Method for encoding texture data of free viewpoint television signals, corresponding method for decoding and texture encoder and decoder

Simple SummaryContent extracted from patent full text and abstract with AI.

This patent presents a method to efficiently encode and decode texture data for free viewpoint television (FVTV) systems, which are created using a limited number of synchronized cameras capturing a scene from different locations. The innovation centers on preprocessing each camera’s video to subtract the background, then encoding each camera's foreground independently using temporal and spatial integer wavelet transforms. The resulting data is compressed and can be decoded in real time, enabling smoother and more accessible rendering of scenes from any desired viewpoint.

Use CasesContent extracted from patent full text and abstract with AI.

• Free viewpoint television broadcasting or streaming, where users can select and move their viewpoint freely within a scene.
• Sports or live event broadcasting with interactive camera angles for viewers.
• Virtual reality (VR) and augmented reality (AR) applications that require efficient multi-view video encoding and decoding.
• Surveillance systems using fewer cameras but with the ability to reconstruct and view scenes from multiple angles.
• Telepresence systems where realistic scene recreation from limited cameras is needed for remote meetings or collaboration.

BenefitsContent extracted from patent full text and abstract with AI.

• Greatly reduces the computational complexity and data rate required for multi-view video encoding, allowing for real-time processing and streaming.
• Independent encoding/decoding per camera enables efficient random access and scalability, which is essential for interactive applications.
• Utilizing wavelet transforms with integer coefficients reduces hardware requirements and is suitable for low-powered devices.
• Efficient handling of background subtraction and focus on foreground ensures bandwidth is used for relevant visual data, enhancing user experience and compression ratios.
• System design is well-suited for sparse camera setups, reducing overall equipment and infrastructure costs.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

Method for encoding texture parts of free viewpoint television signals where the free viewpoint television signals are acquired using a sparse set of synchronized cameras capturing a scene from different spatial locations and where the signals of each camera for a specified time instant are pre-processed by implementing a background subtraction, thereby obtaining a set of subtracted signal frames (14), wherein the subtracted signal frame (14) of each single camera for a specified time is being encoded independently from the subtracted signal frames (14) of the other cameras by the use of a temporal and a spatial integer wavelet transform.