CN110095111A - A kind of construction method of map scene, building system and relevant apparatus - Google Patents
A kind of construction method of map scene, building system and relevant apparatus Download PDFInfo
- Publication number
- CN110095111A CN110095111A CN201910388821.XA CN201910388821A CN110095111A CN 110095111 A CN110095111 A CN 110095111A CN 201910388821 A CN201910388821 A CN 201910388821A CN 110095111 A CN110095111 A CN 110095111A
- Authority
- CN
- China
- Prior art keywords
- key frame
- picture
- scene
- map
- zone
- 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
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C11/00—Photogrammetry or videogrammetry, e.g. stereogrammetry; Photographic surveying
- G01C11/04—Interpretation of pictures
- G01C11/06—Interpretation of pictures by comparison of two or more pictures of the same area
- G01C11/28—Special adaptation for recording picture point data, e.g. for profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/07—Target detection
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Theoretical Computer Science (AREA)
- Processing Or Creating Images (AREA)
Abstract
The application provides a kind of construction method of map scene, comprising: obtains scene picture；Camera pose and the first key frame are obtained using ORB point feature algorithm process scene picture；Zone of mutual visibility domain is determined according to the first key frame, is deleted the zone of mutual visibility domain in the first key frame, is obtained the second key frame；Utilize the second key frame and camera pose building point cloud three-dimensional map.Utilize the total view relationship between picture, it allows total depending on the method that is removed in map of part, so that its content of the three-dimensional map of foundation is constant, the more original map of size can reduce very much, facilitate and carry out subsequent exploitation and check, solves the problems, such as in three-dimensional map due to the excessive caused excessive more contents that are but beyond expression of map of redundant points.The application also provides building system, a kind of computer readable storage medium and a kind of terminal of a kind of map scene, has above-mentioned beneficial effect.
Description
Technical field
This application involves three-dimensional map design field, in particular to a kind of construction method of map scene, building system and
Relevant apparatus.
Background technique
Immediately positioning with build figure (Simultaneously Localization and Mapping, SLAM) be one by
Studying a question for favor, is the critical problem that intelligent robot realizes independent navigation in circumstances not known.SLAM passes through processing
Sensor information achievees the purpose that while constructing environment map and positioning in real time.And common sensor has in vision SLAM
It is more cheap to have compared laser sensor its cost for clear advantage.
In practical applications, visual sensor is big with measurement range, acquisition abundant information, cost performance is high, versatility is good,
The features such as recognizable object, it can obtain the information of the rich and varied property such as color, shape, texture of spatial scene, can be very
It is easy to obtain the information of the rich and varied property such as color, shape, the texture of spatial scene, can easily extract the side of scene
The characteristic informations such as edge, angle point, block are very suitable to identification and matching application in the picture.And for general vision SLAM
The object of environment where its figure of founding a capital assumes that it is static and will not move, for building chart now simultaneously under dynamic environment
It is bad, at present for its solution of vision SLAM under dynamic environment generally by fusion Inertial Measurement Unit, Huo Zhetong
Cross cooperate with laser sensor build figure then to map merge.However according to duplicate in the three-dimensional map of key frame foundation
Point is excessive, is difficult to carry out subsequent map exploitation.
Summary of the invention
The purpose of the application is to provide a kind of construction method of map scene, building system, a kind of computer-readable storage
Medium and a kind of terminal solve the problems, such as that repeatedly point is excessive in existing three-dimensional map.
In order to solve the above technical problems, the application provides a kind of construction method of map scene, specific technical solution is as follows:
Obtain scene picture；
Camera pose and the first key frame are obtained using scene picture described in ORB point feature algorithm process；
Zone of mutual visibility domain is determined according to first key frame, is deleted the zone of mutual visibility domain in first key frame, is obtained
To the second key frame；
Utilize second key frame and camera pose building point cloud three-dimensional map.
