CN106548519A - Augmented reality method based on ORB SLAM and the sense of reality of depth camera - Google Patents
Augmented reality method based on ORB SLAM and the sense of reality of depth camera Download PDFInfo
- Publication number
- CN106548519A CN106548519A CN201610963053.2A CN201610963053A CN106548519A CN 106548519 A CN106548519 A CN 106548519A CN 201610963053 A CN201610963053 A CN 201610963053A CN 106548519 A CN106548519 A CN 106548519A
- Authority
- CN
- China
- Prior art keywords
- camera
- depth
- slam
- real
- map
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/006—Mixed reality
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
Abstract
A kind of augmented reality method based on ORB SLAM and the sense of reality of depth camera, comprises the following steps:1), initial coordinate plane is determined using Marker less technologies, and initialize the slam coordinate systems with true yardstick;2) map, is imported, according to the characteristic point extracted, the direct map for representing and characteristics map is combined;3), monocular SLAM builds true yardstick map, and preserves map;4) dummy object is placed in the scene, and the distance between dummy object and camera are calculated by d engine;5) depth camera, is utilized, scene depth data are obtained;Dummy object range data according to obtaining in d engine carries out depth integration, so as to reach the effect blocked, allows real object and dummy model directly to have very strong interaction sense.The present invention has very strong robustness, realizes that real-world object and virtual data can block interaction.
Description
Technical field
Arrive the invention belongs to SLAM be perceived environmental information technology and obtains scene three-dimensional information technological incorporation with depth camera
AR (Augmented Reality, augmented reality) technical field, is related to a kind of augmented reality method of sense of reality.
Background technology
SLAM (simultaneous localization and mapping are positioned and map structuring immediately) is referred to
Robot creates map under the conditions of self-position is uncertain, in complete graphics communication, while being carried out independently using map
Positioning and navigation.ORB-SLAM(a real-time accurate monocular SLAM system based on ORB
Features, the SLAM systems based on ORB characteristic points), ORB refers to a kind of rotational invariance feature, and whole algorithm is base
Realize in ORB features, the progress of newest ORB-SLAM is to have done half dense scene rebuilding based on the key frame of ORB-SLAM.
Augmented reality be by computer system provide Information application to real world, and by computer generation dummy object, field
Scape or system prompt information superposition in real scene, so as to realize to reality enhancing.AR technologies can be applicable to medical treatment, army
The numerous areas such as thing, industrial maintenance, Entertainment.
The mainstream product for being currently based on SLAM technological development has the Hololens of the Microsoft and Project of Google
Tango.The HoloLens of Microsoft has used AR technologies, and one layer of virtual image is superimposed in real world.It is worldwide in the recent period
Pokemon Go play, and present broad mass market of the AR technologies in game application, have also annotated SLAM from another angle
Importance of the technology in future games.As the popularization of AR concepts and the demand of correlation are gradually apparent, SLAM algorithms are
It is considered as that one of optimal choice of real world is understood in AR technologies, increasing AR is using by the performance of SLAM class algorithms
As the important references index evaluated.
But deficiency and excess still suffers from certain defect with reference to aspect in current AR systems, being mainly manifested in dummy object cannot be very
Good is fused in real scene.As AR systems are all often to be carried out by two-dimensional video by the real scene that camera is seen
Represent, three-dimensional virtual object is placed on before video forever.Even if user want to allow three-dimensional virtual object exist with two-dimensional video
Between two real-world objects, but when there are other objects between this real-world object and dummy object, user still can feel
Outside real scene is swum in dummy object.Cause being primarily due in actual object positioned at virtual mould for this phenomenon
When before type, it is impossible to which the function of dummy model is blocked in realization.
The content of the invention
In order to overcome existing AR technologies accurately calculate object dimensional information, cause to block object and correctly can not be managed
Solution and construct, the poor deficiency of robustness, the present invention provides a kind of robustness preferably based on ORB-SLAM and depth camera
The augmented reality method of sense of reality, by the three-dimensional environment of SLAM perceptual objects, depth camera obtains scene distance information, will be true
Object and dummy model depth integration, reach " parallax " effect as human eye so that gesture recognition, visual angle tracking, scene
Rebuild more accurate, so as to realize that real-world object can block the function of dummy model.
