Google Inc.

Sameer Agarwal

Steven Maxwell Seitz

David Robert Gallup

Dragomir Dimitrov Anguelov

Christian Frueh

Carlos Hernandez Esteban

Simon Fuhrmann

Avanish M Kushal

Systems and methods for the generation of depth data for a scene using images captured by a camera-enabled mobile device are provided. According to a particular implementation of the present disclosure, a reference image can be captured of a scene with an image capture device, such as an image capture device integrated with a camera-enabled mobile device. A short video or sequence of images can then be captured from multiple different poses relative to the reference scene. The captured image and video can then be processed using computer vision techniques to produce an image with associated depth data, such as an RGBZ image.

Computer vision algorithm for capturing and refocusing imagery

A system and machine-implemented method for providing one or more photos associated with a point of interest on a map, the method including receiving an indication of a request from a user to view photos associated with a point of interest on a map, identifying a set of photos associated with the point of interest, wherein the photos comprise at least one of photos taken from the point of interest or photos that depict at least part of the point of interest, ranking the photos within the set of photos according to ranking criteria, wherein the ranking criteria comprises one or more of map context, photo quality, photo type or user request information and providing one or more photos of the set of photos to the user according to the ranking.

System and method for selecting user generated content related to a point of interest

Systems and methods for generating image tour are provided. Method includes constructing image graph comprising primary image nodes and secondary image nodes and edges. Method also includes determining, for each pair of primary nodes, pruned subgraph including pair of primary nodes and first subset of plurality of secondary nodes. Method also includes determining order of plurality of primary nodes based on rendering costs in pruned subgraphs. Method also includes splicing pruned subgraphs together according to determined order of primary nodes to generate spliced graph. Method also includes determining path through spliced graph. Path includes plurality of primary nodes in determined order and second subset of plurality of secondary nodes selected based on rendering costs and turning costs associated with transitioning between pairs of edges in spliced graph. Method also includes providing ordered subset of image set based on determined path for display as image tour on client.

Generating an image tour based on a set of images

A three-dimensional model of a scene is reconstructed from images of the scene. The three-dimensional model comprises parameters describing points of the scene and cameras that captured the scene. The parameters are iteratively modified by adding a correction to the parameters in each iteration. A preconditioner matrix is determined for solving the linear system of equations. To determine the preconditioner matrix, cameras that observed the scene are clustered, each cluster including cameras that observed a portion of the scene. The preconditioner matrix is determined by selectively eliminating terms of the linear system of equations that correspond to interactions across clusters of cameras. The linear system of equations is preconditioned using the preconditioner matrix. The preconditioned linear system of equations is solved to obtain the corrections to the parameters.

Preconditioner for solving linear equations for reconstructing three-dimensional structure of a scene

A system for pose generation consisting of a trajectory system, a pose generation system, an intersection extractor, an object identifier, a constraint generator, and a posegraph solver. The trajectory system identifies a number of trajectories based on input positional data of a bounded area. The pose generation system generates one or more poses based on the trajectories. The intersection extractor identifies one or more possible intersections in the one or more poses. The object identifier identifies an object pair for each possible intersection that represents two positional points at each possible intersection. The constraint generator computes and applies one or more intersection constraints to generate an energy value for each object pair based on their geometric relationship. The posegraph solver then minimizes a total energy value by modifying one or more poses and then generates an improved set of pose trajectories based on the modified one or more poses.

Using image and laser constraints to obtain consistent and improved pose estimates in vehicle pose databases

A system for pose generation consisting of a trajectory system, a pose generation system, an intersection extractor, an object identifier, a constraint generator, and a posegraph solver. The trajectory system identifies a number of trajectories based on input positional data of a bounded area. The pose generation system generates one or more poses based on the trajectories. The intersection extractor identifies one or more possible intersections in the one or more poses. The object identifier identifies an object pair for each possible intersection that represents two positional points at each possible intersection. The constraint generator computes and applies one or more intersection constraints to generate an energy value for each object pair based on their geometric relationship. The posegraph solver then minimizes a total energy value by modifying one or more poses and then generates an improved set of pose trajectories based on the modified one or more poses. A method of pose generation is also presented.

Growing community of inventors

Lake Forest Park, WA, United States of America

Sameer Agarwal