3D Halftone Generation With Randomized Layer Patterns

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In additive manufacturing, deterministic halftoning can lead to the repetition of minor errors and flaws in successive layers of a three-dimensional object, resulting in defective products due to the identical halftone patterns produced from overlapping slices with the same two-dimensional geometry.

Innovation Solution

A method of non-deterministically generating halftone data for overlapping slices in additive manufacturing systems, where the processor generates unique halftone patterns for each layer by shifting the initialization position of the halftone matrix or using error diffusion techniques with random error preloading, ensuring that each layer has a distinct pattern despite identical input geometrical data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If deterministic halftoning is used to generate halftone data from identical input geometrical data, then the halftone pattern is consistent and reproducible, but errors and flaws repeat in successive layers resulting in defective products

Engineering Contradiction:
Improveproduct qualityVSAvoidhalftone pattern consistency
Core Design Contradiction:
ReliabilityVSManufacturing precision

Solution Approach 1:

The patent applies dynamics by making the halftoning process variable rather than static. A random number generator is introduced to create different halftone patterns from the same input geometrical data across successive layers. This dynamic approach ensures that errors do not repeat in the same locations in overlapping slices, thereby improving product quality while maintaining adequate pattern consistency through the underlying geometric constraints

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the parameter of halftone pattern generation from deterministic to non-deterministic by incorporating random number generation. The random parameters are applied specifically to overlapping slices to vary the halftone patterns, which prevents error aggregation while still maintaining the essential geometric fidelity of the original model

Inventive Principle:
Principle #35Parameter changes

2Reliability

If unique halftone patterns are generated for each layer using non-deterministic methods, then errors and flaws do not aggregate, but the halftone generation process becomes more complex

Engineering Contradiction:
Improveerror aggregation preventionVSAvoidhalftone generation process
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies local quality by applying non-deterministic halftoning specifically to overlapping slices rather than uniformly to all layers. The system identifies which slices overlap and applies randomization only to those regions, leaving non-overlapping regions with deterministic processing. This localized approach prevents error aggregation in critical overlapping areas while minimizing the overall complexity increase

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent introduces a random number generator as an intermediary element between the input geometrical data and the halftone pattern output. This intermediary adds the necessary variability to prevent error repetition while maintaining a relatively simple implementation structure that integrates with existing halftoning algorithms

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP3204920B1Generating halftone data for a three-dimensional object
Publication Date: 2021.09.22 HEWLETT PACKARD DEVELOPMENT COMPANY LP
  • EP3204920B1 patent drawingFigure 1
  • EP3204920B1 patent drawingFigure 2
  • EP3204920B1 patent drawingFigure 3

AI summary

A method of forming a three-dimensional object includes non-deterministically generating halftone data for portions of a slice of three-dimensional model data of the three-dimensional object that overlap with a previous slice. The non-deterministically generating halftone data may comprise using randomized initialization for each slice of a three-dimensional object.