Zhejiang Sensetime Technology Development Co., Ltd

Huawei Technologies Co., Limited

Samsung Electronics Co., Ltd.

Zhejiang University

Guofeng Zhang

Hujun Bao

Haomin Liu

Ho Young Lee

Jiwon Kim

Du Sik Park

Xiaowei Yang

Yi He

Chengju Chen

Chongshan Sheng

Jiong Zhou

Chen Li

Yazhen Yuan

Jianping Shi

Qinhong Jiang

Renzhi Wang

Xinglun Liang

Kangkan Wang

Wenzhuo Yang

Zhichao Ye

Youji Feng

Jiacheng Jin

Provided is a positioning method and device, an electronic device and a storage medium. The method includes that: an identification image captured at a first site is acquired; according to identification information of a first identifier extracted from the identification image, position information of the first identifier is determined; and position information of the first site is determined based on the position information of the first identifier and projection information of the first identifier in the identification image.

Positioning method, electronic device and storage medium

An information processing method includes: three-dimensional (3D) point information of a 3D point cloud is obtained; a two-dimensional (2D) point cloud image from projection of the 3D point cloud on a horizontal plane is generated based on the 3D point information; and projection coordinates of 3D points comprised in the 3D point cloud in a reference coordinate system of a reference plane graph are determined based on a consistency degree that the 2D point cloud image has with the reference plane graph, where the reference plane graph is used for representing a projection graph with reference coordinates that is obtained through the projection of a target object on the horizontal plane, and the 3D point cloud is used for representing 3D space information of the target object.

Information processing method and device, positioning method and device, electronic device and storage medium

A method for generating a three-dimensional local human body model are provided in embodiments of the disclosure. The method for generating a three-dimensional local human body model includes: acquiring bone parameter adjustment information of a three-dimensional local human body model; adjusting bone parameters of the three-dimensional local human body model according to the bone parameter adjustment information; and generating the three-dimensional local human body model according to the adjusted bone parameters.

Method, apparatus and device for generating three-dimensional local human body model

A simultaneously localization and dense three-dimensional reconstruction method, being capable of processing rapid motion and frequent closing of a loop in a robust manner, and operating at any time in a large-scale scene. Provided by the method is a key frame-based simultaneous localization and map construction framework. First, depth and color information are used simultaneously, and the framework may operate on central processing unit (CPU) at high speed based on localization of the key frame, and operate in challenging scene in robust manner. To reduce accumulated errors, the method introduces increment bundle adjustment, which may greatly reduce an operation amount and enable local and global bundle adjustment to be completed in a unified framework. Secondly, provided by the method is a key frame-based fusion method, and a model may be generated online and a three-dimensional model may be updated in real time during adjustment of the key frame's pose.

Simultaneous localization and dense three-dimensional reconstruction method

A method for simultaneous localization and mapping is provided, which can reliably handle strong rotation and fast motion. The method provided a simultaneous localization and mapping algorithm framework based on a key frame, which can support rapid local map extension. Under this framework, a new feature tracking method based on multiple homography matrices is provided, and this method is efficient and robust under strong rotation and fast motion. A camera orientation optimization framework based on a sliding window is further provided to increase motion constraint between successive frames with simulated or actual IMU data. Finally, a method for obtaining a real scale of a specific plane and scene is provided in such a manner that a virtual object is placed on a specific plane in real size.

Method for simultaneous localization and mapping

A method and apparatus for processing a virtual object, a device, and a storage medium are provided. The method for processing a virtual object includes: acquiring first bone parameters and first blend deformation data corresponding to a preset three-dimensional virtual object model comprising a plurality of bones, the first blend deformation data indicating a degree of deformation of the preset three-dimensional virtual object; acquiring second bone parameters corresponding to a target three-dimensional virtual object model, and determining a transformation relationship of the second bone parameters with the first bone parameters; and determining second blend deformation data corresponding to the target three-dimensional virtual object model, according to the transformation relationship of the second bone parameters with the first bone parameters, and the first blend deformation data.

Method, apparatus and device for processing deformation of virtual object, and storage medium

A depth recovery method includes: performing feature extraction on a monocular image to obtain a feature image of the monocular image; decoupling the feature image to obtain a scene structure graph of the feature image; performing gradient sensing on the feature image and the scene structure graph to obtain a region-enhanced feature image; and performing depth estimation according to the region-enhanced feature image to obtain a depth image of the monocular image.

Depth recovery methods and apparatuses for monocular image, and computer devices

An image processing method includes obtaining multiple video frames, where the multiple video frames are collected from a same scene at different angles and determining a depth map of each video frame according to corresponding pixels among the multiple video frames; supplementing background missing regions of the multiple video frames according to depth maps of the multiple video frames, to obtain supplemented video frames of the multiple video frames and depth maps of the multiple supplemented video frames. The method also includes generating an alpha image of each video frame according to an occlusion relationship between each of the multiple video frames and a supplemented video frame of each video frame in a background missing region and generating a browsing frame at a specified browsing angle according to the multiple video frames, the supplemented video frames of the multiple video frames, and alpha images of the multiple video frames.

Image processing method and apparatus

Embodiments of the present disclosure disclose a static object reconstruction method and system that are applied to the field of graph and image processing technologies. In the embodiments of the present disclosure, when a static object reconstruction system does not obtain, by means of calculation, an extrinsic camera parameter in a preset time when calculating the extrinsic camera parameter based on a three-dimensional feature point, it indicates that depth data collected by a depth camera is lost or damaged, and a two-dimensional feature point is used to calculate the extrinsic camera parameter to implement alignment of point clouds of a frame of image according to the extrinsic camera parameter. In this way, a two-dimensional feature point and a three-dimensional feature point are mixed, which can implement that a static object can also be successfully reconstructed when depth data collected by a depth camera is lost or damaged.

Static object reconstruction method and system

A camera pose tracking apparatus may track a camera pose based on frames photographed using at least three cameras, may extract and track at least one first feature in multiple-frames, and may track a pose of each camera in each of the multiple-frames based on first features. When the first features are tracked in the multiple-frames, the camera pose tracking apparatus may track each camera pose in each of at least one single-frame based on at least one second feature of each of the at least one single-frame. Each of the at least one second feature may correspond to one of the at least one first feature, and each of the at least one single-frame may be a previous frame of an initial frame of which the number of tracked second features is less than a threshold, among frames consecutive to multiple-frames.

Method and apparatus for camera tracking

An image processing apparatus and method may estimate binocular disparity maps of middle views from among a plurality of views through use of images of the plurality of views. The image processing apparatus may detect a moving object from the middle views based on the binocular disparity maps of the frames. Pixels in the middle views may be separated into dynamic pixels and static pixels through detection of the moving object. The image processing apparatus may apply bundle optimization and a local three-dimensional (3D) line model-based temporal optimization to the middle views so as to enhance binocular disparity values of the static pixels and dynamic pixels.

Growing community of inventors

Hangzhou, China

Guofeng Zhang