Aug 23 2018
In the future, car windshields may become a huge camera sensing objects on the road. Or each window in a home could be converted into a security camera.
University of Utah electrical and computer engineering associate professor Rajesh Menon has discovered a way to create an optics-less camera in which a regular pane of glass or any see-through window can become the lens. (Image credit: Dan Hixson/University of Utah College of Engineering)
University of Utah electrical and computer engineers have found a way to build an optics-less camera in which a standard pane of glass or any see-through window can be used as the lens.
Their innovation was described in a research paper titled, "Computational Imaging Enables a 'See-Through' Lensless Camera," which was published in the latest issue of Optics Express. The paper was co-authored by University of Utah electrical and computer engineering graduate Ganghun Kim.
University of Utah electrical and computer engineering associate professor Rajesh Menon contends that all cameras were created with the idea that humans look at and interpret the pictures. But what if, he asked, one could build a camera that can be interpreted by an algorithm-based computer program?
"Why don't we think from the ground up to design cameras that are optimized for machines and not humans. That's my philosophical point," he says.
If a regular digital camera sensor such as the one used in a mobile phone or an SLR camera is pointed at an object without a lens, it results in an image that resembles a pixelated blob. But within that blob is still sufficient digital information to detect the object if a computer program is suitably designed to identify it. All that is needed is an algorithm to decode the image.
Through a chain of experiments, Menon and his team of scientists clicked a picture of the University of Utah's "U" logo as well as video of an animated stick figure, both screened on an LED light board. An economical, off-the-shelf camera sensor was connected to the side of a plexiglass window, but pointed into the window while the light board was placed in front of the pane at a 90-degree angle from the front of the sensor. The subsequent image from the camera sensor, with help from an algorithm-based computer processor, is a low-resolution picture but positively recognizable. The technique also can create a full-motion video as well as color images, Menon says.
The process requires wrapping reflective tape around the edge of the window. Most of the light emanating from the object in the picture passes via the glass, but just enough - about 1% - scatters via the window and into the camera sensor for the computer algorithm to decipher the image.
Although the resulting photo cannot win a Pulitzer Prize, it would be sufficiently good for applications such as obstacle-avoidance sensors for autonomous cars. But Menon says more robust camera sensors can create higher-resolution images.
Applications for a lensless camera can be almost limitless. Security cameras could be assembled into a home during construction by utilizing the windows as lenses. It could be used in augmented-reality goggles to decrease their bulk. With existing AR glasses, cameras have to be pointed at the user's eyes so as to track their positions, but with this technology they could be placed on the sides of the lens to minimize size. A car windshield could have many cameras along the edges to record more information. Also, the technology also could be applied in the retina or other biometric scanners, which usually have cameras pointed at the eye.
"It's not a one-size-fits-all solution, but it opens up an interesting way to think about imaging systems," Menon says.
Going forward, Menon and his team will further enhance the system, including 3D images, higher color resolution and photographing objects in usual household light. His present experiments involved capturing pictures of self-illuminated images from the light board.