<?xml version="1.0"?>
<metadata xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:dc="http://purl.org/dc/elements/1.1/"><dc:title>Anisotropic reflectance model for predictive rendering of FDM prints</dc:title><dc:creator>Boc,	Ana Gaja	(Avtor)
	</dc:creator><dc:creator>Marolt,	Matija	(Mentor)
	</dc:creator><dc:creator>Rittig,	Tobias	(Komentor)
	</dc:creator><dc:creator>Podpečan,	Vid	(Komentor)
	</dc:creator><dc:subject>light transport simulation</dc:subject><dc:subject>appearance fabrication</dc:subject><dc:subject>3D printing</dc:subject><dc:subject>reflectance modeling</dc:subject><dc:description>Ray tracing is a fundamental rendering technique that simulates the interaction of light with 3D scenes to generate photorealistic images. Rendering speed is intricately linked to the intricacy of the scene and the properties of the materials. This challenge is particularly pronounced when rendering objects produced through Fused Deposition Modeling (FDM), given their high polygon count and the presence of subsurface scattering characteristics within the printing material.

This thesis was dedicated to the development of a shader with dual objectives: first, to approximate object geometry through the utilization of a bump shader and second, to bypass circumvent the computationally intensive simulation of subsurface scattering. The latter goal was realized by constructing a Bidirectional Reflectance Distribution Function (BRDF) function based on material measurements.

In addition to these shader developments, we conducted renderings of objects with explicit geometry, directly derived from the printer's instructions, incorporating subsurface scattering and plastic reflections. These renderings served as benchmarks against which to evaluate our shader results. We also compared our rendered images with photographs of the corresponding physical printed objects with matching light sources and camera positions.

Using our shaders, we successfully achieved a significant improvement in rendering times. The material shader produced results that closely aligned with the benchmark renderings. However, with respect to our bump map shader, while it delivered accurate results for planar surfaces, some discrepancies emerged on spherical surfaces. It is worth noting that these results are considered to be as close to the original as is practically attainable within the shader framework. For a more precise match, it would be more feasible to engage in some sort of geometry preprocessing.</dc:description><dc:date>2023</dc:date><dc:date>2023-11-22 08:05:01</dc:date><dc:type>Magistrsko delo/naloga</dc:type><dc:identifier>152350</dc:identifier><dc:identifier>VisID: 35448</dc:identifier><dc:identifier>COBISS_ID: 177579779</dc:identifier><dc:language>sl</dc:language></metadata>
