Showing posts with label Kevin Kelly. Show all posts
Showing posts with label Kevin Kelly. Show all posts

Monday, March 24, 2014

Megapixel Images Created With One Pixel


The camera replaces the traditional digital pixel grid with an array of tiny micromirrors.
http://www.informationweek.com/megapixel-images-created-with-one-pixel/d/d-id/1047839
Rice University researchers have developed a camera that creates megapixel quality images using one pixel. The technology allows cameras to use a smaller number of sensors to create higher resolution images.
Rice University engineers Richard Baraniuk and Kevin Kelly have been working for two years to build a camera that replaces the traditional digital pixel grid with an array of tiny micromirrors. Each mirror points either toward a single light-sensitive pixel or away from it.
When an image focuses on the mirrors, the light breaks into a random pattern, and their combined intensity is recorded by one pixel. Using a computer algorithm developed by mathematicians during the past two and a half years, the camera software identifies the simplest possible image consistent with the samples.
"Richard Baraniuk and Kevin Kelly are part of a growing movement in the imaging community toward intelligent image processing," said Sean Varah, CEO at MotionDSP Inc., a software company with image processing technology that reconstructs low-res video images into high. "Hardware has limits, and to get past those limits, you need intelligent software. Their work is very promising as it combines both hardware and software expertise."
Recent mathematical findings from scientists Emmanuel Candes at Caltech, David Donoho at Stanford, Terence Tao at University of California Los Angeles, and Justin Romberg at Georgia Tech makes the camera technology possible.
Related to compression algorithms, such as JPEG, the technology requires less compression, battery life and space in memory and storage, Baraniuk said.
Sharper pictures require more pixels. But denser pixel arrays also drain camera batteries faster, and most of the information recorded goes unused. To store images, the camera spends more power to convert them into a fileformat, such as JPEG, removing redundant information and fine details.
"The technology could allow smaller hardware, and lower power, but it also would enable us to build cameras where current digital cameras are blind, such as far infrared ultraviolet and terahertz cameras," Baraniuk said.
The terahertz light band sits between microwaves and infrared proves most interesting to Baraniuk. The research could prove an important technology for use by the military and the U.S. Department of Homeland Security because terahertz can pass through things like clothes and luggage.
Baraniuk said both Texas Instruments Inc., and the Defense Advanced Research Products Agency (DARPA) support the research.
Aside from military or homeland security applications, commercializing the technology in the future could mean developing digital cameras that have the capability to take and process images comparable to trillions of megapixels, said Georgia Tech professor Justin Romberg.
"In the near future, the technology could work well when you need to take pictures at night, where visible light isn't present," Romberg said. "Infrared comes in handy at night when the sun goes down, but objects still radiate heat."
Making the technology smaller, less expensive and practical for consumer goods comes next on the agenda, Baraniuk said. "We'll also look at making the reconstruction process as fast and efficient as possible," he said. "The math tells us random pattern are a good way to do measurements, but there are other ways."
The technology will be demonstrated at the Optical Society of America's annual meeting in Rochester, N.Y.

