CN113362467A - Point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method - Google Patents
Point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method Download PDFInfo
- Publication number
- CN113362467A CN113362467A CN202110634620.0A CN202110634620A CN113362467A CN 113362467 A CN113362467 A CN 113362467A CN 202110634620 A CN202110634620 A CN 202110634620A CN 113362467 A CN113362467 A CN 113362467A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- target
- mobile terminal
- calculating
- pose
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T15/00—3D [Three Dimensional] image rendering
- G06T15/005—General purpose rendering architectures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
Abstract
The invention discloses a mobile terminal three-dimensional pose estimation method based on point cloud preprocessing and ShuffleNet, which comprises the steps of preprocessing at a PC (personal computer) terminal, performing three-dimensional reconstruction on target point cloud data and importing the data into a three-dimensional rendering engine; obtaining two-dimensional photos of a target under different visual angles by adopting a rotation photographing algorithm in a three-dimensional engine, labeling the photos by using a key voxel block extraction algorithm provided by the invention, and establishing a training data set; the detection model of the key voxel blocks of the target is trained by using ShuffleNetv2-YOLOv3 which has light weight and high performance advantages and is suitable for mobile terminal calculation; reading a video stream from a mobile terminal camera, detecting a target key voxel block through a ShuffleNetv2-YOLOv3 model, and calculating a 2D-3D point pair corresponding to a key voxel block central point through RANSAC and EPNP algorithm to obtain a relative pose of a target. And finally, calculating the pose of the target in the actual three-dimensional world by using the advantages of the mobile terminal through data provided by the built-in IMU and the GPS.
Description
Technical Field
The invention belongs to the technical field of computers, and relates to a mobile terminal three-dimensional pose estimation method based on point cloud preprocessing and ShuffleNet, which can be widely applied to the fields of robot grabbing, vehicle intelligent navigation, augmented reality, medical diagnosis and the like.
Background
The three-dimensional pose estimation plays a very key role in the fields of robot grabbing, vehicle intelligent navigation, augmented reality, medical diagnosis and the like. At present, the mainstream pose estimation methods are divided into two categories, one category is an identification method based on a two-dimensional image, the method predicts 1 central point and 8 angular points of an object for an input RGB or RGB-D image, and then obtains the 6D pose of the object through a PNP or EPNP algorithm. The algorithm has good real-time performance but low accuracy. And the other type of the method is used for positioning based on point cloud data, the method firstly uses a deep network to establish a corresponding relation between 3D point cloud data and a 2D image, and then obtains the 6D posture of the object through a PNP or EPNP algorithm. Due to the use of point cloud data, the accuracy is higher than the first class, but comparatively, the speed is lower.
The mobile terminal of the mobile phone has the advantages of high popularization rate and convenience in carrying, but the hardware configuration is far lower than that of a PC (personal computer), and the recognition speed by adopting a conventional algorithm is difficult to meet the requirement. The portable advantages of the laser radar and the depth camera which need to be externally connected are weakened in configuration, and the mobile end only can adopt an RGB video stream identification scheme, so that the pose analysis accuracy is not high.
The invention mainly aims at the requirement of a mobile terminal on target pose estimation in the field of auxiliary industrial application, and provides a mobile terminal three-dimensional pose estimation method based on point cloud preprocessing and ShuffleNet.
The design scheme adopted by the invention is as follows: a mobile terminal three-dimensional pose estimation method based on point cloud preprocessing and ShuffleNet comprises the following steps:
step 1: performing three-dimensional reconstruction on target point cloud data obtained by laser scanning; importing a three-dimensional model obtained by three-dimensional reconstruction into a rendering engine for photographing;
step 2: respectively acquiring two-dimensional photos and camera poses of the target under different visual angles by adopting a positioning rotation photographing algorithm; extracting two-dimensional photo feature points through SIFT, calculating corresponding three-dimensional feature points, dividing a target model into equal-size voxel blocks, and screening target key voxel blocks according to the number of the three-dimensional feature points; generating two-dimensional projection of key voxel blocks on the photo set and establishing a training data set; training a ShuffLeNet feature detection model aiming at a target through a ShufflLeNet v2-YOLOv3 lightweight network;
and step 3: inputting the video stream into a trained detection model of ShuffleNetv2-YOLOv3 target key voxel blocks, identifying the key voxel blocks to obtain 2D-3D matching point pairs, and calculating the relative pose of the target by combining RANSAC and EPNP algorithms;
and 4, step 4: and calculating the absolute pose of the target in the three-dimensional world by combining the GPS and IMU information of the mobile terminal.
