Dinesh K Pai | PubMed Health

PubMed2007

Polarization multiplexing and demultiplexing for appearance-based modeling.

Polarization has been used in numerous prior studies for separating diffuse and specular reflectance components, but in this work we show that it also can be used to separate surface reflectance contributions from individual light sources. Our approach is called polarization multiplexing and it has a significant impact in appearance modeling where the image as a function of illumination direction is needed. Multiple unknown light sources can illuminate the scene simultaneously, and the individual contributions to the overall surface reflectance are estimated. Polarization multiplexing relies on the relationship between the light source direction and the intensity modulation. Inverting this transformation enables the individual intensity contributions to be estimated. In addition to polarization multiplexing, we show that phase histograms from the intensity modulations can be used to estimate scene properties including the number of light sources.

Algorithms↗

PubMed2004

Surface extraction with a three-dimensional freehand ultrasound system.

This paper presents a system for acquiring three-dimensional ultrasound data and extracting surfaces of the examined structures. The acquisition is performed freehand with a PC-based two dimensional ultrasound machine and an optical tracker. The extraction of surfaces from ultrasound data are normally inhibited by speckle, shadowing and gaps in the acquired data. A new method is developed that extracts a surface directly from the irregularly spaced, noisy freehand ultrasound data. The freehand data are first interpolated with radial basis functions and then a mesh is formed along an isosurface of the functional interpolation. The ability of radial basis functions to smooth speckle and interpolate across gaps is demonstrated on a series of experiments with phantoms and human tissue in a water bath. The geometry of the extracted surface matches the external measurements with an average difference ranging from 0.8 to 2.9 mm. These differences are within the range of errors from calibration, resolution and landmark localization. The experiments also show the ability to create continuous and realistic surfaces from scans that require multiple sweeps over a structure.

Animals↗