The technical solution adopted for the present invention to solve the technical problems is:
A kind of augmented reality method based on ORB-SLAM and the sense of reality of depth camera, methods described include following step
Suddenly:
1), determine that initial coordinate is put down using a kind of Marker-less (augmented reality based on a small amount of labelling) technology
Face, and initialize the slam coordinate systems with true yardstick;Marker-less is utilized by first extracting orb characteristic points
Knnmatch (K arest neighbors matching algorithms) is matched, and is tried to achieve homography matrix, further according to perspective transform, is tried to achieve four summits
Coordinate, then try to achieve rotation translation matrix;
2) map, is imported, according to the characteristic point extracted, the direct map for representing and characteristics map is combined.Carry out
SLAM location navigations, pass to the position of virtual camera and attitude in d engine real camera position and attitude;
3), monocular SLAM builds true yardstick map, and preserves map;
4) dummy object is placed in the scene, and the distance between dummy object and camera are calculated by d engine;
5) depth camera, is utilized, scene depth data are obtained;According to the dummy object range data obtained in d engine
Depth integration is carried out, so as to reach the effect blocked, allows real object and dummy model that directly there is very strong interaction sense.
Further, the step 1) in, Marker-less is a tetragon, and polygonal approximation result meets following bar
Part:
1.1) there was only four summits;
1.2) must be convex polygon;
1.3) length on each side can not be too small.
1.4) because the yardstick of marker pictures can be learnt in advance, then slam is built using his full-size(d)
Coordinate system there is real yardstick;
By conditions above, most of profile is excluded, find the most possible part for Marker-less, find so
Candidate contours after, four summits of its polygon are preserved, and are adjusted, then from these candidate regions further
Real Marker-less is filtered out, perspective transform will be carried out to candidate regions, be obtained the front view of Marker-less;
After obtaining the coordinate on four summits, rotation translation matrix is tried to achieve according to formula x=K [R | T] X, wherein:X is space
The coordinate of certain point, relative to world coordinate system, [R | T] is to join matrix outside video camera to the coordinate, for the world coordinates for putting certain
Camera coordinates are transformed to, K is video camera internal reference, for by the image plane of certain spot projection in camera coordinates, x as throws
The pixel coordinate of movie queen.
Further, the step 3) in, environment is perceived by photographic head, the information to obtaining is analyzed, and extracts ring
Feature in border is simultaneously preserved, and is set up environmental map, i.e., by the method for probability statistics, is matched to reach positioning by multiple features.
Further, the step 4) in, calculated between three-dimensional virtual object and camera by Unity3D d engines
Real-time deep information.
The step 5) in, data fusion process is as follows:
5.1) the real depth data of present frame are obtained using depth camera;
5.2) because slam is calculated the coordinate system of true yardstick, after being applied to d engine, d engine it is virtual
Coordinate system can be with the reasonable coincidence of the coordinate system of real world, and obtained from, the depth data of virtual camera also has true
Real yardstick;
5.3) real depth data and virtual depth data are carried out into a depth ratio compared with depth value is before model
Point corresponding to the point of cromogram be just rendered into before model, and cromogram corresponding to point of the depth value behind the model
Point is just rendered into behind model.
Beneficial effects of the present invention are mainly manifested in:With very strong robustness, strenuous exercise's figure can be processed well
As, have than larger leeway unrestrained section closed loop control, reorientation, even full-automatic position initialization;With reference to depth camera, will
The virtual depth data obtained in the truthful data and d engine of camera acquisition carry out depth integration, realize real-world object and void
Intend data and can block interaction so that AR scenes have more real experience and utilization.
Specific embodiment
The invention will be further described below.
