Application and Parametric Design of Line Visual Illusion Graphics in Clothing

Yixin Zou, Yan Wang, Ding-Bang LuhFibres 2023 · DOI

TL;DR

This paper explores the intersection of Op Art and algorithmic modeling, demonstrating how parametric design tools like Grasshopper can be used to create garments that manipulate human perception. By shifting from manual tailoring to variable-driven logic, the authors show how "line visual illusions" can be dynamically adjusted to reshape the perceived silhouette of a wearer in real-time.

What Problem Does This Paper Address?

In traditional fashion design, creating complex visual effects—such as the "slimming" effect of certain stripe patterns—requires extensive manual fabric testing and cumbersome "piece-by-piece" tailoring. This limits the complexity and personalization of the garments.

Furthermore, while "Op Art" (Optical Art) has influenced fashion since the 1960s, there is a lack of rigorous methodology for translating these artistic principles into digital, parametric workflows. The authors aim to bridge this gap by creating a system where visual illusions are not just static prints, but algorithmic structures that can be tuned to achieve specific aesthetic and perceptual goals.

Methods

The researchers employed a multi-stage workflow combining psychological analysis with computational geometry:

  1. Literature & Theoretical Framework: They categorized line illusions into three types: Segmentation (horizontal/vertical spacing), Angle (e.g., the Hering illusion where radial lines make parallel lines look bowed), and Contrast (size perception relative to surroundings).
  2. Semantic Analysis: To ensure the designs met user expectations, they conducted interviews and surveys with 20 fashion enthusiasts. Using a five-level Likert scale and 24 adjective pairs (e.g., "Traditional-Fashionable," "Steady-Dynamic"), they performed a cluster analysis to map specific line patterns to human emotional responses.
  3. Parametric Modeling (Rhino + Grasshopper): The core of the technical work involved building "batteries" (algorithmic nodes) in Grasshopper. Key techniques included:
  4. Graph Mapping: Using non-linear mathematical functions to create unevenly arranged squares and lines.
  5. Polar Arrays & Mirroring: Generating complex radial patterns from simple geometric seeds.
  6. Iso-Curve Mapping: Projecting 2D line patterns onto 3D garment surfaces, allowing the lines to follow the contour of the "body" and define features like the waistline algorithmically.

Datasets

The paper does not use large-scale external datasets (like ImageNet). Instead, it relies on:

  • Primary Perceptual Data: Collected from 20 subjects (fashion students, teachers, and enthusiasts) to evaluate five types of line visual illusion prototypes.
  • Semantic Vocabulary: A refined set of 10 opposite adjective pairs used for the Likert scale evaluation.

Key Insights and Ideas

  • The Helmholtz Paradox in Code: The authors reference the Helmholtz illusion, noting that horizontal stripes can actually make a wearer look thinner than vertical ones. By using parametric sliders, they can fine-tune the "stripe interval" to maximize this effect without redrawing the garment.
  • Math as an Artistic Brush: The research treats mathematical parameters (like rotation angles or "Fillet" radii) as design tools. For example, changing a "Number Slider" from 3 to 5 instantly transforms a triangular pattern into a pentagonal one, which then propagates through the entire garment's visual logic.
  • Dynamic Reshaping: Unlike traditional patterns, these parametric illusions are "aware" of the garment's silhouette. The internal line elements change in response to the outer contour, allowing for a "stretching" effect that can define a waistline or lengthen a torso purely through visual manipulation.

Key Results

  • Workflow Validation: The authors successfully demonstrated a reproducible 4-step logic for "Linear Visual Illusion Clothing Generation": Surface Segmentation → Line Screening → Line Construction → Base Surface Construction.
  • Pattern Complexity: They generated a sophisticated "vortex" line pattern (Figure 7) that would be nearly impossible to draft manually, achieving it through a sequence of algorithmic rotations and offsets.
  • Perceptual Mapping: The cluster analysis (Table 1) successfully grouped design elements into three distinct "classes" (e.g., Class 1: Complicated/Technological/Elegant), providing a roadmap for designers to target specific "vibes" using math.

Limitations and Future Work

  • Digital-Only Scope: The current research focuses on the digital design and visualization phase. The authors acknowledge that they have not yet integrated these results with physical fabrication methods like 3D printing or complex textile weaving.
  • Pattern Diversity: The study primarily focuses on line-based illusions. Future work could expand into other geometric illusions, such as Voronoi diagrams or Moire patterns.
  • Materiality: The paper does not discuss how different fabric types (drape, sheen, texture) might interfere with or enhance the digital visual illusions when translated to the physical world.

Summary updated

Abstract

Abstract This research adopted literature analysis methods to explain the principles and types of line illusion and, secondly, using Grasshopper software generated some fashion design samples with visual illusion element application. Then semantic analysis was applied to investigate audiences’ cognition, with the aim to improve the design. Lastly, Grasshopper software was used to generate new design patterns by changing the set of generated points through algorithm operation and dotted line interference to change the presented linear illusion forms and patterns. The results of this research show that with the support of parametric technology, the opportunities and space for creating line elements in clothing will be further expanded, enabling that real-time pattern or apparel design changes, and adjustments can be realized by parameter variable controlling. The research comprised a new attempt at combining linear visual illusion, fashion design and parametric design, showing the content and design case studies for relevant research in this area.
BibTeX
@inproceedings{zou2023application,
  author = {Yixin Zou and Yan Wang and Ding-Bang Luh},
  title = {{Application and Parametric Design of Line Visual Illusion Graphics in Clothing}},
  booktitle = {Fibres},
  year = {2023},
  doi = {10.2478/ftee-2023-0017},
}
Figure