Wherein, described to obtain phase using scene picture described in ORB point feature algorithm process after obtaining the scene picture
Before seat in the plane appearance and the first key frame, further includes:
Identify the dynamic object in the scene picture；
The target area where the dynamic object is detected to judge whether there is missing inspection picture；
The missing inspection picture if it exists corrects the missing inspection picture using the kinematic data of the dynamic object.
Wherein, before obtaining camera pose and the first key frame using scene picture described in ORB point feature algorithm process, also
Include:
Utilize the gray scale of the unified scene picture of rectangle function in the library OPENCV.
Wherein, determine that zone of mutual visibility domain includes: according to first key frame
Zone of mutual visibility domain is determined using random sampling unification algorism according to first key frame.
The application also provides a kind of building system of map scene, comprising:
Module is obtained, for obtaining scene picture；
Processing module, for obtaining camera pose and the first key using scene picture described in ORB point feature algorithm process
Frame；
Removing module deletes the institute in first key frame for determining zone of mutual visibility domain according to first key frame
Zone of mutual visibility domain is stated, the second key frame is obtained；
Module is constructed, for utilizing second key frame and camera pose building point cloud three-dimensional map.
Wherein, further includes:
Identification module, for identification dynamic object in the scene picture；
Detection module, for detecting the target area where the dynamic object to judge whether there is missing inspection picture；
Correction module, for the missing inspection picture if it exists, described in the kinematic data amendment using the dynamic object
Missing inspection picture.
Wherein, further includes:
Without feature processing block, for the ash using the unified scene picture of rectangle function in the library OPENCV
Degree.
Wherein, the removing module includes:
Zone of mutual visibility domain determination unit, for determining zone of mutual visibility using random sampling unification algorism according to first key frame
Domain.
The application also provides a kind of computer readable storage medium, is stored thereon with computer program, the computer journey
The step of construction method as described above is realized when sequence is executed by processor.
The application also provides a kind of terminal, including memory and processor, has computer program in the memory, institute
State the step of realizing construction method as described above when processor calls the computer program in the memory.
The construction method of a kind of map scene provided herein, comprising: obtain scene picture；Utilize ORB point feature
Scene picture described in algorithm process obtains camera pose and the first key frame；Zone of mutual visibility domain is determined according to first key frame,
The zone of mutual visibility domain in first key frame is deleted, the second key frame is obtained；Utilize second key frame and the phase
Seat in the plane appearance building point cloud three-dimensional map.
The application utilizes the total view relationship between picture, allows the method that removes in map of total view part, so that the three of foundation
Dimension its content of map is constant, and the more original map of size can reduce very much, facilitates and carries out subsequent exploitation and check solve three
Dimension map in due to redundant points excessively caused by map it is excessive be but beyond expression more contents the problem of.The application also provides one kind
Building system, a kind of computer readable storage medium and a kind of terminal of map scene have above-mentioned beneficial effect, herein no longer
It repeats.
Detailed description of the invention
In order to illustrate the technical solutions in the embodiments of the present application or in the prior art more clearly, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of application for those of ordinary skill in the art without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is a kind of flow chart of the construction method of map scene provided by the embodiment of the present application；
Fig. 2 is a kind of building system structure diagram of map scene provided by the embodiment of the present application.
Specific embodiment
To keep the purposes, technical schemes and advantages of the embodiment of the present application clearer, below in conjunction with the embodiment of the present application
In attached drawing, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described embodiment is
Some embodiments of the present application, instead of all the embodiments.Based on the embodiment in the application, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall in the protection scope of this application.
Vision SLAM is broadly divided into 2 kinds of methods at present, is the method based on filtering and the method based on characteristic point respectively.?
On the basis of this, there is the inertia vision SLAM for using inertial sensor as supplement, also have and substitute SLAM system using deep learning
The deep learning SLAM of some part of the inside also has the semantic SLAM that feature is added to semantic information.
