Abstract: A method reconstructs at three-dimensional (3D) real-world scene from a single two-dimensional (2D) image by identifying junctions satisfying geometric constraint of the scene based on intersecting lines, vanishing points, and vanishing lines that are orthogonal to each other. Possible layouts of the scene are generated by sampling the 2D image according to the junctions. Then, an energy function is maximized to select an optimal layout from the possible layouts. The energy function use's a conditional random field (CRF) model to evaluate the possible layouts.

Abstract: Poses of a movable camera relative to an environment are obtained by determining point correspondences from a set of initial images and then applying 2-point motion estimation to the point correspondences to determine a set of initial poses of the camera. A point cloud is generated from the set of initial poses and the point correspondences. Then, for each next image, the point correspondences and corresponding poses are determined, while updating the point cloud.

Abstract: Poses of a movable camera relative to an environment are obtained by determining point correspondences from a set of initial images and then applying 2-point motion estimation to the point correspondences to determine a set of initial poses of the camera. A point cloud is generated from the set of initial poses and the point correspondences. Then, for each next image, the point correspondences and corresponding poses are determined, while updating the point cloud.

Abstract: A method and system extract features from an image acquired of an object with a specular surface by first acquiring an image while illuminating the object with a hue circle generated by a set of lights flashed simultaneously. The lights have different colors and are arranged circularly around a lens of a camera. Then, the features correspond to locations of pixels in the image within a neighborhood of pixels that includes a subset of the colors of the lights.

Abstract: A three-dimensional (3D) location of a reflection point of a ray between a point in a scene (PS) and a center of projection (COP) of a camera of a catadioptric system is determined. The catadioptric system is non-central and includes the camera and a reflector, wherein a surface of the reflector is a quadric surface rotationally symmetric around an axis of symmetry. The 3D location of the reflection point is determined based on a law of reflection, an equation of the reflector, and an equation describing a reflection plane defined by the COP, the PS, and a point of intersection of a normal to the reflector at the reflection point with the axis of symmetry.

Abstract: A method registers data using a set of primitives including points and planes. First, the method selects a first set of primitives from the data in a first coordinate system, wherein the first set of primitives includes at least three primitives and at least one plane. A transformation is predicted from the first coordinate system to a second coordinate system. The first set of primitives is transformed to the second coordinate system using the transformation. A second set of primitives is determined according to the first set of primitives transformed to the second coordinate system. Then, the second coordinate system is registered with the first coordinate system using the first set of primitives in the first coordinate system and the second set of primitives in the second coordinate system. The registration can he used to track a pose of a camera acquiring the data.

Abstract: Three-dimensional data are registered by selecting a first set of primitives from the data in a first coordinate system, wherein the first set of primitives includes at least one plane, at least one point, and a third primitive that is either a point or a plane, and selecting a second set of primitives from the data in a second coordinate system, wherein the second set of primitives includes at least one plane, at least one point, and a third primitive corresponding to the third primitive in the first set of primitives. Then, the planes are registered with each other, as are the points, to obtain registered primitives.

Abstract: A visual hull for a 3D object is generated by using a set of silhouettes extracted from a set of images. First, a set of convex polyhedra is generated as a coarse 3D model of the object. Then for each image, the convex polyhedra are refined by projecting them to the image and determining the intersections with the silhouette in the image. The visual hull of the object is represented as union of the convex polyhedra.

Abstract: A single camera acquires an input image of a scene as observed in an array of spheres, wherein pixels in the input image corresponding to each sphere form a sphere image. A set of virtual cameras are defined for each sphere on a line joining a center of the sphere and a center of projection of the camera, wherein each virtual camera has a different virtual viewpoint and an associated cone of rays, appearing as a circle of pixels on its virtual image plane. A projective texture mapping of each sphere image is applied to all of the virtual cameras on the virtual image plane to produce a virtual camera image comprising circle of pixels. Each virtual camera image for each sphere is then projected to a refocusing geometry using a refocus viewpoint to produce a wide-angle lightfield view, which are averaged to produce a refocused wide-angle image.

Abstract: A pose of an object is estimated by first defining a set of pair features as pairs of geometric primitives, wherein the geometric primitives include oriented surface points, oriented boundary points, and boundary line segments. Model pair features are determined based on the set of pair features for a model of the object. Scene pair features are determined based on the set of pair features from data acquired by a 3D sensor, and then the model pair features are matched with the scene pair features to estimate the pose of the object.

Abstract: A three-dimensional (3D) pose of a 3D object in an environment is determined by extracting features from an image acquired of the environment by a camera. The features are matched to a 3D model of the environment to determine correspondences. A camera reference frame of the image and a world reference frame of the environment are transformed to a corresponding intermediate camera reference frame and a corresponding world reference frame using the correspondences. Geometrical constraints are applied to the intermediate camera reference frame and the intermediate world reference frame to obtain a constrained intermediate world reference frame and a constrained world reference frame. The 3D pose is then determined from parameters of the constrained intermediate world reference frame and the constrained world reference frame.

