Volumetric Viewport Rendering With Default Values for Missing Pixels
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for rendering volumetric scenes are computationally expensive due to the need for inpainting or filtering missing pixel attributes at the decoder side, lacking a format that directs renderers to the most appropriate solution for handling missing pixel values.
Innovation Solution
Encoding metadata in a data stream with suggested attribute values, such as color, reflectance, transparency, or shininess, to be used as default values for missing pixels, along with quality levels indicating visibility of visual artifacts, allowing renderers to efficiently handle missing data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If inpainting or filtering methods are used at the decoder side to handle missing pixel attributes, then the completeness of rendered pixel attributes is improved, but the computational resources required at the decoder/renderer side increase significantly
Solution Approach 1:
The encoder performs preliminary action by computing and transmitting suggested default attribute values in the bitstream before rendering. These pre-computed suggested values are provided in advance to guide the renderer, eliminating the need for complex post-processing operations at the decoder side.
Solution Approach 2:
The suggested default attribute values act as an intermediary between the incomplete volumetric scene representation and the final rendered image. These intermediary values guide the renderer to fill missing pixel attributes efficiently without requiring computationally intensive inpainting or filtering operations.
2Device complexity
If suggested default attribute values are provided in the bitstream, then the computational burden on renderers is reduced, but additional metadata encoding is required at the encoder side
Solution Approach 1:
The suggested default attribute values are extracted as separate metadata from the main volumetric scene representation. This allows the encoder to add only the necessary default values without re-encoding the entire scene, minimizing the overhead while providing useful guidance to renderers.
Solution Approach 2:
The suggested default attribute values serve as a lightweight, disposable guide that provides sufficient information for rendering without requiring complex or permanent solutions. These simple default values are computationally inexpensive to transmit and process, making them an efficient compromise between completeness and computational cost.
Data Source
AI summary
Methods, device and data stream are provided to generate, transmit and decode volumetric scenes. According the present principles, rendering attribute values, like color values, are transmitted to the renderer to be used as a suggested attribute value for pixels for which this attribute is missing. For example, when rendering a viewport image of a volumetric scene, some pixels may miss a color information. In such a case, the present principles provide a suggested default color. These principles apply to various rendering attributes like reflectance, transparency, material or shininess. In an embodiment, an attribute value is associated with a region of the 3D space. In another embodiment, a quality level linked to the default value helps the renderer to select a filling method.