In method based on filtering, SLAM (the Struct SLAM:Visual SLAM of the structure lines based on building interior
With Building Structure Lines), solving thought is to be examined the structure lines of building interior by the method for filtering
It measures and, as the basis for building figure.Figure robustness is preferable for building under dynamic environment for the method, but its scene used has
Limit.Its scene needs to meet the Manhattan world it is assumed that there are the lines in three dominant directions, compares for spacious scene
It can greatly reduce as met its structure lines for building figure for outdoor or large-scale warehouse.
Referring to FIG. 1, Fig. 1 is a kind of flow chart of the construction method of map scene provided by the embodiment of the present application, it should
Construction method includes:
S101: scene picture is obtained；
S102: camera pose and the first key frame are obtained using scene picture described in ORB point feature algorithm process；
ORB (Oriented FASTand Rotated BRIEF) algorithm be current most fast and stable characteristic point detection and
Extraction algorithm, many image mosaics and target tracking technology are realized using ORB feature.
This step is intended to obtain camera using ORB point feature algorithm (namely ORB-SLAM2 algorithm) processing scene picture
Pose and the first key frame.It should be noted that if directly carrying out three-dimensional map using the first key frame obtained in this step
Building, can make in three-dimensional map that duplicate point is excessive, lead to not carry out subsequent map exploitation.
S103: zone of mutual visibility domain is determined according to first key frame, deletes the zone of mutual visibility in first key frame
Domain obtains the second key frame；
Based on S102, this step is intended to determine the zone of mutual visibility domain of the first key frame, as its name suggests, i.e., duplicate region.?
How this is for determining that zone of mutual visibility domain is not construed as limiting, such as can use random sampling unification algorism and determine zone of mutual visibility domain.True
After determining zone of mutual visibility domain, zone of mutual visibility domain is deleted.
It should be noted that being directed to the deletion in zone of mutual visibility domain, it is not necessary to all zone of mutual visibility domains are deleted, but weight
Point is directed to the zone of mutual visibility domain of redundancy, it is generally the case that there are part zone of mutual visibility domains certain auxiliary in the building of three-dimensional map
Effect, these zone of mutual visibility domains for being conducive to three-dimensional map building can not delete.
First key frame obtains the second key frame after this step process, regardless of whether first key frame is deleted altogether
Viewed area is regarded as the second key frame after this step process.
S104: second key frame and camera pose building point cloud three-dimensional map are utilized.
This step constructs point cloud three-dimensional map according to the pose of the second key frame and camera, herein for wherein constructing
Point cloud three-dimensional map is not construed as limiting, such as some cloud three-dimensional maps can be created in PCL point cloud library.
The embodiment of the present application utilizes the total view relationship between picture, allows the total view partially method that removes in map, so that building
Its content of vertical three-dimensional map is constant, and the more original map of size can reduce very much, facilitates and carries out subsequent exploitation and check, solves
Determined in three-dimensional map due to redundant points excessively caused by map it is excessive be but beyond expression more contents the problem of.
Based on the above embodiment, as preferred embodiment, between S101 and S102, can also include:
S201: the dynamic object in the scene picture is identified；
This step usually can use deep learning algorithm of target detection and identify dynamic object in scene picture, herein
Which kind of deep learning algorithm of target detection is not especially limited.
S202: the target area where the dynamic object is detected to judge whether there is missing inspection picture；
This step can use the target area where sliding window algorithm detection dynamic object, to judge whether there is leakage
Examine picture.
S203: the missing inspection picture if it exists corrects the missing inspection picture using the kinematic data of the dynamic object.
Although the case where can reach quickly detection at present identifies target, and there is also missing inspections, when missing inspection occurs
Figure can be built to three-dimensional by, which waiting, causes very big interference.Based on this, a kind of method for correcting picture is present embodiments provided.
Due to the video or picture stream of acquisition be all it is continuous in time, so the movement of object be also it is continuous,
Just appearing in certain period of time when there is missing inspection has contrast with other pictures, utilizes this phenomenon, the present embodiment
The method of proposition can use the kinematics of dynamic object to correct missing inspection picture, reduces and builds figure bring to three-dimensional due to missing inspection
Unnecessary trouble.
