3-D scans from smartphones?
- By Mark Pomerleau
- Jan 07, 2016
3-D printing has been used by astronauts to manufacture tools on the International Space Station and by defense contractors to develop liquid-propellant rocket engine applications for the Air Force. It has even been suggested as a new business for the U.S. Postal Service. High-quality 3-D printing, however, depends on having accurate three-dimensional scans of the objects to be manufactured. And while 3-D printers are becoming affordable even for hobbyists, the same cannot be said for the scanners.
That may change, though, thanks to researchers from Brown University who want to use off-the-shelf digital cameras as readily deployable and reliable 3-D scanners.
“One of the things my lab has been focusing on is getting 3-D image capture from relatively low-cost components,” said Gabriel Taubin, associate professor of engineering and computer science at Brown. “The 3-D scanners on the market today are either very expensive or are unable to do high-resolution image capture, so they can’t be used for applications where details are important.”
In high-quality 3-D scanning, images are typically captured by using a technique called structured light, in which a projector casts light patterns over an object and a camera captures the images shown under the series of light beams. This structured light method only works when both the pattern projector and camera are precisely synchronized, which requires specialized and costly hardware.
With funding from the National Science Foundation, Taubin’s team has developed an algorithm that enables the structured light process to be used without the synchronization between the projector and camera, which should allow off-the-shelf cameras to be used for 3-D scanning. The key requirement is that they support burst mode (shooting several successive frames per second), which many DSLR cameras and smartphones can do.
The algorithm calibrates the timing of the image sequence following the capture of burst of images by the camera using information within the projected pattern. Pixel by pixel, the algorithm works to assemble a new sequence of images that captures each pattern as a whole. Finally, a standard structured light 3-D reconstruction algorithm can create a single 3-D image of the object.
During testing, the researchers used a standard structured light projector, but the team hopes to develop a structured light flash that could be used as camera attachment, now that there's an algorithm that can properly assemble the images.
“We think this could be a significant step in making precise and accurate 3-D scanning cheaper and more accessible,” Taubin said.
Mark Pomerleau is a former editorial fellow with GCN and Defense Systems.