Simplification of 3D Computer-aided Design Models to Improve Rendering Performance
Heinonen, Antti (2019)
2019
Tietotekniikka
Tieto- ja sähkötekniikan tiedekunta - Faculty of Computing and Electrical Engineering
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2019-01-25
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tty-201901251165
https://urn.fi/URN:NBN:fi:tty-201901251165
Tiivistelmä
Visualization of three-dimensional (3D) computer-aided design model is an integral part of the design process. Large assemblies such as plant or building designs contain a substantial amount of geometric data. New constraints for visualization performance and the amount of geometric data are set by the advent of mobile devices and virtual reality headsets. Our goal is to improve visualization performance and reduce memory consumption by simplifying 3D models while retaining the output simplification quality stable regardless of the geometric complexity of the input mesh.
We research the current state of 3D mesh simplification methods that use geometry decimation. We design and implement our own data structure for geometry decimation. Based on the existing research, we select and use an edge decimation method for model simplification. In order to free the user from configuring edge decimation level per model by hand, and to retain a stable quality of the simplification output, we propose a threshold parameter, \textit{edge decimation cost threshold}. The threshold is calculated by multiplying the length of the model’s bounding box diagonal with a user-defined scale parameter.
Our results show that the edge decimation cost threshold works as expected. The geometry decimation algorithm manages to simplify models with round surfaces with an excellent simplification rate. Based on the edge decimation cost threshold, the algorithm terminates the geometry decimation for models that have a large number of planar surfaces. Without the threshold, the simplification leads to large geometric errors quickly. The visualization performance improvement from the simplification scales almost at the same rate as the simplification rate.
We research the current state of 3D mesh simplification methods that use geometry decimation. We design and implement our own data structure for geometry decimation. Based on the existing research, we select and use an edge decimation method for model simplification. In order to free the user from configuring edge decimation level per model by hand, and to retain a stable quality of the simplification output, we propose a threshold parameter, \textit{edge decimation cost threshold}. The threshold is calculated by multiplying the length of the model’s bounding box diagonal with a user-defined scale parameter.
Our results show that the edge decimation cost threshold works as expected. The geometry decimation algorithm manages to simplify models with round surfaces with an excellent simplification rate. Based on the edge decimation cost threshold, the algorithm terminates the geometry decimation for models that have a large number of planar surfaces. Without the threshold, the simplification leads to large geometric errors quickly. The visualization performance improvement from the simplification scales almost at the same rate as the simplification rate.