CN106846479B - Three-dimensional visialization of tunnel system and method based on depth camera - Google Patents
Three-dimensional visialization of tunnel system and method based on depth camera Download PDFInfo
- Publication number
- CN106846479B CN106846479B CN201710081447.XA CN201710081447A CN106846479B CN 106846479 B CN106846479 B CN 106846479B CN 201710081447 A CN201710081447 A CN 201710081447A CN 106846479 B CN106846479 B CN 106846479B
- Authority
- CN
- China
- Prior art keywords
- depth
- camera
- tunnel
- depth camera
- rotating device
- 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.)
- Active
Links
Classifications
-
- 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
- G06T17/05—Geographic models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10028—Range image; Depth image; 3D point clouds
Landscapes
- Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Geometry (AREA)
- Software Systems (AREA)
- Remote Sensing (AREA)
- Computer Graphics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Processing Or Creating Images (AREA)
Abstract
The invention discloses a kind of three-dimensional visialization of tunnel system and method based on depth camera, including depth camera device, rotating device, the depth camera device includes microphone assembly, infrared camera, colour imagery shot and infrared camera, the depth camera device is mounted on the rotating device, the rotating device drives depth camera device to carry out 180 ° of rotation on vertical direction, acquires the image of tunnel upper half；It is provided with angular transducer on rotating device, detects the rotation angle of rotating device, while being provided with elevation-angle controller on rotating device, the acquisition angles of controlling depth cam device.The present invention can also be observed that geological state whole in tunnel, such as the trend at joint, crack by the analysis to tunnel correlated color information, to carry out more targeted scientific construction.
Description
Technical field
The present invention relates to a kind of three-dimensional visialization of tunnel system and method based on depth camera.
Background technique
As the country that a science and technology, economic various aspects develop rapidly, will necessarily be built more during city is built
More tunnel, subway meet the needs of in traffic.Certainly, in order to avoid the generation of accident, in order to which tunnel is better anticipated
Situation allows the process of tunnelling to be in progress more smooth, we carry out three-dimensional modeling to tunnel using depth camera, thus
Allowing and is showed on three-dimensional space the case where entire tunnel, let us has more intuitive understanding to tunnel information,
To avoid the generation of some contingencies.
For at present, the application range of depth camera is than wide.Its main exploitation is the reason is that be used for instantly very
Popular somatic sensation television game, being able to use family has better experience.However as being constantly progressive for technology, depth camera is virtual
Real aspect can also have very novel application, and some boutiques apply to this virtual technology on fitting room, Ke Huwu
It need to try on i.e. and true clothes effect can be observed；In engineering, application range mainly includes the weight of small-scale scene (object)
It builds, the 3D scanning and printing of object facility；And the depth camera of low cost is added in robot field, and utilize machine
Device people carries out some measurements not needed under the adverse circumstances such as very high-precision dangerous area or ground end high-altitude and prospecting operation；
In terms of medicine, the application idea of depth camera is also that quite extensively, the research project of Bern university of Switzerland is deep using exploitation
Camera system is spent, replaces doctor to dissect corpse using sound control and body-sensing software, University of Washington then grinds in laboratory
Surgical mechanical arm can be controlled by the transformation to depth camera by studying carefully, and utilize its sensitive negative-feedback function
Operation can be executed to control arm, or obtain the rehabilitation shape of current patient by the body kinematics situation of capture postoperative patient
Condition etc..
In terms of tunnel detection, advance geologic prediction mainly uses engineering geophysical method, i.e., to tunnel physical parameter into
The method of row analysis.Because all there is the application range and detection accuracy of oneself in various geophysical prospecting methods, our usually bases
Difference physical features possessed by detected object are detected using two or more effective geophysical prospecting method, and to result
Carry out comprehensive analysis.From the point of view of the engineering practice effort of many years and experience, in the work physical prospecting of tunnel geological forecast work
Mainly the methods of flexible wave reflection method, echolation, infrared detecting method, DC electrical method, these types of method are all method
Various aspects geological information in front of front and side is tunneled to complete-section tunnel boring machine and carries out detection analysis, for tunneling boring tunnel digging
It is substantially taken into account into the case where machine rear fewer.Three-dimensional visialization of tunnel technology may be implemented to characterize complicated tunnel, break
The processing of face and local section, while the triangle gridding to tunnel shape, surface area meter may be implemented in the design of some systems
It calculates and tunnel contour line deviates calculating, infrared temperature data can be shown, any cross section screenshot and some parts are thin
The displaying of section.Further, by way of software and hardware combining, system can be detected tunnel deformation, and to its original
Because being analyzed.The accuracy of inspection is also relatively high (+/- 2mm within 25m).Tunnel state deterioration multidate information can be grasped,
And system operatio is mostly fairly simple, not examinate person's technical level limitation.