The invention combines the advantages of two types of three-dimensional pose estimation algorithms, and firstly carries out preprocessing at the PC end. And reconstructing a three-dimensional model for the target point cloud data through a Delauay algorithm. The method adopts a positioning rotation photographing and key voxel block extraction algorithm to automatically generate a target voxel characteristic detection data set, and adopts a ShuffleNetv2-YOLOv3 training characteristic detection model which has the advantages of light weight and high performance and is suitable for mobile-end calculation.
The invention fully utilizes the advantages of the mobile terminal hardware equipment in the identification stage and introduces GPS and IMU data to position the pose of the mobile terminal. And detecting key voxel blocks of the target by using a trained ShuffleNetv2-YOLOv3 model, and calculating by using RANSAC and EPNP algorithms to obtain the relative pose between the target and the mobile terminal camera. And finally, calculating the absolute pose of the target in the three-dimensional world. At present, the popularity of the mobile terminal of the mobile phone exceeds 90%, the invention can provide position and pose estimation which does not depend on a depth camera and laser equipment and meets the requirements of industrial auxiliary application in real time and accuracy at the mobile terminal, and has the advantages of portability, practicability and easy popularization.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention.
Fig. 2 is a schematic block diagram of an embodiment of the present invention.
Detailed Description
In order to facilitate the understanding and implementation of the present invention for those of ordinary skill in the art, the present invention is further described in detail with reference to the accompanying drawings and examples, it is to be understood that the embodiments described herein are merely illustrative and explanatory of the present invention and are not restrictive thereof.
Referring to fig. 1 and fig. 2, the method for estimating the three-dimensional pose of the mobile terminal based on point cloud preprocessing and ShuffleNet provided by the invention comprises the following steps:
step 1: performing three-dimensional reconstruction on target point cloud data obtained by laser scanning; importing a three-dimensional model obtained by three-dimensional reconstruction into a rendering engine for photographing;
the specific implementation of the embodiment includes the following sub-steps:
step 1.1: performing three-dimensional reconstruction on target point cloud data obtained by laser scanning through a Delauay algorithm to obtain a target three-dimensional model;
step 1.2: and importing the target three-dimensional model into a rendering engine, calculating a target three-dimensional model bounding box and a central point, and moving the target three-dimensional model to enable the central point to reach an origin point.
Step 2: respectively acquiring two-dimensional photos and camera poses of the target under different visual angles by adopting a positioning rotation photographing algorithm; extracting two-dimensional photo feature points through SIFT, calculating corresponding three-dimensional feature points, dividing a target model into equal-size voxel blocks, and screening target key voxel blocks according to the number of the three-dimensional feature points; generating two-dimensional projection of key voxel blocks on the photo set and establishing a training data set; training a ShuffLeNet feature detection model aiming at a target through a ShufflLeNet v2-YOLOv3 lightweight network;
the specific implementation of the embodiment includes the following sub-steps:
step 2.1: shooting a target in a three-dimensional engine to obtain a target photo;
step 2.2: detecting two-dimensional feature points in the target picture through an SIFT algorithm to obtain a feature point set K ═ K1,...,kn}；
Step 2.3: calculating each characteristic point k by a screen ray projection algorithmiCorresponding three-dimensional coordinate point piAnd recording the three-dimensional feature point set corresponding to the two-dimensional feature point as P ═ P1,...,pn}；
Step 2.4: the camera rotates around the target and continues to take pictures of the target, and the step 2.2 and the step 2.3 are repeated until a multi-view picture of the target is obtained and a three-dimensional feature point set PS is obtained through calculation1,...,PNN is the number of photos, and P is a three-dimensional feature point set of each photo;
step 2.5: dividing a target three-dimensional voxel (Volume Pixel) into M blocks B ═ B of the same voxel size1,...,bM}; setting the frequency of the three-dimensional feature points appearing in each voxel block as a voxel block weight q, and screening m blocks KB with the maximum weight as { b }1,...,bmAs key voxel block, where m<M；
Step 2.6: taking the key voxel blocks as categories, calculating the areas of the key voxel blocks on the two-dimensional photo set according to a projection transformation formula, and generating marking information to obtain a training data set;
step 2.7: and training ShuffleNetv2-YOLOv3 through the generated data set to obtain a detection model aiming at the target key voxel block.
