PERANCANGAN PIPELINE FOTOGRAMETRI PADA PROSES REVERSE ENGINEERING MODEL TIGA DIMENSI DI PT PRESA GENTA ENGINEERING
PT Presa Genta Engineering (PGE) is a manufacturing company that implements an engineer-to-order (ETO) system. The company faces issues in the reverse engineering process, which is time-consuming and has inconsistent accuracy. This research aims to apply photogrammetry techniques in the reverse e...
Saved in:
Main Author: | |
---|---|
Format: | Final Project |
Language: | Indonesia |
Online Access: | https://digilib.itb.ac.id/gdl/view/86117 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Institut Teknologi Bandung |
Language: | Indonesia |
Summary: | PT Presa Genta Engineering (PGE) is a manufacturing company that implements an
engineer-to-order (ETO) system. The company faces issues in the reverse engineering
process, which is time-consuming and has inconsistent accuracy. This research aims to
apply photogrammetry techniques in the reverse engineering process of three-dimensional
models of PT PGE's products as an alternative method for conducting the process.
Photogrammetry was chosen as a solution to overcome the limitations of manual methods
in measuring and creating computer-aided designs of three-dimensional products.
In this study, a photogrammetry-based processing pipeline was designed, integrating the
use of smartphones and open-source software to automate the creation of threedimensional
models.
The
pipeline
consists
of
six
stages
of
data
collection
and
processing,
which
is image capture, image import, structure from motion, multi-view stereo, mesh
texturing, and mesh cleaning. This pipeline was tested through experiments with two
factors, the image capture angle and the frame interval, verified using basic geometric
objects, a prism, and a pyramid. The verified testing was conducted on two objects
representing prismatic components, the base part, and cylindrical components, the
rotating part.
The research results indicate that photogrammetry can produce three-dimensional models
with good accuracy. Additionally, the choice of image capture angle and frame interval
did not significantly affect the accuracy level or processing time requirements. It was also
found that adding image capture angles for the rotating part improved accuracy. The
results of this study can be further developed by adding more factors and levels to the
experiments.
|
---|