A kind of augmented reality method based on ORB-SLAM and the sense of reality of depth camera, comprises the following steps:
1) initial coordinate plane is determined using Marker-less technologies, extract ORB characteristic points, try to achieve rotation translation matrix,
And initialize the slam coordinate systems with true yardstick.
As Marker-less is a tetragon, its polygonal approximation result should meet following condition:
1.1) there was only four summits;
1.2) must be convex polygon;
1.3) length on each side can not be too small.
1.4) because the yardstick of marker pictures can be learnt in advance, then just can be caused using his full-size(d)
The coordinate system that slam builds has real yardstick.
By more than, several conditions, can exclude most of profile, so as to find the most possible portion for Marker-less
Point, after finding such candidate contours, four summits of its polygon are preserved, and does appropriate adjustment, then from this
Real Marker-less is filtered out further in a little candidate regions.The information in Marker-less is extracted for convenience,
Perspective transform is carried out to candidate regions, the front view of Marker-less is obtained.
Know the coordinate on four summits, rotation translation matrix is tried to achieve according to formula x=K [R | T] X.Wherein:X is space
The coordinate (relative to world coordinate system) of point, [R | T] it is outside video camera, to join matrix, the world coordinate transformation for certain is put is to take the photograph
Camera coordinate, K are video camera internal references, for the picture by the image plane of certain spot projection in camera coordinates, after x as projections
Plain coordinate;
2) SLAM location navigations.
According to the characteristic point extracted, the direct map for representing is combined with the characteristics map for building, in constructing environment
The track of camera is estimated simultaneously, and the position of virtual camera and attitude in d engine are passed in real camera position and attitude;
3) monocular SLAM builds true yardstick map.
Environment is perceived by vision (photographic head), the information to obtaining is analyzed, and the feature in extraction environment is simultaneously preserved,
Set up environmental map.The ultimate principle of SLAM is the method by probability statistics, matches to reach positioning and subtract by multiple features
Few position error.Vision SLAM method passes through observation model and completes accurate interframe movement parameter estimation, reduces algorithm high
The restriction of complexity is required;
4) according to design requirement, three-dimensional virtual object is placed on into correct position, is commonly held within the 3rd step and is determined
Plan-position.Real-time deep information between three-dimensional virtual object and camera is calculated by Unity3D d engines.Can lead to
Cross function " UNITY_SAMPLE_DEPTH " and obtain depth information, but this depth information is non-linear depth information, it is impossible to be straight
Connect and merge with real scene depth information, in addition it is also necessary to dummy object linear depth is obtained by processing.Process false code is as follows:
4.1) pass through _ DepthPower sets the parameter that non-linear depth is transformed into linear depth;
4.2) non-linear depth d of dummy object is calculated using function UNITY_SAMPLE_DEPTH;
4.3) using function pow and 4.1) in parameter d is converted to into linear depth;
5) depth camera is utilized, depth data is mapped in true environment, reach dummy object mutual with real-world object
The effect blocked.
Depth camera can be rapidly completed the identification to target and follow the trail of, and can be obtained between object more by range information
Abundant position relationship, that is, distinguish prospect and background, obtains depth data in scene in real time.Then obtain in d engine
The scene depth data that dummy object depth data is obtained with camera are merged, and realize phase between dummy object and real-world object
The effect mutually blocked, so that AR scenes have higher sense of reality.
Data fusion step:
5.1) the real depth data of present frame are obtained using depth camera;
5.2) because slam is calculated the coordinate system of true yardstick, after being applied to d engine, d engine it is virtual
Coordinate system can be with the reasonable coincidence of the coordinate system of real world, and obtained from, the depth data of virtual camera also has true
Real yardstick;
5.3) real depth data and virtual depth data are carried out into a depth ratio compared with depth value is before model
Point corresponding to the point of cromogram be just rendered into before model, and cromogram corresponding to point of the depth value behind the model
Point is just rendered into behind model.
