Realistic pattern formations on surfaces by adding arbitrary roughness

Siqing Li,Leevan Ling,Steven J. Ruuth,Xuemeng Wang

2024-01-11

Abstract:We are interested in generating surfaces with arbitrary roughness and forming patterns on the surfaces. Two methods are applied to construct rough surfaces. In the first method, some superposition of wave functions with random frequencies and angles of propagation are used to get periodic rough surfaces with analytic parametric equations. The amplitude of such surfaces is also an important variable in the provided eigenvalue analysis for the Laplace-Beltrami operator and in the generation of pattern formation. Numerical experiments show that the patterns become irregular as the amplitude and frequency of the rough surface increase. For the sake of easy generalization to closed manifolds, we propose a second construction method for rough surfaces, which uses random nodal values and discretized heat filters. We provide numerical evidence that both surface {construction methods} yield comparable patterns to those {observed} in real-life animals.

Numerical Analysis

What problem does this paper attempt to address?

### Problems the Paper Attempts to Solve

This paper aims to generate surfaces with arbitrary roughness and form patterns on these surfaces. Specifically, the authors propose two methods to construct rough surfaces:

1. **First Method**: Generate periodic rough surfaces by superimposing wave functions with random frequencies and propagation angles, which have analytical parametric equations. The amplitude in this method is an important variable in the eigenvalue analysis of the Laplace-Beltrami operator and also affects pattern formation. Numerical experiments show that as the amplitude and frequency of the rough surface increase, the patterns become irregular.

2. **Second Method**: To facilitate extension to closed manifolds, a method based on random node values and discrete heat filters is proposed to construct rough surfaces. Numerical evidence shows that the patterns generated by these two surface construction methods are similar to those on real animal skin.

### Background and Motivation

Reaction-Diffusion Systems (RDS) have been widely applied in the study of pattern formation, especially in generating animal skin patterns. However, most studies have focused on smooth surfaces, neglecting the actual roughness of the skin surface. This roughness can cause patterns to no longer be strictly spots or stripes. Therefore, combining surface properties with reaction-diffusion systems can provide an opportunity to improve the generation of realistic patterns.

### Main Contributions

1. **Generation of Rough Surfaces**: Two methods for generating surfaces with arbitrary roughness are proposed, and the geometric properties and eigenvalues of the Laplace-Beltrami operator of these surfaces are analyzed.
2. **Numerical Simulation of Pattern Formation**: By solving the partial differential equations (PDE) of the reaction-diffusion system on rough surfaces, the impact of roughness on pattern formation is studied. Numerical experiments show that as the amplitude and frequency of the rough surface increase, the patterns become more irregular.
3. **Practical Application**: The similarity between the generated patterns and real animal skin patterns is demonstrated, verifying the effectiveness of the methods.

### Method Overview

- **First Method**: Generate periodic rough surfaces by superimposing trigonometric functions and solve the heat equation and reaction-diffusion system using the finite difference method.
- **Second Method**: Generate rough surfaces through random node values and discrete heat filters, and verify their effectiveness through numerical methods.

### Conclusion

The rough surfaces generated by these two methods can produce patterns similar to those on real animal skin, providing a new approach for generating more realistic biological patterns. Additionally, these methods can be extended to other types of manifolds, such as generating rough surfaces on red blood cells.