In conclusion need to overcome some known difficulties during tunnel scene imaging, it on the one hand will be to extensive
Scene carry out rebuild be to need to occupy sizable memory, and reconstruction process needs to expend considerable time, another party
Face color information in tunnel adverse circumstances is not that can easily extract very much, due to the deficiency of hardware condition facility,
It needs to spend more energy during the processing of later period color information.And environment is poor in tunnel, some quicker
The equipment of sense is commonly used in tunnel and can inevitably reduce the service life, it is therefore desirable to increase good safeguard measure.
Summary of the invention
The present invention to solve the above-mentioned problems, proposes a kind of three-dimensional visialization of tunnel system based on depth camera and side
Depth camera is carried on complete-section tunnel boring machine and is subject to mechanical rotary device and protective device by method, the present invention,
In the tunneling process of complete-section tunnel boring machine, tunnel cross-section experienced is filmed one by one during being advanced.Pass through
Overall condition in current tunnel is obtained to the processing of tunnel vision information, analysis.
To achieve the goals above, the present invention adopts the following technical scheme that：
A kind of three-dimensional visialization of tunnel system based on depth camera, including depth camera device, rotating device, it is described
Depth camera device includes microphone assembly, infrared camera, colour imagery shot and infrared camera, the depth camera
Device is mounted on the rotating device, and the rotating device drives 180 ° on depth camera device progress vertical direction
Rotation acquires the image of tunnel upper half；
It is provided with angular transducer on the rotating device, detects the rotation angle of rotating device, while on rotating device
It is provided with elevation-angle controller, the acquisition angles of controlling depth cam device.
The depth camera device is set on bracket, and the bracket is fixed on rotating device.
The depth camera device is provided with waterproof case.
Modeling method based on above system, includes the following steps：
(1) depth information and image information of the acquisition of depth camera device are obtained；
(2) acquired image is sampled, floating-point, smoothing techniques, depth camera dress is obtained from image
The motion track set；
(3) large-scale virtual space is established, the information being collected is fused in Virtual Space according to the sampling interval, is formed
Threedimensional model.
In the step (1), the depth camera being mounted on complete-section tunnel boring machine is as the advance two of machine obtains
The depth and colour information in tunnel is taken as constantly to add wherein the part for repeating to take can be used as the optimization of photographic intelligence
Add tunnel information, the lower limit of redundancy can't be set.
In the step (2), depth image data is converted according to threshold value setting, the data except threshold range
Setting distance is invalid.
In the step (2), depth data is subjected to sampling processing, increases the speed of processing, while to the weight in scene
It builds object to be handled, its reconstruction model can be optimized by smoothing techniques to mobile small-sized object, while by smooth
Algorithm is denoised, and the dynamic change in certain scenes has also been handled, any partition being not exhibited by raw video
Or sky can be also filled, as camera is closer to object, by using the data of new higher precision, body surface can quilt
Continuous optimization.
In the step (2), its posture is constantly obtained when camera is mobile by using the registration Algorithm of interactive, this
Sample system knows the relative pose of the camera when preceding camera is relative to start frame always.Two are generally used for posture tracking
Kind registration Algorithm.The first is using the point cloud that will be calculated the point cloud got from reconstructed object with obtain from depth image data
It is registrated, or individually uses and such as the data of the different field angles of Same Scene are registrated；Second algorithm can
To re-establishing the tracking that can obtain higher precision when cube is handled as a result, the still object for moving in scene
The algorithm may be not healthy and strong enough, if the tracking in scene is interrupted, needs to take the photograph the position of camera and last
As head aligned in position can just continue to track.
In the step (2), the size of smoothing parameter is set according to the movement speed of depth camera device.