Based on the above embodiment, as preferred embodiment, before S102, can also include:
Utilize the gray scale of the unified scene picture of rectangle function in the library OPENCV.
This step is intended to carry out without characteristic processing, herein for the concrete mode or means of no characteristic processing scene picture
It is not construed as limiting, such as the rectangle function that can use in the library OPENCV is unified by the gray scale in region.
The various embodiments described above are then based on, the construction method process of another preferred map scene of the application can be such that
Step 1 obtains scene picture；
Dynamic object in step 2, the identification scene picture；
Target area where step 3, the detection dynamic object is to judge whether there is missing inspection picture；
Step 4, if it exists the missing inspection picture correct the missing inspection figure using the kinematic data of the dynamic object
Piece；The missing inspection picture if it does not exist, is directly entered step 5；
Step 5 utilizes the gray scale of the unified scene picture of rectangle function in the library OPENCV；
Step 6 obtains camera pose and the first key frame using scene picture described in ORB point feature algorithm process；
Step 7 determines zone of mutual visibility domain according to first key frame, deletes the zone of mutual visibility in first key frame
Domain obtains the second key frame；
Step 8 utilizes second key frame and camera pose building point cloud three-dimensional map.
Then the embodiment of the present application has the advantages that
(1) dynamic object identification efficiently and is easily completed using deep learning target detection, avoids dynamic environment
The problem of object contributions vision SLAM system of lower movement.
(2) abnormal frame is compensated by sliding window, can failed to avoid because being detected caused by the other factors such as fuzzy pictures,
Key effect is played to building three-dimensional map.
(3) extra content is got rid of depending on relationship using total between picture, so that three-dimensional map its content established is constant,
The more original map of size can reduce very much, facilitate and carry out subsequent exploitation and check.
A kind of building system of map scene provided by the embodiments of the present application is introduced below, building described below
System can correspond to each other reference with a kind of above-described construction method of map scene.
Referring to fig. 2, Fig. 2 is a kind of building system structure diagram of map scene provided by the embodiment of the present application, should
Building system may include:
Module 100 is obtained, for obtaining scene picture；
Processing module 200 is closed for obtaining camera pose and first using scene picture described in ORB point feature algorithm process
Key frame；
Removing module 300 is deleted in first key frame for determining zone of mutual visibility domain according to first key frame
The zone of mutual visibility domain, obtains the second key frame；
Module 400 is constructed, for utilizing second key frame and camera pose building point cloud three-dimensional map.
Based on the above embodiment, as preferred embodiment, the building system can also include:
Identification module, for identification dynamic object in the scene picture；
Detection module, for detecting the target area where the dynamic object to judge whether there is missing inspection picture；
Correction module, for the missing inspection picture if it exists, described in the kinematic data amendment using the dynamic object
Missing inspection picture.
Based on the above embodiment, as preferred embodiment, the building system can also include:
Without feature processing block, for the ash using the unified scene picture of rectangle function in the library OPENCV
Degree.
Based on the above embodiment, as preferred embodiment, the removing module 300 includes:
Zone of mutual visibility domain determination unit, for determining zone of mutual visibility using random sampling unification algorism according to first key frame
Domain.
Present invention also provides a kind of computer readable storage mediums, have computer program thereon, the computer program
It is performed and step provided by above-described embodiment may be implemented.The storage medium may include: USB flash disk, mobile hard disk, read-only deposit
Reservoir (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or
The various media that can store program code such as CD.
Present invention also provides a kind of terminals, may include memory and processor, have computer in the memory
When the processor calls the computer program in the memory, step provided by above-described embodiment is may be implemented in program.
Certain terminal can also include various network interfaces, the components such as power supply.