And step 3: inputting the video stream into a trained detection model of ShuffleNetv2-YOLOv3 target key voxel blocks, identifying the key voxel blocks to obtain 2D-3D matching point pairs, and calculating the relative pose of the target by combining an EPNP algorithm;
the specific implementation of the embodiment includes the following sub-steps:
step 3.1: reading a video stream, inputting a trained detection model of ShuffleNetv2-YOLOv3 key voxel blocks, and outputting the detection model into two-dimensional areas corresponding to a plurality of key voxel blocks;
step 3.2: calculating the detected central point of the two-dimensional region, and forming a 2D-3D matching point pair with the central point of the corresponding key voxel block;
step 3.3: and calculating the relative pose between the target and the mobile terminal camera through RANSAC and EPNP algorithms.
And 4, step 4: calculating the pose of the mobile end camera in the three-dimensional world through the GPS and IMU information of the mobile end, and calculating the absolute pose of the target in the three-dimensional world by combining the relative position between the target and the mobile end camera;
the specific implementation of the embodiment includes the following sub-steps:
step 4.1: reading GPS and IMU data of a mobile terminal;
step 4.2: calculating the positioning of the mobile terminal in the three-dimensional world through the data acquired in the step 4.1;
step 4.3: and (3) calculating the absolute pose of the target in the three-dimensional world according to the moving end pose calculated in the step 4.2 and the relative pose of the target calculated in the step 3.
The method comprises the steps of estimating the real-time pose at a mobile terminal, reading a video stream from a camera at the mobile terminal, obtaining feature points of a target in an RGB image by using a ShuffleNet model, and obtaining the relative pose through RANSAC and EPNP algorithms. The invention fully utilizes the advantages of the mobile terminal, and calculates the pose of the target in the actual three-dimensional world by combining the relative poses through the positioning information provided by the GPS and IMU of the mobile terminal.
It should be understood that parts of the specification not set forth in detail are well within the prior art.
It should be understood that the above description of the preferred embodiments is given for clarity and not for any purpose of limitation, and that various changes, substitutions and alterations can be made herein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (5)
1. A mobile terminal three-dimensional pose estimation method based on point cloud preprocessing and ShuffleNet comprises the following steps:
step 1: performing three-dimensional reconstruction on target point cloud data obtained by laser scanning; importing a three-dimensional model obtained by three-dimensional reconstruction into a rendering engine for photographing;
step 2: respectively acquiring two-dimensional photos and camera poses of the target under different visual angles by adopting a positioning rotation photographing algorithm; extracting two-dimensional photo feature points through SIFT, calculating corresponding three-dimensional feature points, dividing a target model into equal-size voxel blocks, and screening target key voxel blocks according to the number of the three-dimensional feature points; generating two-dimensional projection of key voxel blocks on the photo set and establishing a training data set; training a ShuffLeNet feature detection model aiming at a target through a ShufflLeNet v2-YOLOv3 lightweight network;
and step 3: inputting the video stream into a trained detection model of ShuffleNetv2-YOLOv3 target key voxel blocks, identifying the key voxel blocks to obtain 2D-3D matching point pairs, and calculating the relative pose of the target through RANSAC and EPNP algorithms;
and 4, step 4: and calculating the absolute pose of the target in the three-dimensional world by combining the GPS and IMU information of the mobile terminal.
2. The point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method according to claim 1, characterized in that: in the step 1, performing three-dimensional reconstruction on target point cloud data obtained by laser scanning through a Delauay algorithm to obtain a target three-dimensional model; and importing the target three-dimensional model into a rendering engine, calculating a target three-dimensional model bounding box and a central point, and moving the target three-dimensional model to enable the central point to reach an origin point.