It is to represent camera view by the depth image data fusion generated from known poses camera in the step (3)
The cube of scenery in wild range.This fusion to depth data is frame by frame, to be carried out continuously, finally, regarding from sensor
Point position counterweight is established cube and carries out light projection, the point cloud being calculated using ray casting algorithm, then calculates its normal vector,
With the input picture registration of point cloud and next frame with normal vector, the pose of next frame input picture is calculated.It is so a circulation
Process.The dot matrix cloud of reconstruction can generate the three-dimensional reconstruction cube rendered.
In the step (3), the precision of model is set by the way that the size of sampling step length is arranged.Pay attention to setting for sampling step length
It sets range to have to be larger than 0 and be less than minimum volume axis voxel resolution, the sampling step length value beyond this range will will lead to weight
Threedimensional model after building lacks the details on curved surface or curved surface.
In the step (3), this system focuses on realizing large-scale scene modeling, special due to tunnel model
Property, therefore attempt to set up that comparison is a wide range of in the side that tunnelling is advanced, because of the limitation of GPU memory, system is using sharp
It is modeled with CPU memory.In modeling process by the way of processed offline.Winding detection and winding optimization are added in algorithm,
According to winding optimization as a result, update point coordinate so that winding place rebuild twice can be aligned.When camera rotation or
When the mobile distance of person is more than certain threshold value, key frame is done into present frame addition and carries out winding detection, winding detection
It is if there is matched image, the SURF of the matching image stored in memory is special first by finding matched key frame
Sign point and depth image index out again and.The SURF point of given two field pictures describes sub- Ui and Um, with FLANN lookup algorithm come
Establish the matching relationship of SURF, if it is possible to which the SURF point quantity matched is no more than given threshold, then it is assumed that this is not one
Effective matching, the matching relationship of SURF is established by matching, and the matching between the SURF established by previous step is calculated with RANSAC
Method estimates the pose between two frames, optimizes camera pose with LM algorithm optimization re-projection error again after obtaining pose.It is excellent again with ICP
Change the pose that above-mentioned algorithm is calculated, if the error between match point is less than given threshold, then it is assumed that this is one effective
Winding.
Beneficial effects of the present invention are：
(1) tunnel threedimensional model can integrally be showed, allow user for entire tunnel have more intuitively cognition with
Impression；
(2) by the analysis for tunnel correlated color information, it can also be observed that geological state whole in tunnel, such as
The trend etc. in crack, to carry out more targeted scientific construction.Overall model is to system of analysis rock organ verifying relevant parameter
The influence generated for complete-section tunnel boring machine has certain effect.
Detailed description of the invention
Fig. 1 is schematic diagram sectional view of the invention.
Fig. 2 is depth camera device figure of the invention.
Fig. 3 is design flow diagram of the present invention.
Wherein, 1 depth camera is indicated, 2 indicate mechanical rotary device, and 3 indicate tunnel surface, and 4 indicate shell protection dress
It sets, 9 indicate microphone array, and 10 indicate infrared camera, and 11 indicate colour imagery shot, and 12 indicate infrared camera, and 13 indicate
The elevation angle controls motor, and 14 indicate bracket.
Specific embodiment：
The invention will be further described with embodiment with reference to the accompanying drawing.
Since scene is larger, and the mode handled in real time is suitable consuming GPU memory, therefore we use offline
The method for handling data, i.e., first collect all information in tunnel, then handle on this basis data, to keep away
The case where exempting from Out of Memory.Wherein, processed offline is also classified into both of which, one is first small-scale model foundation is got up,
Then all small-scale model of place are stitched together one by one, are in addition then first to get all data, it is unified to data
Analysis processing is done, last threedimensional model is calculated.We using the latter method because for split-join model,
It wherein there will certainly be some errors, this is poorer than the precision of the latter much.Depth camera to from multiple angles by obtaining
The depth image data got is merged, to rebuild the single frames smooth surface model of object.When sensor is mobile, shine
The position of camera and pose information are recorded, these information include position and orientation.Due to it is understood that each frame
Being associated between the posture and frame and frame of image, the data that multiframe acquires from different perspectives can be fused into what single frames had been rebuild
Pinpoint cube.We are envisioned that down a huge virtual cube in space, and the inside is our real worlds
Scene, when our movable sensors, depth data information is continuously added.