Each embodiment is described in a progressive manner in specification, the highlights of each of the examples are with other realities
The difference of example is applied, the same or similar parts in each embodiment may refer to each other.For embodiment provide system and
Speech, since it is corresponding with the method that embodiment provides, so being described relatively simple, related place is referring to method part illustration
?.
Specific examples are used herein to illustrate the principle and implementation manner of the present application, and above embodiments are said
It is bright to be merely used to help understand the present processes and its core concept.It should be pointed out that for the ordinary skill of the art
For personnel, under the premise of not departing from the application principle, can also to the application, some improvement and modification can also be carried out, these improvement
It is also fallen into the protection scope of the claim of this application with modification.
It should also be noted that, in the present specification, relational terms such as first and second and the like be used merely to by
One entity or operation are distinguished with another entity or operation, without necessarily requiring or implying these entities or operation
Between there are any actual relationship or orders.Moreover, the terms "include", "comprise" or its any other variant meaning
Covering non-exclusive inclusion, so that the process, method, article or equipment for including a series of elements not only includes that
A little elements, but also including other elements that are not explicitly listed, or further include for this process, method, article or
The intrinsic element of equipment.In the absence of more restrictions, the element limited by sentence "including a ...", is not arranged
Except there is also other identical elements in the process, method, article or apparatus that includes the element.
Claims (10)
1. a kind of construction method of map scene characterized by comprising
Obtain scene picture；
Camera pose and the first key frame are obtained using scene picture described in ORB point feature algorithm process；
Zone of mutual visibility domain is determined according to first key frame, deletes the zone of mutual visibility domain in first key frame, obtains the
Two key frames；
Utilize second key frame and camera pose building point cloud three-dimensional map.
2. construction method according to claim 1, which is characterized in that described to utilize ORB after obtaining the scene picture
Scene picture described in point feature algorithm process obtains before camera pose and the first key frame, further includes:
Identify the dynamic object in the scene picture；
The target area where the dynamic object is detected to judge whether there is missing inspection picture；
The missing inspection picture if it exists corrects the missing inspection picture using the kinematic data of the dynamic object.
3. construction method according to claim 1, which is characterized in that utilize scene figure described in ORB point feature algorithm process
Piece obtains before camera pose and the first key frame, further includes:
Utilize the gray scale of the unified scene picture of rectangle function in the library OPENCV.
4. construction method according to claim 1, which is characterized in that determine that zone of mutual visibility domain is wrapped according to first key frame
It includes:
Zone of mutual visibility domain is determined using random sampling unification algorism according to first key frame.
5. a kind of building system of map scene characterized by comprising
Module is obtained, for obtaining scene picture；
Processing module, for obtaining camera pose and the first key frame using scene picture described in ORB point feature algorithm process；
Removing module is deleted described total in first key frame for determining zone of mutual visibility domain according to first key frame
Viewed area obtains the second key frame；
Module is constructed, for utilizing second key frame and camera pose building point cloud three-dimensional map.
6. building system according to claim 5, which is characterized in that further include:
Identification module, for identification dynamic object in the scene picture；
Detection module, for detecting the target area where the dynamic object to judge whether there is missing inspection picture；
Correction module corrects the missing inspection using the kinematic data of the dynamic object for the missing inspection picture if it exists
Picture.
7. building system according to claim 5, which is characterized in that further include:
Without feature processing block, for the gray scale using the unified scene picture of rectangle function in the library OPENCV.
8. building system according to claim 5, which is characterized in that the removing module includes:
Zone of mutual visibility domain determination unit, for determining zone of mutual visibility domain using random sampling unification algorism according to first key frame.
9. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program quilt
The step of processor realizes construction method according to any one of claims 1-4 when executing.