Claims (5)
1. a kind of augmented reality method based on ORB SLAM and the sense of reality of depth camera, it is characterised in that:Methods described bag
Include following steps:
1), initial coordinate plane is determined using Marker-less technologies, and initialize the slam coordinate systems with true yardstick;
Marker-less is matched using knnmatch, is tried to achieve homography matrix by first extracting orb characteristic points, is become further according to perspective
Change, try to achieve the coordinate on four summits, then try to achieve rotation translation matrix;
2) map, is imported, according to the characteristic point extracted, the direct map for representing and characteristics map is combined, SLAM is carried out fixed
Position navigation, passes to the position of virtual camera and attitude in d engine real camera position and attitude;
3), monocular SLAM builds true yardstick map, and preserves map;
4) dummy object is placed in the scene, and the distance between dummy object and camera are calculated by d engine;
5) depth camera, is utilized, scene depth data are obtained;Dummy object range data according to obtaining in d engine is carried out
Depth integration, so as to reach the effect blocked, allows real object and dummy model directly to have very strong interaction sense.
2. the augmented reality method based on ORB SLAM and the sense of reality of depth camera as claimed in claim 1, its feature exist
In:The step 1) in, Marker-less is a tetragon, and polygonal approximation result meets following condition:
1.1) there was only four summits;
1.2) must be convex polygon;
1.3) length on each side can not be too small;
1.4) because the yardstick of marker pictures can be learnt in advance, then cause the seat of slam structures using his full-size(d)
Mark system is with real yardstick;
By conditions above, most of profile is excluded, find the most possible part for Marker-less, find such time
After selecting profile, four summits of its polygon are preserved, and is adjusted, then further screened from these candidate regions
Go out real Marker-less, perspective transform will be carried out to candidate regions, obtain the front view of Marker-less;
After obtaining the coordinate on four summits, rotation translation matrix is tried to achieve according to formula x=K [R | T] X, wherein:X is space point
Coordinate, relative to world coordinate system, [R | T] is to join matrix outside video camera to the coordinate, for the world coordinate transformation for putting certain
For camera coordinates, K is video camera internal reference, for by the image plane of certain spot projection in camera coordinates, x is after projecting
Pixel coordinate.
3. the augmented reality method based on ORB SLAM and the sense of reality of depth camera as claimed in claim 1 or 2, its feature
It is:The step 3) in, environment is perceived by photographic head, the information to obtaining is analyzed, and the feature in extraction environment is simultaneously
Preserve, set up environmental map, i.e., by the method for probability statistics, match to reach positioning by multiple features.
4. the augmented reality method based on ORB SLAM and the sense of reality of depth camera as claimed in claim 1 or 2, its feature
It is:The step 4) in, the real-time deep letter between three-dimensional virtual object and camera is calculated by Unity3D d engines
Breath.
5. the augmented reality method based on ORB SLAM and the sense of reality of depth camera as claimed in claim 1 or 2, its feature
It is:The step 5) in, data fusion process is as follows:
5.1) the real depth data of present frame are obtained using depth camera;
5.2) because slam is calculated the coordinate system of true yardstick, after being applied to d engine, the virtual coordinates of d engine
