OSA-SPIE presents Dr. Joseph N. Mait of the US Army Research Laboratory and his colloquium talk “Millimeter Wave Computational Imaging” on Tues., January 27, at 4:10 pm in Webster B17.
Abstract: Imaging with millimeter waves embodies the worst of two worlds. At long wavelength radio frequencies (> 1 cm) one can measure phase and image coherently relatively inexpensively. At millimeter wavelengths (GHz frequencies), measuring phase is costly. Components, such as amplifiers, must be low noise, which increases cost. At visible wavelengths imaging is so inexpensive, consumers scoff at any smartphone, tablet, or PC without a camera. Plastic lenses and high-count pixel arrays make incoherent imaging a commodity. Physics limits the size of a millimeter image produced by a 60-cm aperture to approximately 200 x 200. That is, megapixel images are not possible even for a vehicle-borne system. Moreover, the technology does not yet exist for integrated detector arrays and inexpensive optics.
Given that measurements at millimeter wavelengths are few and costly, to create images requires measurements that are rich in information. This is the realm of computational imaging. Computational imaging attempts to break conventional paradigms in imager design by combining optics with electronics in an optimal fashion for image and sensor processing. In a computational imaging system the optics and signal processing are no longer designed independently. Further, describing the system as computational underscores the fact that the burden of image formation does not fall solely on the optics.
I present several examples of computational imaging applied to millimeter waves, including extended depth-of-field imaging, compressive holography, and compression applied to extended depth-of-field imaging.
Please meet our guest speaker and share refreshments, 3:45-4:10 p.m. in the foyer on floor G above the lecture hall.