3. The method for estimating the three-dimensional pose of the moving end based on the point cloud preprocessing and the ShuffleNet according to claim 1, wherein the concrete implementation of the step 2 comprises the following substeps:
step 2.1: shooting a target in a three-dimensional engine to obtain a target photo;
step 2.2: detecting two-dimensional feature points in the target picture through an SIFT algorithm to obtain a feature point set K ═ K1,...,kn}；
Step 2.3: calculating each characteristic point k by a screen ray projection algorithmiCorresponding three-dimensional coordinate point piAnd recording the three-dimensional feature point set corresponding to the two-dimensional feature point as P ═ P1,...,pn}；
Step 2.4: the camera rotates around the target and continues to take pictures of the target, and the step 2.2 and the step 2.3 are repeated until a multi-view picture of the target is obtained and a three-dimensional feature point set PS is obtained through calculation1,...,PNN is the number of photos, and P is a three-dimensional feature point set of each photo;
step 2.5: dividing the target three-dimensional voxel into M blocks B ═ B with the same voxel size1,...,bM}; setting the frequency of the three-dimensional feature points appearing in each voxel block as a voxel block weight q, and screening m blocks KB with the maximum weight as { b }1,...,bmAs key voxel block, where m<M；
Step 2.6: taking the key voxel blocks as categories, calculating the areas of the key voxel blocks on the two-dimensional photo set according to a projection transformation formula, and generating marking information to obtain a training data set;
step 2.7: and training ShuffleNetv2-YOLOv3 through the generated data set to obtain a detection model aiming at the target key voxel block.
4. The method for estimating the three-dimensional pose of the moving end based on the point cloud preprocessing and the ShuffleNet according to claim 1, wherein the concrete implementation of the step 3 comprises the following substeps:
step 3.1: reading a video stream, inputting a trained detection model of ShuffleNetv2-YOLOv3 key voxel blocks, and outputting the detection model into two-dimensional areas corresponding to a plurality of key voxel blocks;
step 3.2: calculating the detected central point of the two-dimensional region, and forming a 2D-3D matching point pair with the central point of the corresponding key voxel block;
step 3.3: and calculating the relative pose between the target and the mobile terminal camera through RANSAC and EPNP algorithms.
5. The method for estimating the three-dimensional pose of the moving end based on the point cloud preprocessing and the ShuffleNet according to any one of claims 1 to 4, wherein the specific implementation of the step 4 comprises the following substeps:
step 4.1: reading GPS and IMU data of a mobile terminal;
step 4.2: calculating the positioning of the mobile terminal in the three-dimensional world through the data acquired in the step 4.1;
step 4.3: and (3) calculating the absolute pose of the target in the three-dimensional world according to the moving end pose calculated in the step 4.2 and the relative pose of the target calculated in the step 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110634620.0A CN113362467B (en) | 2021-06-08 | 2021-06-08 | Point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110634620.0A CN113362467B (en) | 2021-06-08 | 2021-06-08 | Point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113362467A true CN113362467A (en) | 2021-09-07 |
| CN113362467B CN113362467B (en) | 2023-04-07 |
Family
ID=77532983
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110634620.0A Active CN113362467B (en) | 2021-06-08 | 2021-06-08 | Point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113362467B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115222731A (en) * | 2022-09-07 | 2022-10-21 | 西南交通大学 | Train fastener abnormity detection method based on two-dimensional image-point cloud mapping |
| CN115578265A (en) * | 2022-12-06 | 2023-01-06 | 中汽智联技术有限公司 | Point cloud enhancement method, system and storage medium |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090232353A1 (en) * | 2006-11-10 | 2009-09-17 | University Of Maryland | Method and system for markerless motion capture using multiple cameras |
