Reverse engineering program that creates customizable furniture CAD models with only its images
Every piece of furniture – an almirah, a desk, or even a stool – that we buy is based on either seeing it somewhere or loving it because we’ve owned something similar before. This is especially true for vintage furniture that cannot be found very often in stores, as it is full of the latest designs and products. Also, you might also like a piece from a friend or relative’s house, but you might not be comfortable asking where they got it from. Some of us may not want the same item, just a slightly edited version. However, we may not know where to find it.
To deal with all of this, a team of researchers from the University of Washington and Shandong University developed a method called “Fabrication-Aware Reverse Engineering for Carpentry”. The team’s method proposes to generate manufacturing plans from images of carpentry elements.
How do they do?
The researchers said that an individual may like a piece of furniture but may not want the same product in terms of dimensions. You might want to make it slightly taller and wider, they said, and for that, take a few photos from different angles of the object you want. This reverse engineering consists of taking as input a set of images of a framed object, then generating a CAD model, ready to build a replica of the object or a modified version of it.
âOur method uses domain-specific constraints to retrieve not only valid geometry, but semantically valid part assembly, using a combination of image-based and geometric optimization techniques,â read the summary of the paper.
They added that the team demonstrated the method on a range of wooden objects and furniture. The team said they can automatically get designs that are not only easy to edit, but also precise recreations of the ground truth. They further show how reverse engineering for carpentry can be used to make a physical replica of the captured object. In addition, a customized version of the product can also be obtained by directly editing the reconstructed model in CAD software.
There are also challenges. âArriving at a fabricable solution requires identifying the parts and optimizing their precise shapes and the part-to-part connections constraining those shapesâ are just a few of the challenges the researchers highlighted in their paper.
The article will be published in Eurographics Symposium on Geometry Processing.