System can be with the reasonable coincidence of the coordinate system of real world, and obtained from, the depth data of virtual camera also has real
Yardstick;
5.3) real depth data and virtual depth data are carried out into a depth ratio compared with point of the depth value before model
The point of corresponding cromogram is just rendered into before model, and the point of the cromogram corresponding to point of the depth value behind the model is just
It is rendered into behind model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610963053.2A CN106548519A (en) | 2016-11-04 | 2016-11-04 | Augmented reality method based on ORB SLAM and the sense of reality of depth camera |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610963053.2A CN106548519A (en) | 2016-11-04 | 2016-11-04 | Augmented reality method based on ORB SLAM and the sense of reality of depth camera |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106548519A true CN106548519A (en) | 2017-03-29 |
Family
ID=58394247
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610963053.2A Pending CN106548519A (en) | 2016-11-04 | 2016-11-04 | Augmented reality method based on ORB SLAM and the sense of reality of depth camera |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106548519A (en) |
Cited By (28)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107185245A (en) * | 2017-05-31 | 2017-09-22 | 武汉秀宝软件有限公司 | A kind of actual situation synchronous display method and system based on SLAM technologies |
CN107204015A (en) * | 2017-05-27 | 2017-09-26 | 中山大学 | Instant positioning based on color image and infrared image fusion is with building drawing system |
CN107564107A (en) * | 2017-07-19 | 2018-01-09 | 中国农业大学 | A kind of design method and equipment of augmented reality implementation tool |
CN107564089A (en) * | 2017-08-10 | 2018-01-09 | 腾讯科技(深圳)有限公司 | Three dimensional image processing method, device, storage medium and computer equipment |
CN107766855A (en) * | 2017-10-25 | 2018-03-06 | 南京阿凡达机器人科技有限公司 | Chess piece localization method, system, storage medium and robot based on machine vision |
CN107958466A (en) * | 2017-12-01 | 2018-04-24 | 大唐国信滨海海上风力发电有限公司 | A kind of tracking of the Slam algorithm optimizations based on model |
CN108109208A (en) * | 2017-12-01 | 2018-06-01 | 同济大学 | A kind of marine wind electric field augmented reality method |
CN108364344A (en) * | 2018-02-08 | 2018-08-03 | 重庆邮电大学 | A kind of monocular real-time three-dimensional method for reconstructing based on loopback test |
CN108537844A (en) * | 2018-03-16 | 2018-09-14 | 上海交通大学 | A kind of vision SLAM winding detection methods of fusion geological information |
CN108537889A (en) * | 2018-03-26 | 2018-09-14 | 广东欧珀移动通信有限公司 | Method of adjustment, device, storage medium and the electronic equipment of augmented reality model |
CN108596964A (en) * | 2018-05-02 | 2018-09-28 | 厦门美图之家科技有限公司 | Depth data acquisition methods, device and readable storage medium storing program for executing |
CN108735052A (en) * | 2018-05-09 | 2018-11-02 | 北京航空航天大学青岛研究院 | A kind of augmented reality experiment with falling objects method based on SLAM |
CN108805985A (en) * | 2018-03-23 | 2018-11-13 | 福建数博讯信息科技有限公司 | Virtual Space method and apparatus |
CN108876900A (en) * | 2018-05-11 | 2018-11-23 | 重庆爱奇艺智能科技有限公司 | A kind of virtual target projective techniques merged with reality scene and system |
CN108905208A (en) * | 2018-06-21 | 2018-11-30 | 珠海金山网络游戏科技有限公司 | A kind of electronic gaming method and device based on augmented reality |
CN109471521A (en) * | 2018-09-05 | 2019-03-15 | 华东计算技术研究所(中国电子科技集团公司第三十二研究所) | Virtual and real shielding interaction method and system in AR environment |
CN109636916A (en) * | 2018-07-17 | 2019-04-16 | 北京理工大学 | A kind of a wide range of virtual reality roaming system and method for dynamic calibration |