| US20150269438A1 (en) * | 2014-03-18 | 2015-09-24 | Sri International | Real-time system for multi-modal 3d geospatial mapping, object recognition, scene annotation and analytics |
| EP3276575A1 (en) * | 2016-07-25 | 2018-01-31 | Nuctech Company Limited | Method, apparatus and system for reconstructing images of 3d surface |
| WO2018047687A1 (en) * | 2016-09-12 | 2018-03-15 | パナソニックＩｐマネジメント株式会社 | Three-dimensional model generating device and three-dimensional model generating method |
| WO2018148924A1 (en) * | 2017-02-17 | 2018-08-23 | 深圳市大疆创新科技有限公司 | Method and device for reconstructing three-dimensional point cloud |
| CN109241990A (en) * | 2018-07-19 | 2019-01-18 | 杭州电子科技大学 | A kind of threedimensional model mask method propagated based on multi-tag |
| CN109387204A (en) * | 2018-09-26 | 2019-02-26 | 东北大学 | The synchronous positioning of the mobile robot of dynamic environment and patterning process in faced chamber |
| CN109961504A (en) * | 2017-12-25 | 2019-07-02 | 宏达国际电子股份有限公司 | Method for reconstructing three-dimensional model, electronic device and non-transient computer-readable recording medium |
| US20190213402A1 (en) * | 2018-01-09 | 2019-07-11 | Futurewei Technologies, Inc. | Head pose and distraction estimation |
| CN110517209A (en) * | 2018-05-21 | 2019-11-29 | 北京京东尚科信息技术有限公司 | Data processing method, device, system and computer readable storage medium |
| CN111814683A (en) * | 2020-07-09 | 2020-10-23 | 北京航空航天大学 | Robust visual SLAM method based on semantic prior and deep learning features |
| CN111899301A (en) * | 2020-06-02 | 2020-11-06 | 广州中国科学院先进技术研究所 | Workpiece 6D pose estimation method based on deep learning |
| US20210073641A1 (en) * | 2019-09-05 | 2021-03-11 | Kabushiki Kaisha Toshiba | Learning device, learning system, and learning method |
| CN112489129A (en) * | 2020-12-18 | 2021-03-12 | 深圳市优必选科技股份有限公司 | Pose recognition model training method and device, pose recognition method and terminal equipment |
| US20210158023A1 (en) * | 2018-05-04 | 2021-05-27 | Northeastern University | System and Method for Generating Image Landmarks |
| US20210407039A1 (en) * | 2020-06-30 | 2021-12-30 | Intel Corporation | Apparatus and method for approximate trilinear interpolation for scene reconstruction |
-
2021
- 2021-06-08 CN CN202110634620.0A patent/CN113362467B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090232353A1 (en) * | 2006-11-10 | 2009-09-17 | University Of Maryland | Method and system for markerless motion capture using multiple cameras |
| US20150269438A1 (en) * | 2014-03-18 | 2015-09-24 | Sri International | Real-time system for multi-modal 3d geospatial mapping, object recognition, scene annotation and analytics |
| EP3276575A1 (en) * | 2016-07-25 | 2018-01-31 | Nuctech Company Limited | Method, apparatus and system for reconstructing images of 3d surface |
| WO2018047687A1 (en) * | 2016-09-12 | 2018-03-15 | パナソニックＩｐマネジメント株式会社 | Three-dimensional model generating device and three-dimensional model generating method |
| WO2018148924A1 (en) * | 2017-02-17 | 2018-08-23 | 深圳市大疆创新科技有限公司 | Method and device for reconstructing three-dimensional point cloud |
| CN109961504A (en) * | 2017-12-25 | 2019-07-02 | 宏达国际电子股份有限公司 | Method for reconstructing three-dimensional model, electronic device and non-transient computer-readable recording medium |
| US20190213402A1 (en) * | 2018-01-09 | 2019-07-11 | Futurewei Technologies, Inc. | Head pose and distraction estimation |
| US20210158023A1 (en) * | 2018-05-04 | 2021-05-27 | Northeastern University | System and Method for Generating Image Landmarks |
| CN110517209A (en) * | 2018-05-21 | 2019-11-29 | 北京京东尚科信息技术有限公司 | Data processing method, device, system and computer readable storage medium |
| CN109241990A (en) * | 2018-07-19 | 2019-01-18 | 杭州电子科技大学 | A kind of threedimensional model mask method propagated based on multi-tag |
| CN109387204A (en) * | 2018-09-26 | 2019-02-26 | 东北大学 | The synchronous positioning of the mobile robot of dynamic environment and patterning process in faced chamber |