Firstly, depth camera is carried on complete-section tunnel boring machine machine, and hardware rotation device is utilized, complete
Scanning carries out 180 ° of scanning-tunnelling upper half area panoramas during section tunnel boring machine advances, because of complete-section tunnel boring machine
Forward speed to be much smaller than the scanning speed of depth camera, and the multiple scanning of depth camera is more advantageous to scene
The reconstruction of real information, and the true motion track of depth camera camera also has algorithm later and is calculated, therefore I
Can leave out of consideration depth camera video camera true and motion track it is whether regular.
Further, the relevant depth information of scene, depth herein are got using the infrared camera of depth camera
Degree information can embody the position of the current taken specific camera of object, under in addition obtaining its colour information and storing
Come.
Further, photographic intelligence is handled, depth image data is converted according to threshold value setting, in threshold value
Data setting distance except range be it is invalid, in this way can be by some special objects exclusions except three-dimensional reconstruction.
Further, depth data after processing is registrated, includes the latest position for calculating camera, track position
Acquisition can be calculated according to trajectory calculation algorithm, the calculating of this parameter can make system understand camera relative to start frame
When camera position, and data are carried out with the process of smoothing techniques, the size for smoothing parameter can move according to camera
Dynamic speed determines, thus can ensure that detail of information retains number.
Further, virtual cubic space is established, before requirement namely tunnel of the Virtual Space herein for Y-axis
It is bigger into direction.The difficulty of large scale scene modeling mainly in this respect, the requirement ratio due to Real-time modeling set for memory
It is larger, therefore by the way of processed offline.By by after smoothing denoising depth real-coded GA and camera position believe
Breath handled, certainly also handle scene in some other small dynamic change or wisps movement or disappearance, after from
Sensor viewpoint position project to the light for re-establishing cube, and lighting after reconstruction is required for cloud sequence can render
Three-dimension Reconstruction Model.
Further, the precision of model is set by the way that the size of sampling step length is arranged, the size of sampling step length will consider
To the application of memory and the runing time and the exquisite degree of last model etc. of program.
The present invention is equipped on complete-section tunnel boring machine.
As shown in Figure 1 and Figure 2, the depth camera 3-D imaging system that complete-section tunnel boring machine carries, including can incite somebody to action
The mechanical device of camera is rotated, the overall space in tunnel can be subjected to three-dimensional modeling.Whole system is including being depth camera
Head can rotate the mechanical rotary device that can take tunnel upper half, protect the canning of camera, and can be right
Tunnel scene is capable of the depth camera of three-dimensional modeling.Build the three-dimensional of tunnel scene by the three-D imaging method of design
Mould.
The speed advanced according to complete-section tunnel boring machine may be implemented to determine itself in the mechanical device for rotating camera
The speed of rotation, it is ensured that certain positions of shooting tunnel space will not be missed.
Due to subterranean tunnel space Duo Shui, moist situation, general device is mounted on complete-section tunnel boring machine facility
It is easy to be influenced by severe external condition, leads to shortening for service life, therefore we will be in system using protective device
It is protected than more sensitive cam device, and this canning can't influence the shooting effect of camera.
By depth photo imaging detection go out tunnel relative to camera distance so as to building complete tunnel
Threedimensional model.
Above-mentioned, although the foregoing specific embodiments of the present invention is described with reference to the accompanying drawings, not protects model to the present invention
The limitation enclosed, those skilled in the art should understand that, based on the technical solutions of the present invention, those skilled in the art are not
Need to make the creative labor the various modifications or changes that can be made still within protection scope of the present invention.
Claims (10)
1. a kind of three-dimensional visialization of tunnel system based on depth camera, it is characterized in that：Including depth camera device, rotating dress
It sets, the depth camera device includes microphone assembly, infrared camera, colour imagery shot and infrared camera, the depth
Degree cam device is mounted on the rotating device, and the rotating device drives depth camera device to carry out on vertical direction
180 ° of rotation, acquire tunnel upper half image；
It is provided with angular transducer on the rotating device, detects the rotation angle of rotating device, while being arranged on rotating device
There are elevation-angle controller, the acquisition angles of controlling depth cam device.
2. a kind of three-dimensional visialization of tunnel system based on depth camera as described in claim 1, it is characterized in that：The depth
Cam device is set on bracket, and the bracket is fixed on rotating device.
