Abstract: A method for three-dimensional (3D) image reconstruction includes the steps of registering an orthographic image of a scene, combining a photogrammetric image and a technical drawing of the scene to form a co-registered orthographic and perspective (COP) image, and reconstructing a 3D image from the COP image.

Abstract: A monocular input image is transformed to give it an enhanced three dimensional appearance by creating at least two output images. Foreground and background objects are segmented in the input image and transformed differently from each other, so that the foreground objects appear to stand out from the background. Given a sequence of input images, the foreground objects will appear to move differently from the background objects in the output images.

Abstract: A method and apparatus for high-performance rendering and hit-testing of a window tree is provided. A window tree may be rendered using an application programming interface provided by the present invention. The application programming interface provides support for world-transforms, enabling entire sub-trees of the window tree to be rotated and scaled during rendering. In order to quickly render and hit-test the transformed nodes of the window tree, a stack-based implementation of the “painter's algorithm” is utilized to achieve fast rendering. By storing all state information on a stack regarding each node in the window tree and building new data structures containing rendering information for each node and its children, any portion of the sub tree may be rendered on demand.

Abstract: A method determines mappings between high-dimensional measured samples and a reduced-dimensional manifold. Samples are acquired of a physical system. The samples have a high number of dimensions. A low number of dimensions are determined for a manifold embedded in a space of the high-dimensional samples. Local charts having the dimensions of the low number of dimensions of the manifold are fitted to selected high-dimensional samples. The charts are then connected to determine a forward mapping from any high-dimensional sample to a coordinate on the manifold and a reverse mapping from any coordinate on the manifold to a corresponding point in high-dimensional sample space.

Abstract: A photo realistic talking head creation system and method comprising: a template; a video camera having an image output signal of a subject; a mixer for mixing the template and the image output signal of the subject into a composite image, and an output signal representational of the composite image; a prompter having a partially reflecting mirror therebetween the video camera and the subject, an input for receiving the output signal of the mixer representational of the composite image, the partially reflecting mirror adapted to allow the video camera to collect the image of the subject therethrough and the subject to view the composite image and to align the image of the subject therewith the template; storage means having an input for receiving the output image signal of the video camera representational of the collected image of the subject, and storing the image of the subject substantially aligned therewith the template.

Abstract: Disclosed herein are a method for generating volume data from images of an object viewed from diverse view points. This method comprises the steps of: (a) in each of discrete coordinate positions within a coordinate system for rendering a volume, obtaining a value corresponding to the distance to the boundary of a viewed volume defined by the contour of the image of said object and by the point of view of said image; and (b) in each of said coordinate positions, on the basis of a plurality of said values obtained for said images of said object, determining a value corresponding to the distance to the object surface and then maintaining said value in a manner corresponded to said coordinate position. In the method, multi-valued data can be assigned to each vertex or lattice point, and therefore, a small number of voxels can represent precisely the three-dimensional shape.

Abstract: A method of visualizing the impact of security flaws or breaches in a network. A 3-D visualization tool that simulates 3-D space on a monitor interfaces with a security database which relationally associates security events with the network elements affected thereby. The security events are visually depicted in a first section of simulated 3-D space and the network elements are depicted in a second section of simulated 3-D space. Relationship lines are drawn between displayed categories of security events and the displayed network elements in order to aid an analyst to visualize the impact of security breaches on the organization. Various other properties of the network elements may also be displayed such as the role of the network device within the organization, and the business functions of the organization.

Abstract: A system and method for generating a visualized data representation preserving independent variable geometric relationships is described. A pair of convex clusters are selected. Each convex cluster is rendered on a display. Each convex cluster has a center of mass located at an original fixed distance from a common origin, and is oriented along a vector formed at a fixed angle from a common polar axis. A span is measured between the centers of mass of each convex cluster. For each convex cluster, a segment is measured from the center of mass of each convex shape to a point closest to the other convex shape along the span. A new fixed distance from the common origin for the center of mass for one of the convex clusters located along the vector for that convex cluster is evaluated if the span is less than the sum of the segments of the convex clusters. The pair of convex clusters are displayed rendered using at least the new fixed distance for the center of mass of the one convex cluster.