CN109636852A (en) * | 2018-11-23 | 2019-04-16 | 浙江工业大学 | A kind of monocular SLAM initial method |
CN110009732A (en) * | 2019-04-11 | 2019-07-12 | 司岚光电科技(苏州)有限公司 | Based on GMS characteristic matching towards complicated large scale scene three-dimensional reconstruction method |
CN110120098A (en) * | 2018-02-05 | 2019-08-13 | 浙江商汤科技开发有限公司 | Scene size estimation and augmented reality control method, device and electronic equipment |
CN110599432A (en) * | 2018-06-12 | 2019-12-20 | 光宝电子(广州)有限公司 | Image processing system and image processing method |
CN110827376A (en) * | 2018-08-09 | 2020-02-21 | 北京微播视界科技有限公司 | Augmented reality multi-plane model animation interaction method, device, equipment and storage medium |
CN110827411A (en) * | 2018-08-09 | 2020-02-21 | 北京微播视界科技有限公司 | Self-adaptive environment augmented reality model display method, device, equipment and storage medium |
CN110910484A (en) * | 2019-12-03 | 2020-03-24 | 上海世长信息科技有限公司 | SLAM-based object mapping method from two-dimensional image to three-dimensional real scene |
CN111815696A (en) * | 2019-04-11 | 2020-10-23 | 曜科智能科技(上海)有限公司 | Depth map optimization method, device, equipment and medium based on semantic instance segmentation |
US10853649B2 (en) * | 2018-08-27 | 2020-12-01 | Dell Products, L.P. | Context-aware hazard detection using world-facing cameras in virtual, augmented, and mixed reality (xR) applications |
CN112258633A (en) * | 2020-10-23 | 2021-01-22 | 华中科技大学鄂州工业技术研究院 | High-precision scene reconstruction method and device based on SLAM technology |
CN112308980A (en) * | 2020-10-30 | 2021-02-02 | 脸萌有限公司 | Augmented reality interactive display method and equipment |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103489214A (en) * | 2013-09-10 | 2014-01-01 | 北京邮电大学 | Virtual reality occlusion handling method, based on virtual model pretreatment, in augmented reality system |
-
2016
- 2016-11-04 CN CN201610963053.2A patent/CN106548519A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103489214A (en) * | 2013-09-10 | 2014-01-01 | 北京邮电大学 | Virtual reality occlusion handling method, based on virtual model pretreatment, in augmented reality system |
Non-Patent Citations (4)
Title |
---|
HAOMIN LIU ET AL: "Robust Keyframe-based Monocular SLAM for Augmented Reality", 《2016 IEEE INTERNATIONAL SYMPOSIUM ON MIXED AND AUGMENTED REALITY》 * |
刘浩敏 等: "基于单目视觉的同时定位与地图构建方法综述", 《计算机辅助设计与图形学学报》 * |
李红波 等: "动态变换背景帧的虚实遮挡处理方法", 《计算机工程与设计》 * |
饶玲珊 等: "增强现实游戏的场景重建和运动物体跟踪技术", 《计算机工程与应用》 * |
Cited By (42)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107204015A (en) * | 2017-05-27 | 2017-09-26 | 中山大学 | Instant positioning based on color image and infrared image fusion is with building drawing system |
CN107204015B (en) * | 2017-05-27 | 2021-06-08 | 中山大学 | Instant positioning and mapping system based on color image and infrared image fusion |
CN107185245A (en) * | 2017-05-31 | 2017-09-22 | 武汉秀宝软件有限公司 | A kind of actual situation synchronous display method and system based on SLAM technologies |
CN107564107A (en) * | 2017-07-19 | 2018-01-09 | 中国农业大学 | A kind of design method and equipment of augmented reality implementation tool |
CN107564089B (en) * | 2017-08-10 | 2022-03-01 | 腾讯科技(深圳)有限公司 | Three-dimensional image processing method, device, storage medium and computer equipment |
CN107564089A (en) * | 2017-08-10 | 2018-01-09 | 腾讯科技(深圳)有限公司 | Three dimensional image processing method, device, storage medium and computer equipment |
CN107766855A (en) * | 2017-10-25 | 2018-03-06 | 南京阿凡达机器人科技有限公司 | Chess piece localization method, system, storage medium and robot based on machine vision |
CN107766855B (en) * | 2017-10-25 | 2021-09-07 | 南京阿凡达机器人科技有限公司 | Chessman positioning method and system based on machine vision, storage medium and robot |
CN108109208A (en) * | 2017-12-01 | 2018-06-01 | 同济大学 | A kind of marine wind electric field augmented reality method |
