Handheld portable optical scanner and method of using

摘要

A system and methods for real-time or near-real time processing and post-processing of RGB-D image data using a handheld portable device and using the results for a variety of applications. The disclosure is based on the combination of off-the-shelf equipment (e.g. an RGB-D camera and a smartphone/tablet computer) in a self-contained unit capable of performing complex spatial reasoning tasks using highly optimized computer vision algorithms. New applications are disclosed using the instantaneous results obtained and the wireless connectivity of the host device for remote collaboration. One method includes steps of projecting a dot pattern from a light source onto a plurality of points on a scene, measuring distances to the points, and digitally reconstructing an image or images of the scene, such as a 3D view of the scene. A plurality of images may also be stitched together to re-position an orientation of the view of the scene.

主权项

1. A mobile handheld imaging system for creating a digital 3D model of the environment by capturing a plurality of images of a scene, in real time, comprising:
an image/depth capture module for capturing a plurality of digital images to obtain depth information and color information;a handheld computer system having a processor, a display with user interface controls, and a communications module, wherein the processor is in communication with the image/depth capture module; anda non-transitory computer readable medium having encoded thereon a set of instructions executable by the processor to perform one or more operations, comprising:
capturing the plurality of digital images of a scene; utilizing an initial digital image of the plurality of digital images to initialize a 3D model; and
computing an estimate of position and orientation (pose) for the image/depth capture module at time of capture for each of the plurality of captured digital images acquired from the image/depth capture module subsequent to utilizing the initial digital image of the plurality of digital images to initialize the 3D model, by registering each of the plurality of captured digital images against at least a subset of the 3D model represented digitally on the non-transitory computer readable medium; extending the 3D model by transforming a location of new points from at least one of a plurality of the subsequent captured digital images captured subsequent to the initial digital image into the coordinate system; of the 3D model by using the subsequent captured digital image's pose, and adding the new points into the 3D model if the overlap of the subsequent captured digital image with the 3D model falls below a threshold, otherwise discarding at least one of the plurality of new points of the subsequent captured digital image; using the estimates of poses, and the depth information and color information of all of the plurality of captured digital images which have been used to initialize or extend the 3D model, and that together represent the 3D model, as an input to a mathematical global optimization process which optimizes the input poses and the depth information and color information by global registration of the 3D model thereby increasing the accuracy and optical quality of the captured 3D model; and displaying a rendering of the plurality of captured digital images as a visualization of the 3D model in the mobile handheld imaging system.