Abstract: An image is segmented into superpixels by constructing a graph with vertices connected by edges, wherein each vertex corresponds to a pixel in the image, and each edge is associated with a weight indicating a similarity of the corresponding pixels, A subset of edges in the graph are selected to segment the graph into subgraphs, wherein the selecting maximizes an objective function based on an entropy rate and a balancing term. The edges with maximum gains are added to the graph until a number of subgraphs is equal to some threshold.

Abstract: A location and orientation in an environment is determined by acquiring a set of one or more real omni-directional images of an unknown skyline in the environment from an unknown location and an unknown orientation in the environment by an omni-directional camera. A set of virtual omni-directional images is synthesized from a 3D model of the environment, wherein each virtual omni-directional image is associated with a known skyline, a known location and a known orientation. Each real omni-directional image is compared with the set of virtual omni-directional images to determine a best matching virtual omni-directional image with the associated known location and known orientation that correspond to the unknown location and orientation.

Abstract: An image is segmented into superpixels by constructing a graph with vertices connected by edges, wherein each vertex corresponds to a pixel in the image, and each edge is associated with a weight indicating a similarity of the corresponding pixels, A subset of edges in the graph are selected to segment the graph into subgraphs, wherein the selecting maximizes an objective function based on an entropy rate and a balancing term. The edges with maximum gains are added to the graph until a number of subgraphs is equal to some threshold.

Abstract: A three-dimensional (3D) location of a reflection point of a ray between a point in a scene (PS) and a center of projection (COP) of a camera of a catadioptric system is determined. The catadioptric system is non-central and includes the camera and a reflector, wherein a surface of the reflector is a quadric surface rotationally symmetric around an axis of symmetry. The 3D location of the reflection point is determined based on a law of reflection, an equation of the reflector, and an equation describing a reflection plane defined by the COP, the PS, and a point of intersection of a normal to the reflector at the reflection point with the axis of symmetry.

Abstract: A location and orientation in an environment is determined by first acquiring a real omni-directional image of an unknown skyline in the environment. A set of virtual omni-directional images of known skylines are synthesized from a 3D model of the environment, wherein each virtual omni-directional image is associated with a known location and orientation. The real omni-directional image with each virtual omni-directional images to determine a best matching virtual omni-directional image with the associated known location and orientation.

Abstract: Three-dimensional points acquired of an object in a sensor coordinate system are registered with planes modeling the object in a world coordinate system by determining correspondences between the points and the planes. Points are transformed to an intermediate coordinate system using the correspondences and transformation parameters. The planes are transformed to an intermediate world coordinate system using world rotation and translation parameters. Intermediate rotation and translation parameters are determined by applying coplanarity constraints and orthogonality constraints to the points in the intermediate sensor coordinate system and the planes in the intermediate world coordinate system. Then, rotation and translation parameters between the sensor and world coordinate systems are determined to register the points with the planes.

Abstract: A pose of an object is determine by acquiring sets of images of the object by a camera, wherein the object has a thread arranged on a surface such that a local region of the object appears substantially spherical, wherein the camera is at a different point of view for each set, and wherein each image in each set is acquired while the scene is illuminated from a different direction. A set of features is extracted from each image, wherein the features correspond to points on the surface having normals towards the camera. A parametric line is fitted to the points for each image, wherein the line lies on a plane joining a center of the camera and an axis of the object. Then, geometric constraints are applied to lines to determine the pose of the object.

Abstract: Embodiment of invention discloses a system and a method for determining a three-dimensional (3D) location of a folding point of a ray between a point in a scene (PS) and a center of projection (COP) of a camera of a catadioptric system. One embodiment maps the catadioptric system, including 3D locations of the PS and the COP on a two-dimensional (2D) plane defined by an axis of symmetry of a folding optical element and the PS to produce a conic and 2D locations of the PS and COP on the 2D plane, and determines a 2D location of the folding point on the 2D plane based on the conic, the 2D locations of the PS and the COP. Next, the embodiment determines the 3D location of the folding point from the 2D location of the folding point on the 2D plane.

Abstract: Embodiment of invention discloses a system and a method for determining a three-dimensional (3D) location of a folding point of a ray between a point in a scene (PS) and a center of projection (COP) of a camera of a catadioptric system. One embodiment maps the catadioptric system, including 3D locations of the PS and the COP on a two-dimensional (2D) plane defined by an axis of symmetry of a folding optical element and the PS to produce a conic and 2D locations of the PS and COP on the 2D plane, and determines a 2D location of the folding point on the 2D plane based on the conic, the 2D locations of the PS and the COP. Next, the embodiment determines the 3D location of the folding point from the 2D location of the folding point on the 2D plane.