CN107958466B (en) * | 2017-12-01 | 2022-03-29 | 大唐国信滨海海上风力发电有限公司 | Slam algorithm optimization model-based tracking method |
CN107958466A (en) * | 2017-12-01 | 2018-04-24 | 大唐国信滨海海上风力发电有限公司 | A kind of tracking of the Slam algorithm optimizations based on model |
CN108109208B (en) * | 2017-12-01 | 2022-02-08 | 同济大学 | Augmented reality method for offshore wind farm |
CN110120098B (en) * | 2018-02-05 | 2023-10-13 | 浙江商汤科技开发有限公司 | Scene scale estimation and augmented reality control method and device and electronic equipment |
CN110120098A (en) * | 2018-02-05 | 2019-08-13 | 浙江商汤科技开发有限公司 | Scene size estimation and augmented reality control method, device and electronic equipment |
CN108364344A (en) * | 2018-02-08 | 2018-08-03 | 重庆邮电大学 | A kind of monocular real-time three-dimensional method for reconstructing based on loopback test |
CN108537844A (en) * | 2018-03-16 | 2018-09-14 | 上海交通大学 | A kind of vision SLAM winding detection methods of fusion geological information |
CN108537844B (en) * | 2018-03-16 | 2021-11-26 | 上海交通大学 | Visual SLAM loop detection method fusing geometric information |
CN108805985B (en) * | 2018-03-23 | 2022-02-15 | 福建数博讯信息科技有限公司 | Virtual space method and device |
CN108805985A (en) * | 2018-03-23 | 2018-11-13 | 福建数博讯信息科技有限公司 | Virtual Space method and apparatus |
WO2019184889A1 (en) * | 2018-03-26 | 2019-10-03 | Oppo广东移动通信有限公司 | Method and apparatus for adjusting augmented reality model, storage medium, and electronic device |
CN108537889A (en) * | 2018-03-26 | 2018-09-14 | 广东欧珀移动通信有限公司 | Method of adjustment, device, storage medium and the electronic equipment of augmented reality model |
CN108596964A (en) * | 2018-05-02 | 2018-09-28 | 厦门美图之家科技有限公司 | Depth data acquisition methods, device and readable storage medium storing program for executing |
CN108735052A (en) * | 2018-05-09 | 2018-11-02 | 北京航空航天大学青岛研究院 | A kind of augmented reality experiment with falling objects method based on SLAM |
CN108876900A (en) * | 2018-05-11 | 2018-11-23 | 重庆爱奇艺智能科技有限公司 | A kind of virtual target projective techniques merged with reality scene and system |
CN110599432A (en) * | 2018-06-12 | 2019-12-20 | 光宝电子(广州)有限公司 | Image processing system and image processing method |
CN110599432B (en) * | 2018-06-12 | 2023-02-24 | 光宝电子(广州)有限公司 | Image processing system and image processing method |
CN108905208A (en) * | 2018-06-21 | 2018-11-30 | 珠海金山网络游戏科技有限公司 | A kind of electronic gaming method and device based on augmented reality |
CN109636916A (en) * | 2018-07-17 | 2019-04-16 | 北京理工大学 | A kind of a wide range of virtual reality roaming system and method for dynamic calibration |
CN109636916B (en) * | 2018-07-17 | 2022-12-02 | 北京理工大学 | Dynamic calibration large-range virtual reality roaming system and method |
CN110827376A (en) * | 2018-08-09 | 2020-02-21 | 北京微播视界科技有限公司 | Augmented reality multi-plane model animation interaction method, device, equipment and storage medium |
CN110827411A (en) * | 2018-08-09 | 2020-02-21 | 北京微播视界科技有限公司 | Self-adaptive environment augmented reality model display method, device, equipment and storage medium |
US10853649B2 (en) * | 2018-08-27 | 2020-12-01 | Dell Products, L.P. | Context-aware hazard detection using world-facing cameras in virtual, augmented, and mixed reality (xR) applications |
CN109471521A (en) * | 2018-09-05 | 2019-03-15 | 华东计算技术研究所(中国电子科技集团公司第三十二研究所) | Virtual and real shielding interaction method and system in AR environment |
CN109636852A (en) * | 2018-11-23 | 2019-04-16 | 浙江工业大学 | A kind of monocular SLAM initial method |
CN111815696A (en) * | 2019-04-11 | 2020-10-23 | 曜科智能科技(上海)有限公司 | Depth map optimization method, device, equipment and medium based on semantic instance segmentation |
CN110009732A (en) * | 2019-04-11 | 2019-07-12 | 司岚光电科技(苏州)有限公司 | Based on GMS characteristic matching towards complicated large scale scene three-dimensional reconstruction method |