Abstract: A method and apparatus is disclosed for generating a m-degree form in a n-dimension space. A framework is used to generate the form. The framework comprises a defined number of sections, lines and control points depending on the complexity of the form to generate. An iterative solving process is then used in order to provide the algebraic coefficients of the m-degree form using at least data related to the control points.

Abstract: An image processing apparatus and an image processing program capable of sufficiently representing changes in undulations of a water surface without deforming the surface of an object. A water surface object B is so set as to be larger than a water surface area to be actually represented and to has undulations formed in the neighborhood of a boundary line portion meeting a land object A. The place position is determined so that the position of this entire water surface object B is moved at predetermined time (frame) intervals. Thus, it is possible to change in shape only the boundary portion where the water surface and the land meet, and to represent undulations of the water surface as if water is undulating. Furthermore, the shapes of polygons composing the water surface object B are not individually changed, but the entire water surface object B having a particular shape is moved.

Abstract: An apparatus is disclosed for producing an animation of a group including a plurality of members which are dynamically controlled.

Abstract: There is provided a method for augmented reality guided instrument positioning. A graphics guide is determined for positioning an instrument that provides a substantially unobstructed view of the instrument during the positioning. The graphics guide is rendered in alignment with a predetermined path to be traversed during the positioning of the instrument.

Abstract: A technique for solving an inverse-kinematic problem by interpolating solutions from examples. Example poses or motions of an object are collected and annotated. The annotations are essentially parameters for a function—i.e., the function X(p) generates degree-of-freedom values of an object that is posed in a manner that satisfies parameters p. The analytic function X is interpolated from these examples and improved automatically based on kinematic measurements. Preferably, the interpolation is created by taking a weighted sum of cardinal basis functions having linear and radial parts. Preferably, the interpolation is a weighted sum of cardinal basis functions having linear and radial portions.

Abstract: An omnidirectional video camera captures images of the environment while moving along several intersecting paths forming an irregular grid. These paths define the boundaries of a set of image loops within the environment. For arbitrary viewpoints within each image loop, a 4D plenoptic function may be reconstructed from the group of images captured at the loop boundary. For an observer viewpoint, a strip of pixels is extracted from an image in the loop in front of the observer and paired with a strip of pixels extracted from another image on the opposite side of the image loop. A new image is generated for an observer viewpoint by warping pairs of such strips of pixels according to the 4D plenoptic function, blending each pair, and then stitching the resulting strips of pixels together.

Abstract: Images of a reference object for detecting each camera position are captured by cameras 14, 16 and the relative position between the cameras 14 and 16 is obtained by using the captured images. Then, a target 18 is placed on a turntable 10 onto which reference marks 10a for detecting position are formed, and while the turntable 10 is rotated, images are captured by the cameras 14 and 16. Based on the reference marks 10a on the turntable 10, the positional relationship between the camera 14 and the target 18 is determined, and, based on the relative position of the cameras 14 and 16, the positional relationship between the target 18 and the camera 16 is determined. A three-dimensional model of the target is produced based on the images captured by the cameras 14 and 16. In this manner, a target of relatively large size is easily three-dimensionally modeled.

Abstract: Animation data is produced in a data processing system that has storage, a processor, a visual display (102) and a manual input device (103). A simulated three-dimensional world-space is displayed to a user and an animatable actor is displayed in the world-space. First input data is received from a user specifying a desired location and orientation of the actor to which the actor is to animate towards from a starting location. The processor is instructed to generate first animation data using a first processing technique possibly using a forward kinematic solver. The processor is instructed to generate second animation data using a second processing technique, in response to the same input data, possibly using an inverse kinematic solver. Second input data defines a blend parameter for blending between the first animation data and the second animation data.

Abstract: Provided is a method and apparatus for representation and rendering of three-dimensional still and animated objects, using Binary Volumetric Octree. The three-dimensional object representation based on depth images, requiring relatively small storage space and allowing for fast and high quality rendering.