Vertex Ambient Occlusion Calculation via Triangular Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining vertex ambient occlusion values in image processing are computationally expensive and require significant memory usage, leading to high operational costs and suboptimal picture quality in rendering.
Innovation Solution
A method involving triangular surface segmentation to calculate ambient occlusion values, where errors in connecting edges are used to simplify the model, reducing calculation complexity and memory usage while maintaining accurate ambient occlusion values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ambient occlusion values are calculated for each sampling point in real-time using fragment shaders, then picture quality is improved, but computational cost and memory usage increase significantly
Solution Approach 1:
The patent divides the model surface into triangular surfaces and calculates ambient occlusion values at vertex level rather than at every fragment level. This segmentation approach maintains accuracy while reducing computational burden by working with a coarser grid of vertices and triangles.
Solution Approach 2:
The patent performs preliminary calculations of ambient occlusion values at the vertex level before rendering. By pre-computing these values and storing them in vertex attributes, the expensive rendering operations are shifted to the CPU or GPU preprocessing stage, reducing the burden during actual rendering.
2Measurement precision
If ambient occlusion values are calculated for each sampling point in real-time, then rendering accuracy is improved, but memory usage and operational overheads increase
Solution Approach 1:
The patent merges the ambient occlusion calculation with the vertex processing pipeline. By calculating AO values at vertices and storing them in vertex attributes (combining geometry and lighting data), the system reduces the need for separate per-fragment calculations and reduces memory bandwidth requirements.
Solution Approach 2:
The patent performs preliminary computation of ambient occlusion values during model processing or preprocessing stages, storing the results in vertex attributes. This shifts the heavy computational workload from real-time rendering to preprocessing, reducing memory usage during the actual rendering operation.
3Measurement precision
If complex rendering techniques are used to improve picture quality, then visual realism is enhanced, but operational costs and processing time increase
Solution Approach 1:
The patent uses a simplified vertex-based representation to copy the essential ambient occlusion information needed for rendering, rather than performing complex per-fragment calculations. This copying approach preserves the necessary visual information while dramatically reducing processing requirements.
Solution Approach 2:
The patent changes the parameter representation from per-fragment sampling to per-vertex calculation. By transforming the problem into vertex space and using vertex attributes to store AO values, the system achieves better rendering efficiency while maintaining picture quality through appropriate parameter selection and storage.
Data Source
AI summary
A method includes performing a triangular surface segmentation on a to-be-processed model to obtain a triangular surface model, the triangular surface model includes a plurality of triangles and a plurality of vertices, each triangle is associated with three vertices that are connected by connecting edges. The method also includes calculating ambient occlusion values respectively of the plurality of vertices, a first ambient occlusion value of a first vertex is calculated based on first one or more triangles associated with the first vertex. The method also includes calculating errors respectively associated with a plurality of connecting edges based on ambient occlusion values of the plurality of vertices, and collapsing a first connecting edge between two vertices based on the errors to obtain a simplified triangular surface model. Apparatus and non-transitory computer-readable storage medium counterpart embodiments are also contemplated.