CN111815696B (en) * | 2019-04-11 | 2023-08-22 | 曜科智能科技(上海)有限公司 | Depth map optimization method, device, equipment and medium based on semantic instance segmentation |
CN110009732B (en) * | 2019-04-11 | 2023-10-03 | 司岚光电科技(苏州)有限公司 | GMS feature matching-based three-dimensional reconstruction method for complex large-scale scene |
CN110910484A (en) * | 2019-12-03 | 2020-03-24 | 上海世长信息科技有限公司 | SLAM-based object mapping method from two-dimensional image to three-dimensional real scene |
CN112258633A (en) * | 2020-10-23 | 2021-01-22 | 华中科技大学鄂州工业技术研究院 | High-precision scene reconstruction method and device based on SLAM technology |
CN112258633B (en) * | 2020-10-23 | 2023-02-28 | 华中科技大学鄂州工业技术研究院 | SLAM technology-based scene high-precision reconstruction method and device |
CN112308980A (en) * | 2020-10-30 | 2021-02-02 | 脸萌有限公司 | Augmented reality interactive display method and equipment |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106548519A (en) | Augmented reality method based on ORB SLAM and the sense of reality of depth camera | |
CN109166149B (en) | Positioning and three-dimensional line frame structure reconstruction method and system integrating binocular camera and IMU | |
CN105631861B (en) | Restore the method for 3 D human body posture from unmarked monocular image in conjunction with height map | |
CN106228538B (en) | Binocular vision indoor orientation method based on logo | |
Hagbi et al. | Shape recognition and pose estimation for mobile augmented reality | |
CN109671120A (en) | A kind of monocular SLAM initial method and system based on wheel type encoder | |
CN101681423B (en) | Method of capturing, processing, and rendering images | |
US20170193710A1 (en) | System and method for generating a mixed reality environment | |
CN103177269B (en) | For estimating the apparatus and method of object gesture | |
US8675972B2 (en) | Method and device for determining the pose of a three-dimensional object in an image and method and device for creating at least one key image for object tracking | |
CN107990899A (en) | A kind of localization method and system based on SLAM | |
CN105528082A (en) | Three-dimensional space and hand gesture recognition tracing interactive method, device and system | |
KR20090110357A (en) | Augmented reality method and devices using a real time automatic tracking of marker-free textured planar geometrical objects in a video stream | |
CN103994765B (en) | Positioning method of inertial sensor | |
CN108955718A (en) | A kind of visual odometry and its localization method, robot and storage medium | |
Tulyakov et al. | Robust real-time extreme head pose estimation | |
CN113706699B (en) | Data processing method and device, electronic equipment and computer readable storage medium | |
CN111127524A (en) | Method, system and device for tracking trajectory and reconstructing three-dimensional image | |
WO2021002025A1 (en) | Skeleton recognition method, skeleton recognition program, skeleton recognition system, learning method, learning program, and learning device | |
CN109978919A (en) | A kind of vehicle positioning method and system based on monocular camera | |
Zou et al. | Automatic reconstruction of 3D human motion pose from uncalibrated monocular video sequences based on markerless human motion tracking | |
CN111784775A (en) | Identification-assisted visual inertia augmented reality registration method | |
CN104166995B (en) | Harris-SIFT binocular vision positioning method based on horse pace measurement | |
CN109584347A (en) | A kind of augmented reality mutual occlusion processing method based on active apparent model | |
CN107101632A (en) | Space positioning apparatus and method based on multi-cam and many markers |
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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170329 |