
Innovation Solution
A method that uses Arbitrary View Generation and Free-view Watermark Detection to embed and extract watermarks in virtual views by interpolating light field images, employing high pass filtering and correlation-based schemes to determine the virtual camera's position and orientation, and correlating rendered watermarks with the rendered image to detect their existence.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If virtual views are generated by interpolating multiple camera views for free-view television, then user viewing flexibility and adaptability are improved, but watermark detection capability deteriorates
Solution Approach 1:
Watermarks are embedded in the original multi-view video content before transmission to the user. This preliminary embedding ensures that watermark information is already present in the source material, allowing for later detection even after the content is processed through virtual view generation and interpolation operations.
Solution Approach 2:
The system performs watermark detection on the received multi-view video content and uses the detection results to identify the source of illegal transmissions. This feedback mechanism enables copyright protection by tracing back to the original transmitting device through detected watermark patterns.
2Measurement precision
If watermarks are embedded on recorded views from multiple cameras, then copyright identification capability is improved, but detection capability on virtual views deteriorates
Solution Approach 1:
Watermarks are embedded in the original multi-view video content before transmission to the user. This preliminary embedding ensures that watermark information is already present in the source material, allowing for later detection even after the content is processed through virtual view generation and interpolation operations.
Solution Approach 2:
The watermark embedding and detection system is designed to work across multiple camera views and virtual views simultaneously. The same watermarking approach is applied to all views, creating a universal solution that maintains copyright identification capability whether the content is displayed as original camera views or generated virtual views.
3Adaptability or versatility
If set-top-box creates virtual views using interpolation algorithms, then user choice and adaptability are improved, but watermark detection reliability deteriorates
Solution Approach 1:
Watermarks are embedded in the original multi-view video content before transmission to the user. This preliminary embedding ensures that watermark information is already present in the source material, allowing for later detection even after the content is processed through virtual view generation and interpolation operations.
Solution Approach 2:
The system performs watermark detection on the received multi-view video content and uses the detection results to identify the source of illegal transmissions. This feedback mechanism enables copyright protection by tracing back to the original transmitting device through detected watermark patterns.
Data Source
Figure 1
Figure 2~4
Figure 5~6
AI summary
It is known that TV-viewers might record personal videos for their arbitrarily selected views in the upcoming consumer electronic product, free-view TV, it is apparent that solutions for copyright and copy protection problems should be proposed for the free-view TV against misusing such content. The proposed method embeds the same watermark into the transmitted multiple-views of the free-view TV by exploiting the spatial masking properties of the Human Visual System. Extraction of the watermark from any virtual view is achieved after correlating this view by a 'rendered watermark', which is obtained through Light Field Rendering, after determining the unknown location of the virtual camera through a homography between focal and image planes.