360-Degree Video Streaming Spatial Object Segmentation

The patent segments the 360-degree video content into multiple spatial objects (e.g., spherical harmonics coefficients grouped by frequency bands). This allows the system to transmit only the necessary segments corresponding to the user's current viewport, rather than transmitting the entire video content. The segmentation enables selective transmission of high-quality video data for the visible region while reducing overall data volume.

The patent applies local quality enhancement by transmitting high-resolution video data only for the user's current viewport region, while using lower resolution or compressed representations for other regions. This is achieved by encoding spatial objects at different quality levels and selectively transmitting those relevant to the user's field of view, thereby improving perceived video quality without proportionally increasing total data transmission volume.

The patent implements preliminary action by pre-processing and pre-positioning multiple spatial objects corresponding to different viewport regions before the user actually looks in those directions. When the user changes their field of view, the system can immediately switch to the pre-prepared spatial object for the new viewport region, avoiding the need to fetch and process new data in real-time. This reduces the perceived latency when users rotate their view.

The patent employs dynamic adaptation by continuously monitoring the user's viewport changes and dynamically adjusting which spatial objects are transmitted and displayed. The system adapts the video stream in real-time based on the user's head orientation and movement, switching between different pre-encoded spatial objects to maintain optimal data transmission efficiency while responding to viewport changes.

The patent segments the high-resolution video content into multiple lower-resolution spatial objects (e.g., dividing the spherical video into frequency bands or angular regions). Each spatial object can be decoded independently at a lower computational complexity, and then combined to form the complete high-resolution viewport. This segmentation allows terminals with limited decoding capability to process multiple small segments rather than one large high-resolution stream.

The patent changes the resolution parameter dynamically based on the user's viewport and the terminal's capabilities. Instead of transmitting a single fixed maximum resolution stream, the system adjusts the resolution of transmitted spatial objects according to the user's field of view and device capabilities, allowing high-quality viewing without requiring the terminal to support maximum resolution decoding for all content.

The patent applies different compression rates to different spatial regions based on their importance to the user's viewing experience. The central viewport region, which the user is most likely to observe, is transmitted with lower compression (higher quality), while peripheral regions use higher compression rates. This local quality adaptation maintains excellent viewing experience in the critical viewing area while improving overall data transmission efficiency through aggressive compression of less critical regions.

The patent dynamically changes compression parameters based on the user's viewport position, movement speed, and network conditions. When the user's view is stable, higher quality encoding is used. When the user rotates quickly or network bandwidth is limited, the system adjusts compression rates adaptively. This parameter adaptation allows the system to optimize the trade-off between transmission efficiency and viewing experience in real-time.