Voxel-to-Mesh Pipeline for Thin-Feature-Preserving Smoothing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional methods for converting voxel data into faceted object representations often result in highly unsmoothed meshes with undesired local features like wavy ridges and stair-like surfaces, and indiscriminate mesh smoothing algorithms can alter the underlying topology of the object design, invalidating specific design goals and constraints.

Innovation Solution

A computational pipeline that includes feature detection and labeling of thin features, adaptive pressing, and iterative mesh smoothing processes to preserve the topology while smoothing unwanted local features, using techniques like morphological opening, second-order smoothing, and voxel averaging, along with constrained mesh pressing to adjust facet boundaries.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If voxel data is processed using conventional methods, then processing time and computational resources are excessive, but the quality and detail of the generated representations may be compromised

Engineering Contradiction:
Improveprocessing speedVSAvoidcomputational resource requirements
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The processing system divides voxel data into multiple patches or chunks that can be processed independently and in parallel. This segmentation allows the computational workload to be distributed across multiple processing units, reducing the time and resources required for each individual processing operation while maintaining overall representation quality.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transforms the processing approach by moving from direct high-resolution voxel processing to a multi-scale representation system. By creating representations at different levels of detail and combining them hierarchically, the system achieves efficient processing without sacrificing the ability to represent fine details when needed.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Manufacturing precision

If high-resolution voxel data is processed in detail, then representation quality improves, but processing time increases significantly

Engineering Contradiction:
Improverepresentation qualityVSAvoidprocessing time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The system dynamically adjusts the level of processing detail applied to different regions of the voxel data based on their importance and the desired output resolution. Critical regions receive higher processing detail while less important areas use simplified processing, optimizing the balance between representation quality and processing time.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent implements preliminary processing steps that create coarse representations and identify important features before detailed processing. This preliminary action allows the system to focus computational resources on critical areas that will most impact representation quality, reducing overall processing time while maintaining high-quality output.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4371078B1Processing pipelines for generation of faceted object representations from voxel data
Publication Date: 2026.05.20 SIEMENS INDUSTRY SOFTWARE INC
  • EP4371078B1 patent drawingFigure 1
  • EP4371078B1 patent drawingFigure 2
  • EP4371078B1 patent drawingFigure 3

AI summary

A computing system (100) may include a voxel access engine (108) configured to access voxel data (210, 310) and a voxel processing engine (110). The voxel processing engine (110) may identify and label thin features in the voxel data (210, 310), smooth the voxel data (210, 310) to preserve the thin features, and convert the voxel data (210, 310) to form a faceted representation (320). The voxel processing engine (110) may also adaptively perform a pressing process on the faceted representation (320). Responsive to a determination that a pressing reapplication criterion is satisfied, the voxel processing engine (110) may modify the voxel data (210, 310), convert the modified voxel data (340) to form the faceted representation (320) of the object, and perform the pressing process on the faceted representation (320) of the object formed through the modified voxel data (340).