Multifocal planes head-mounted displays.

Stereoscopic head-mounted displays (HMD's) provide an effective capability to create dynamic virtual environments. For a user of such environments, virtual objects would be displayed ideally at the appropriate distances, and natural concordant accommodation and convergence would be provided. Under such image display conditions, the user perceives these objects as if they were objects in a real environment. Current HMD technology requires convergent eye movements. However, it is currently limited by fixed visual accommodation, which is inconsistent with real-world vision. A prototype multiplanar volumetric projection display based on a stack of laminated planes was built for medical visualization as discussed in a paper presented at a 1999 Advanced Research Projects Agency workshop (Sullivan, Advanced Research Projects Agency, Arlington, Va., 1999). We show how such technology can be engineered to create a set of virtual planes appropriately configured in visual space to suppress conflicts of convergence and accommodation in HMD's. Although some scanning mechanism could be employed to create a set of desirable planes from a two-dimensional conventional display, multiplanar technology accomplishes such function with no moving parts. Based on optical principles and human vision, we present a comprehensive investigation of the engineering specification of multiplanar technology for integration in HMD's. Using selected human visual acuity and stereoacuity criteria, we show that the display requires at most 27 equally spaced planes, which is within the capability of current research and development display devices, located within a maximal 26-mm-wide stack. We further show that the necessary in-plane resolution is of the order of 5 microm.

"KnowWare: virtual reality maps for blind people".

Much has been done to give blind people access to text-based information through computers. Unfortunately, the recent development of Windows operating systems and GUIs jeopardize these gains. Spatial information which is becoming more important to the sighted has never been widely available to blind people. Through a combination of gesture input with audio output, the KnowWare system offers blind users a new way to access spatial information. An invisible virtual map is defined on the desktop. When the user touches a feature on the map with the tip of his index finger, the computer tells him what he has touched by means of computer-generated speech or synthesized sound codes. Touching is detected by means of an overhead video camera that looks down at the user's hands as they rest upon the desktop. Specialized processors analyze the image of the user's hands and locate the fingertips. Because KnowWare is an electronic medium, it is possible to instantly zoom in on a particular state or country to examine it in more detail. Similarly, panning and overlays are possible. Twenty blind subjects have tested KnowWare and have been able to use it as well as they do traditional raised-line maps.

Oxidation and hydrolysis determination of sulfur amino acids in food and feed ingredients: collaborative study.

Samples of 6 food and feed ingredients and a purified protein, beta-lactoglobulin, were analyzed by 7 laboratories to determine the concentrations of cysteine as cysteic acid and methionine as methionine sulfone. Samples were oxidized by reaction with performic acid before hydrolysis with 6N HCl. The free amino acids were then separated and measured by ion-exchange chromatography on dedicated amino acid analyzers. Each laboratory was provided with a detailed method as well as sealed vials containing solutions of standards. For the determination of cysteine as cysteic acid, the coefficients of variation between laboratories for duplicate samples ranged from 7.13 to 10.8% for the 6 ingredients. For the determination of methionine as methionine sulfone, the coefficients of variation between laboratories for duplicate samples ranged from 1.18 to 12.8% for the 6 ingredients. Cysteine and methionine recoveries were determined by analysis of beta-lactoglobulin and were based on expected levels of each amino acid from amino acid sequence data. The mean recovery of cysteine was 95% with a range of 91-101%. The mean recovery of methionine was 101% with a range of 98-106%. This method has been adopted official first action.