| US20210073641A1 (en) * | 2019-09-05 | 2021-03-11 | Kabushiki Kaisha Toshiba | Learning device, learning system, and learning method |
| CN111899301A (en) * | 2020-06-02 | 2020-11-06 | 广州中国科学院先进技术研究所 | Workpiece 6D pose estimation method based on deep learning |
| US20210407039A1 (en) * | 2020-06-30 | 2021-12-30 | Intel Corporation | Apparatus and method for approximate trilinear interpolation for scene reconstruction |
| CN111814683A (en) * | 2020-07-09 | 2020-10-23 | 北京航空航天大学 | Robust visual SLAM method based on semantic prior and deep learning features |
| CN112489129A (en) * | 2020-12-18 | 2021-03-12 | 深圳市优必选科技股份有限公司 | Pose recognition model training method and device, pose recognition method and terminal equipment |
Non-Patent Citations (4)
| Title |
|---|
| MENGDAN FENG 等: "2D3D-MatchNet: Learning to Match Keypoints Across 2D Image and 3D Point Cloud", 《2019 INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA)》 * |
| 付泰等: "一种具有精确位姿的飞机CAD模型投影图像库设计方法", 《计算机科学》 * |
| 詹川 著: "《反垃圾邮件信息过滤技术研究》", 31 May 2016, 成都：电子科技大学出版社出版 * |
| 马小秋等: "基于透视式增强现实的BIM建筑三维重建仿真", 《计算机仿真》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115222731A (en) * | 2022-09-07 | 2022-10-21 | 西南交通大学 | Train fastener abnormity detection method based on two-dimensional image-point cloud mapping |
| CN115222731B (en) * | 2022-09-07 | 2022-12-02 | 西南交通大学 | Train fastener abnormity detection method based on two-dimensional image-point cloud mapping |
| CN115578265A (en) * | 2022-12-06 | 2023-01-06 | 中汽智联技术有限公司 | Point cloud enhancement method, system and storage medium |
| CN115578265B (en) * | 2022-12-06 | 2023-04-07 | 中汽智联技术有限公司 | Point cloud enhancement method, system and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113362467B (en) | 2023-04-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11393173B2 (en) | Mobile augmented reality system | |
| CN107330439B (en) | Method for determining posture of object in image, client and server | |
| EP2915140B1 (en) | Fast initialization for monocular visual slam | |
| US20180189974A1 (en) | Machine learning based model localization system | |
| KR102097016B1 (en) | Apparatus and methdo for analayzing motion | |
| US10043097B2 (en) | Image abstraction system | |
| Wei et al. | Applications of structure from motion: a survey | |
| CN110458865B (en) | In situ formation of planar natural feature targets | |
| US20100194679A1 (en) | Gesture recognition system and method thereof | |
| CN108958469B (en) | Method for adding hyperlinks in virtual world based on augmented reality | |
| CN113362467B (en) | Point cloud preprocessing and ShuffleNet-based mobile terminal three-dimensional pose estimation method | |
| Pan et al. | Rapid scene reconstruction on mobile phones from panoramic images | |
| CN103345736A (en) | Virtual viewpoint rendering method | |
| TW201222288A (en) | Image retrieving system and method and computer program product thereof | |
| KR102464271B1 (en) | Pose acquisition method, apparatus, electronic device, storage medium and program | |
| CN115008454A (en) | Robot online hand-eye calibration method based on multi-frame pseudo label data enhancement | |
| Zhu et al. | Large-scale architectural asset extraction from panoramic imagery | |
| CN111914790A (en) | Real-time human body rotation angle identification method based on double cameras under different scenes | |
| CN109118576A (en) | Large scene three-dimensional reconstruction system and method for reconstructing based on BDS location-based service | |
| CN104680520A (en) | Field three-dimensional information investigation method and system | |
| Kikuchi et al. | Automatic diminished reality-based virtual demolition method using semantic segmentation and generative adversarial network for landscape assessment | |
| Kim et al. | Vision-based all-in-one solution for augmented reality and its storytelling applications | |
| CN112258435A (en) | Image processing method and related product | |
| CN105631938A (en) | Image processing method and electronic equipment | |
| WO2022253043A1 (en) | Facial deformation compensation method for facial depth image, and imaging apparatus and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |