Large Scale Painting of Photographs by Interactive Optimization

Dartmouth Researcher Invents Smart Paint Spray

A Dartmouth researcher together with his colleagues has invented a smart paint spray which robotically reproduces photographs as significant murals. The computerized technique that spray paints a photo is not expected to issue a trend of giant graffiti, but can be utilised in digital fabrication, digital and visual arts, together with artistic stylization and other application. The findings had appeared in the journal Computer & Graphics and the project is in partnership between ETH Zurich, Disney Research Zurich, Dartmouth College and Columbia University. The spray can system is an innovative twist on computer aided painting that had been initiated in early 1960s and is well-studied topic among artists and scientist.

 Spray paint tends to be economical and easy to use making large-scale spray painted mural common in modern urban culture. However, it could be logistically and technically difficult for skilled artists in covering walls of buildings as well as other large surfaces. The researchers desired to create a method of helping non-artists in creating perfect reproductions of photographs as large-scale murals utilising spray painting and hence they developed a computer aided method which tends to use an ordinary paint can, tracking the position of the can related to the wall or canvas, recognizing what image it would want to paint.

Fast & Light-Weight/Two Webcams/QR Coded Cubes

As the person wave the pre-programmed spray can around the canvas, the system tends to automatically operate the spray on/off button to provide the specific image as spray painting. The prototype seems to be fast and light-weight and includes two webcams together with QR coded cubes for the purpose of tracking as well as a small actuation device for the spray can that is attached through a 3D printed mount. The paint commands are transmitted through radio directly connected to a servo-motor operating the spray nozzle.

On a close by computer, the real time process regulates the optimal amount of paint of the current colour to spray at the tracked location of the spray can. The outcome is that the painting portrays itself as the user waves the can around without the user needing to know the image beforehand.Owing to spray paint difficulty in obtaining permission foruniversity buildings, the researchers have tested the automated painting system on large sheets of paper gathered into mural-size paintings.

Permitting Unskilled Users – Creating Physical Piece of Art

Though this system seems to presently support only painting on flat surface, one possible advantage of the new technique over standard printing is that it could be usable on more complex curved painting surfaces. Co-author Wojciech Jarosz, assistant professor of computer science at Dartmouth who was a senior research scientist earlier at Disney Research Zurich stated that computationally assisted painting methods are limited to the computer and in this research, it is shown that by combining computer graphics and computer vision techniques, they can bring such assistance technology to the physical world even for very traditional painting medium which has created an unconventional kind of digital fabrication.

Their assistive approach seems to be like a modern take on `paint by numbers’ for spray painting and most important was that they wanted to maintain the aesthetic aspects of physical spray painting and the physical experience of holding and waving a physical spray can while permitting unskilled users in creating physical piece of art.