Thursday, June 14, 2012

InView -Issued Patent on Compressive Sensing Camera Design


Compressive Sensing dramatically streamlines the opto-electrical architecture of 

high-performance shortwave infrared cameras



AUSTIN, Texas, Jun 14, 2012 (BUSINESS WIRE) -- The United States Patent and Trademark Office has issued patent 8,199,244 to Rice University. Patent 8,199,244 describes the basic architecture and design of a Compressive Sensing Camera, including the combination of a spatial light modular, a measurement basis, a single or small number of detector diodes, an analog-to-digital convertor and a reconstruction algorithm.
InView has an exclusive license from Rice University to apply patent 8,199,244 to imaging and hyperspectral imaging for wavelengths of ultraviolet, visual, infrared, and terahertz.
Bob Bridge, CEO of InView, said, "InView is pleased that the foundational invention of Compressive Sensing Imaging by Dr. Richard Baraniuk and Dr. Kevin Kelly of Rice University has been recognized. They built their breakthrough camera inventions on top of basic mathematical concepts previously developed, and were the first to demonstrate the application of compressive sampling theory to the design of imagers and cameras."
Bridge went on to say, "InView has defined additional application-focused IP that surrounds the basic inventions from Rice with detailed Compressive Sensing imaging implementation concepts. The combined patent portfolio benefits from a first-mover advantage, with first-mover research IP from Rice and first-mover development IP from InView."
Compressive Sensing Imaging allows cost-effective, high-resolution cameras to be built using a much smaller number of photodiode detectors than is required in a conventional camera. InView is initially using Compressive Sensing to introduce a family of low-cost, high-resolution Shortwave Infrared (SWIR) cameras. InView will additionally be applying the technology to deliver high-performance, cost-effective cameras in additional wavebands and to deliver hyperspectral imagers. Compressive Sensing shortwave infrared cameras are well suited for security, surveillance and maritime navigation applications because of their unique ability to see through impairments such as fog, haze, smoke and dust.
Patent 8,199,244 claims priority to the provisional filing date of April 2005.
InView Technology Corporation, www.inviewcorp.com , was incorporated in 2009 to commercialize the Compressive Sensing Imaging technology developed at Rice University. InView is located in Austin, Texas, and has received funding from a variety of venture capital, industrial and governmental sources.
SOURCE: InView Technology Corporation



Tuesday, April 10, 2012

Single-pixel camera has multiple futures: Terahertz version adds new potential to unique invention

http://news.rice.edu/2008/10/14/single-pixel-camera-has-multiple-futures/
A terahertz version of the single-pixel camera developed by Rice University researchers could lead to breakthrough technologies in security, telecom, signal processing and medicine.
The research, recently published online in Applied Physics Letters, describes a way to replace the expensive, multipixel sensor arrays used in current terahertz imaging systems with a single sensor.
Two keys to the system are the ongoing development of a modulator that would feed a rapid-fire series of randomized images to the sensor, and the compressed sensing algorithm that turns the raw data into an image.
The advances "could make for very inexpensive security and scientific cameras in the near future," said Richard Baraniuk, Rice's Cameron Professor in Electrical and Computer Engineering, who helped make a proof-of-concept prototype that used 600 sheets of copper (which blocks terahertz radiation) through which random holes had been punched as the modulator.
"There's very good reason to believe you could build a terahertz modulator that could do that same task electrically, and very fast," said Daniel Mittleman, a Rice professor in electrical and computer engineering, who is testing a 4-by-4 array of metamaterials supplied by Los Alamos National Laboratory that become opaque to terahertz radiation when a voltage is applied.
"There are lots of applications for terahertz imaging, if you could make a real-time imager that's sensitive enough. Some of them are pretty science-fictiony, but some are pretty realistic," Mittleman said. "I think this is really promising."
Terahertz radiation, which occupies space in the electromagnetic spectrum between infrared and microwave, penetrates fabric, wood, plastic and even clouds, but not metal or water. Unlike X-rays, T-rays are not harmful, and cheap T-ray cameras may someday be used for security screening in airports, supplementing traditional X-ray scanners and walk-through portals.
In addition, hyperspectral capabilities inherent in even basic single-pixel cameras make them useful for all kinds of things, said Kevin Kelly, associate professor in electrical and computer engineering. "Current cameras break an image down into red, green and blue. But this system breaks down every pixel into all the individual wavelengths that make up a color.
"If you want to know whether that green object over there is a bunch of trees or a tank painted green, this system will tell you," he said.
Rice introduced its research into the single-pixel camera two years ago. The technology made waves when it was introduced, and advances have come quickly, particularly in the compressive sensing algorithms that make it possible to do with one pixel what takes commercial digital cameras millions.
Read the paper, "A Single-Pixel Terahertz Imaging System Based on Compressed Sensing," here:
http://scitation.a … _1-div0.html
Source: Rice University