Sparse Volumetric Effects Codec for Fast Quality-Preserving Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage and processing methods for volumetric data are inefficient, particularly for sparse data, as they do not account for the structure of 3-dimensional data, lack effective compression, and are not optimized for real-time use cases.
Innovation Solution
A method for compressing sparse volumetric data involves subdividing the data into blocks, applying a spatial-frequency transform, frequency-dependent quantization, and entropy encoding to generate a compressed bitstream, which can be efficiently decoded with minimal quality loss.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If volumetric data is stored explicitly with full precision, then reconstruction quality is maintained, but memory consumption becomes prohibitively expensive
Solution Approach 1:
The patent divides volumetric data into multiple octrees, where each octree represents a hierarchical decomposition of 3D space. This segmentation allows the system to manage and compress large volumetric datasets by processing them in manageable hierarchical units, reducing memory requirements while preserving reconstruction quality through selective storage of significant coefficients.
Solution Approach 2:
The patent extracts and stores only the most significant frequency coefficients after applying a spherical harmonic transform and wavelet transform. By identifying and retaining only the coefficients that contribute meaningfully to reconstruction quality, the system achieves efficient compression without sacrificing essential visual information, thereby reducing memory consumption while maintaining acceptable reconstruction fidelity.
2Productivity
If existing compression methods like VDB or JPEG are applied to volumetric data, then some compression is achieved, but they do not efficiently represent 3-dimensional sparse data structure
Solution Approach 1:
The patent transforms volumetric data from spatial domain to frequency domain using spherical harmonic transforms and wavelet transforms. This dimensional transformation allows the data to be represented in a form that better captures the inherent 3D structure and sparsity patterns, enabling more efficient compression algorithms that are specifically adapted to volumetric data characteristics rather than treating it as 2D image slices.
Solution Approach 2:
The patent employs adjustable compression parameters including octree depth, spherical harmonic order, and wavelet decomposition levels. These parameters can be tuned to balance compression ratio against reconstruction quality, allowing the system to adapt to different application requirements and data types, thereby improving both compression efficiency and 3D representation capability.
3Quantity of substance
If sparse volumetric data is compressed using traditional methods, then file size is reduced, but decoding speed is insufficient for real-time applications
Solution Approach 1:
The patent pre-computes and stores transformation matrices and basis functions during the encoding phase. By preparing these computational elements in advance, the decoding process can proceed more quickly by simply applying the pre-computed transformations rather than calculating them in real-time, thus achieving faster decoding speeds suitable for real-time applications while maintaining acceptable file sizes.
Data Source
AI summary
A control circuit is configured to divide a volumetric dataset into a plurality of volumetric blocks, apply a spatial-frequency transform to each of the volumetric blocks to obtain transform-domain coefficients, quantize the transform-domain coefficients according to one or more quantization parameters to provide quantized coefficients, reorder the quantized coefficients based on a scanning order determined to reduce coefficient differences, compose a collection of symbols for each volumetric block, the collection of symbols including at least a block header and a sparse representation of non-zero coefficients, entropy encode at least some of the symbols to generate a compressed bitstream, and store the compressed bitstream in memory.