3. a kind of three-dimensional visialization of tunnel system based on depth camera as described in claim 1, it is characterized in that：The depth
Cam device is provided with waterproof case.
4. based on the modeling method of system as claimed in any one of claims 1-3, it is characterized in that：Include the following steps：
(1) depth information and image information of the acquisition of depth camera device are obtained；
(2) acquired image is sampled, floating-point, smoothing techniques, depth camera device is obtained from image
Motion track；
(3) large-scale virtual space is established, the information of acquisition is fused in Virtual Space according to the sampling interval, forms three-dimensional mould
Type.
5. modeling method as claimed in claim 4, it is characterized in that：In the step (1), it is mounted in complete-section tunnel boring machine
On depth camera with before machine so that obtain tunnel depth and colour information, wherein the part for repeating to take can
As the optimization of photographic intelligence, tunnel information is as constantly added, the lower limit of redundancy can't be set.
6. modeling method as claimed in claim 4, it is characterized in that：In the step (2), it is arranged according to threshold value to depth image
Data are converted, and the data setting distance except threshold range is invalid.
7. modeling method as claimed in claim 4, it is characterized in that：In the step (2), calculated by using the registration of interactive
Method constantly obtains its posture when camera is mobile, and posture tracking is used will calculate the point Yun Yucong got from reconstructed object
The point cloud obtained in depth image data is registrated, or individually using the data to the different field angles of Same Scene into
Row registration.
8. modeling method as claimed in claim 4, it is characterized in that：In the step (2), according to the movement of depth camera device
The size of speed setting smoothing parameter, is to represent camera by the depth image data fusion generated from known poses camera
The cube of scenery within the vision, this fusion to depth data is frame by frame, to be carried out continuously, finally, from sensor
Viewpoint position carries out light projection, the point cloud being calculated using ray casting algorithm to cube is re-established, then calculates its normal direction
Amount is registrated with the input picture of point cloud and next frame with normal vector, calculates the pose of next frame input picture.
9. modeling method as claimed in claim 4, it is characterized in that：In the step (3), using offline place in modeling process
The mode of reason is optimized using winding detection and winding, according to winding optimization as a result, the coordinate of point is updated, so that the ground of winding
The result alignment that side is rebuild twice.
10. modeling method as claimed in claim 9, it is characterized in that：In the step (3), when camera rotates or mobile
When distance is more than threshold value, key frame is done into present frame addition and carries out winding detection, winding detection passes through searching first
Matched key frame, if there is matched image, by the SURF characteristic point and depth map of the matching image stored in memory
Come as indexing out again；The SURF of given two field pictures describes sub- Ui and Um, the matching relationship of SURF is established with FLANN, such as
The SURF point quantity that fruit can match is no more than given threshold, then it is assumed that this is not an effective matching, is built by matching
The matching relationship of vertical SURF estimates the pose between two frames with RANSAC algorithm by the matching between the SURF of previous step foundation,
It obtains after pose optimizing camera pose with LM algorithm optimization re-projection error again, the pose being calculated with ICP re-optimization, if
Error between match point is less than given threshold, then it is assumed that this is an effective winding.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710081447.XA CN106846479B (en) | 2017-02-15 | 2017-02-15 | Three-dimensional visialization of tunnel system and method based on depth camera |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710081447.XA CN106846479B (en) | 2017-02-15 | 2017-02-15 | Three-dimensional visialization of tunnel system and method based on depth camera |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106846479A CN106846479A (en) | 2017-06-13 |
| CN106846479B true CN106846479B (en) | 2018-11-30 |
Family
ID=59127287
