Qualcomm Patent | Scalable 3D mapping system

Patent: Scalable 3D mapping system

Publication Number: 10185775

Publication Date: 2019-01-22

Applicants: Qualcomm

Abstract

Disclosed are a system, apparatus, and method for multiple client simultaneous localization and mapping. Tracking and mapping may be performed locally and independently by each of a plurality of clients. At configurable points in time map data may be sent to a server for stitching and fusion. In response to successful stitching and fusion to one or more maps known to the server, updated position and orientation information relative to the server’s maps may be sent back to the clients. Clients may update their local map data with the received server location data. Clients may receive additional map data from the server, which can be used for extending their maps. Clients may send queries to the server for 3D maps, and the queries may include metadata.

Background
Computer vision is a field that includes methods and systems for acquiring, analyzing, processing, and understanding images (e.g., real world image captures) to provide an event or result. For example, one computer vision technique is Simultaneous Localization and Mapping (SLAM), which can process the input of a single camera and continuously build up a three dimensional (3D) model (e.g., reconstructed map) of an environment as the camera moves in Six Degrees of Freedom (6DOF). SLAM systems can simultaneously track the pose of the camera with respect to the 3D model while mapping the 3D model. Keyframe-based visual SLAM systems can process discretely selected frames from the incoming camera image stream or feed. Keyframe-based visual SLAM systems assume general camera motion and apply structure-from-motion techniques to create 3D feature maps.

Modern keyframe-based computer vision (e.g., SLAM) systems subdivide work into parallel tracking and mapping (PTAM) threads. Both the tracking and mapping threads may be processed in parallel, but asynchronously. The tracking thread may perform at a full frame rate, while mapping is typically more computationally intensive and thus slower. Scaling computer vision to large areas and letting multiple clients/users or robots participate in the processing of computer vision work creates the need for stitching two or more separate map pieces/sections together. In general, stitching refers to the discovery of overlapping portions from two or more maps and determining the corresponding 7DOF similarity transform (composed of a 3DOF orientation, a 3DOF position, and a 1D scale). If one of the maps covers a much larger area than the other, this is sometimes called “place recognition.” After successful stitching or place recognition, map fusion may be performed. Map fusion, or simply “fusion” typically describes the processing of data or information from separate maps to combine into a single map. For example fusion may be performed with a form of Structure from Motion (SfM) technique applied to the image information from the separate source maps.

With respect to some types of computer vision techniques, the four tasks of tracking, mapping, stitching, and fusion may have increasing computational requirements as additional data or information is processed. To support many maps, a single and independent user/client may be unable to process all data associated with tracking, mapping, stitching, and fusion. However, offloading mapping to a server may cause clients to become reliant upon the server for content. Clients may rely upon the connection to the server to generate real-time local map dependent content. For example, such content may be used in Augmented Reality (AR) applications. Additionally, maps on servers are typically not scalable or well organized. Therefore, improved techniques are desirable.

Summary

Embodiments disclosed herein may relate to a method performed by a server for implementing a 3D map database. The method may include: receiving one or more keyframes and associated metadata from a client; storing the received one or more keyframes in the 3D map database; tagging the received one or more keyframes in the 3D map database according to the received metadata; receiving a client request for map data; searching the 3D map database for map data according to the client request; and providing, in response to the searching, a resulting map data to the client requesting the map data.

Embodiments disclosed herein may relate to a server for implementing a 3D map database. The server may be configured to: receive one or more keyframes and associated metadata from a client; store the received one or more keyframes in the 3D map database; tag the received one or more keyframes in the 3D map database according to the received metadata; receive a client request for map data search the 3D map database for map data according to the client request; and provide, in response to the searching, a resulting map data to the client requesting the map data.

Embodiments disclosed herein may relate to a machine readable non-transitory storage medium having stored therein program instructions that are executable by a server’s processor to: receive one or more keyframes and associated metadata from a client; store the received one or more keyframes in the 3D map database; tag the received one or more keyframes in the 3D map database according to the received metadata; receive a client request for map data search the 3D map database for map data according to the client request; and provide, in response to the searching, a resulting map data to the client requesting the map data.

Embodiments disclosed herein may relate to an apparatus for implementing a 3D map database. The apparatus may include: means for receiving one or more keyframes and associated metadata from a client; means for storing the received one or more keyframes in the 3D map database; means for tagging the received one or more keyframes in the 3D map database according to the received metadata; means for receiving a client request for map data; searching the 3D map database for map data according to the client request; and means for providing, in response to the searching, a resulting map data to the client requesting the map data.

发表评论

电子邮件地址不会被公开。 必填项已用*标注