In 1992 research physicist Rick Puetter, working with graduate student Robert Piña, began developing a computational approach for improving astronomical images captured with new mid-infrared cameras then under development at the University of California, San Diego. Their goal was to undo the blurring effects of optical diffraction in the telescope-an unavoidable corruption of image quality due to the basic properties of light-without generating false features unrepresented in the original raw data. In this way, significantly more detail could be extracted from the original data with high confidence that no false details, hence misleading scientific results, were being created in the process. The final result of this effort was the Pixon Method, a revolutionary new way to extract the maximum amount of justifiable information contained in an image data set.
While the Pixon Method had yielded spectacular results in a number of initial test cases, the slow speed of the numerical algorithm used to implement the method's principles threatened to limit its applications to either small images or to those few projects where very powerful computers could be used to run the software. Even then, computation times could be days or longer.
In 1995, Amos Yahil, an astrophysicist at the Stony Brook University, devised a new algorithm that decreased Pixon Method computation times by a factor of 10,000 or more. Recognizing that the improved techniques could be applied across a wide range of fields, Puetter and Yahil formed Pixon LLC in 1997 and began offering to a broad spectrum of end users an entirely new and superior means for getting the most from their images.
Shortly after its formation, Pixon introduced Pixon Software, a suite of flexible software tools that implement the Pixon Method either as a general image-processing application on a stand-alone workstation or as a specialized image-processing solution to be embedded within specific imaging or video systems. In 1999, Pixon, in collaboration with DigiVision, Inc., a leader in real-time video products, developed PixonVision, a dedicated hardware implementation of the Pixon Method capable of processing standard video in real-time. That same year, Pixon Method was granted a US patent for its core image processing technology.
Leveraging the company's expertise in information theory, Pixon recently developed the Algebron Method, a wholly new approach to the systematic analysis of complex historical data that enables efficient characterization and forecasting of future outcomes.
Today, Pixon's solutions support critical imaging applications in defense, security, medical care, microscopy-truly in any field where pictures make the difference.
|