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710081447.XA Active CN106846479B (en) | 2017-02-15 | 2017-02-15 | Three-dimensional visialization of tunnel system and method based on depth camera |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106846479B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107886129B (en) * | 2017-11-13 | 2021-06-08 | 湖南大学 | Mobile robot map closed-loop detection method based on visual word bag |
| CN110186651B (en) * | 2018-02-23 | 2021-08-03 | 宁波舜宇车载光学技术有限公司 | MTF test equipment for lens |
| CN110415329B (en) * | 2018-04-26 | 2023-10-13 | 财团法人工业技术研究院 | Three-dimensional modeling device and calibration method applied to same |
| CN109375263B (en) * | 2018-12-04 | 2020-04-21 | 山东大学 | Earthquake advanced prediction device, system and method suitable for drilling and blasting method tunnel |
| US11263809B2 (en) | 2019-09-06 | 2022-03-01 | Shandong University | TBM-mounted virtual reconstruction system and method for surrounding rock structure of tunnel |
| CN111258410B (en) * | 2020-05-06 | 2020-08-04 | 北京深光科技有限公司 | Man-machine interaction equipment |
| CN113284242B (en) * | 2021-06-18 | 2022-05-31 | 中铁隧道局集团有限公司 | Tunnel design system based on big data and AI |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413352A (en) * | 2013-07-29 | 2013-11-27 | 西北工业大学 | Scene three-dimensional reconstruction method based on RGBD multi-sensor fusion |
| CN104599314A (en) * | 2014-06-12 | 2015-05-06 | 深圳奥比中光科技有限公司 | Three-dimensional model reconstruction method and system |
| KR20160019613A (en) * | 2014-08-11 | 2016-02-22 | 한국전력공사 | Apparatus for modeling 3d of underground structure and method thereof |
-
2017
- 2017-02-15 CN CN201710081447.XA patent/CN106846479B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103413352A (en) * | 2013-07-29 | 2013-11-27 | 西北工业大学 | Scene three-dimensional reconstruction method based on RGBD multi-sensor fusion |
| CN104599314A (en) * | 2014-06-12 | 2015-05-06 | 深圳奥比中光科技有限公司 | Three-dimensional model reconstruction method and system |
| KR20160019613A (en) * | 2014-08-11 | 2016-02-22 | 한국전력공사 | Apparatus for modeling 3d of underground structure and method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 基于Kinect深度信息的实时三维重建和滤波算法研究;陈晓明等;《计算机应用研究》;20130430;第30卷(第4期);1216-1218 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106846479A (en) | 2017-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106846479B (en) | Three-dimensional visialization of tunnel system and method based on depth camera | |
| JP7427188B2 (en) | 3D pose acquisition method and device | |
| US10855936B2 (en) | Skeleton-based effects and background replacement | |
| US10605602B2 (en) | Surveying system | |
| US10484669B2 (en) | Inertial measurement unit progress estimation | |
| CN107836012B (en) | Projection image generation method and device, and mapping method between image pixel and depth value | |
| Ventura et al. | Wide-area scene mapping for mobile visual tracking | |
| US20190332866A1 (en) | Method and apparatus for 3-d auto tagging | |
| JP6974873B2 (en) | Devices and methods for retrieving depth information from the scene | |
| CN108961395A (en) | A method of three dimensional spatial scene is rebuild based on taking pictures | |
| US11928778B2 (en) | Method for human body model reconstruction and reconstruction system | |
| Negahdaripour et al. | Mosaic-based positioning and improved motion-estimation methods for automatic navigation of submersible vehicles | |
| JP4743818B2 (en) | Image processing apparatus, image processing method, and computer program | |
| RU2572637C2 (en) | Parallel or serial reconstructions in online and offline modes for 3d measurements of rooms | |
| US10659686B2 (en) | Conversion of an interactive multi-view image data set into a video | |
| CN109461208A (en) | Three-dimensional map processing method, device, medium and calculating equipment | |
| WO2018019272A1 (en) | Method and apparatus for realizing augmented reality on the basis of plane detection | |
| CN110319772A (en) | Visual large-span distance measurement method based on unmanned aerial vehicle | |
| KR20210058686A (en) | Device and method of implementing simultaneous localization and mapping | |
| CN109508679A (en) | Realize method, apparatus, equipment and the storage medium of eyeball three-dimensional eye tracking | |
| WO2023159668A1 (en) | System and method of capturing large scale scenes using wearable inertial measurement devices and light detection and ranging sensors | |
| WO2019119597A1 (en) | Method for implementing planar recording and panoramic recording by coordination between mobile terminal and lens assembly and lens assembly | |
| CN109166181A (en) | A kind of mixing motion capture system based on deep learning | |
| CN108253962A (en) | New energy pilotless automobile localization method under a kind of low light environment | |
| Gomez-Lahoz et al. | Recovering traditions in the digital era: the use of blimps for modelling the archaeological cultural heritage |
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 |