10. a kind of terminal, which is characterized in that including memory and processor, there is computer program in the memory, it is described
Processor realizes the step of construction method according to any one of claims 1-4 when calling the computer program in the memory
Suddenly.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910388821.XA CN110095111A (en) | 2019-05-10 | 2019-05-10 | A kind of construction method of map scene, building system and relevant apparatus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910388821.XA CN110095111A (en) | 2019-05-10 | 2019-05-10 | A kind of construction method of map scene, building system and relevant apparatus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110095111A true CN110095111A (en) | 2019-08-06 |
Family
ID=67447635
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910388821.XA Pending CN110095111A (en) | 2019-05-10 | 2019-05-10 | A kind of construction method of map scene, building system and relevant apparatus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110095111A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110827395A (en) * | 2019-09-09 | 2020-02-21 | 广东工业大学 | Instant positioning and map construction method suitable for dynamic environment |
| CN111914715A (en) * | 2020-07-24 | 2020-11-10 | 廊坊和易生活网络科技股份有限公司 | Intelligent vehicle target real-time detection and positioning method based on bionic vision |
| CN115376051A (en) * | 2022-10-25 | 2022-11-22 | 杭州华橙软件技术有限公司 | Key frame management method and device, SLAM method and electronic equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831439A (en) * | 2012-08-15 | 2012-12-19 | 深圳先进技术研究院 | Gesture tracking method and gesture tracking system |
| CN106885574A (en) * | 2017-02-15 | 2017-06-23 | 北京大学深圳研究生院 | A kind of monocular vision robot synchronous superposition method based on weight tracking strategy |
| JP2017224280A (en) * | 2016-05-09 | 2017-12-21 | ツーアンツ インク．ＴｗｏＡｎｔｚ Ｉｎｃ． | Visual positioning-based navigation apparatus and method |
| CN107545582A (en) * | 2017-07-04 | 2018-01-05 | 深圳大学 | Video multi-target tracking and device based on fuzzy logic |
| CN108961385A (en) * | 2017-05-22 | 2018-12-07 | 中国人民解放军信息工程大学 | A kind of SLAM patterning process and device |
| CN109307508A (en) * | 2018-08-29 | 2019-02-05 | 中国科学院合肥物质科学研究院 | A kind of panorama inertial navigation SLAM method based on more key frames |
| CN109579840A (en) * | 2018-10-25 | 2019-04-05 | 中国科学院上海微系统与信息技术研究所 | A kind of close coupling binocular vision inertia SLAM method of dotted line Fusion Features |
-
2019
- 2019-05-10 CN CN201910388821.XA patent/CN110095111A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102831439A (en) * | 2012-08-15 | 2012-12-19 | 深圳先进技术研究院 | Gesture tracking method and gesture tracking system |
| JP2017224280A (en) * | 2016-05-09 | 2017-12-21 | ツーアンツ インク．ＴｗｏＡｎｔｚ Ｉｎｃ． | Visual positioning-based navigation apparatus and method |
| CN106885574A (en) * | 2017-02-15 | 2017-06-23 | 北京大学深圳研究生院 | A kind of monocular vision robot synchronous superposition method based on weight tracking strategy |
| CN108961385A (en) * | 2017-05-22 | 2018-12-07 | 中国人民解放军信息工程大学 | A kind of SLAM patterning process and device |
| CN107545582A (en) * | 2017-07-04 | 2018-01-05 | 深圳大学 | Video multi-target tracking and device based on fuzzy logic |
| CN109307508A (en) * | 2018-08-29 | 2019-02-05 | 中国科学院合肥物质科学研究院 | A kind of panorama inertial navigation SLAM method based on more key frames |
| CN109579840A (en) * | 2018-10-25 | 2019-04-05 | 中国科学院上海微系统与信息技术研究所 | A kind of close coupling binocular vision inertia SLAM method of dotted line Fusion Features |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110827395A (en) * | 2019-09-09 | 2020-02-21 | 广东工业大学 | Instant positioning and map construction method suitable for dynamic environment |
| CN110827395B (en) * | 2019-09-09 | 2023-01-20 | 广东工业大学 | Instant positioning and map construction method suitable for dynamic environment |
| CN111914715A (en) * | 2020-07-24 | 2020-11-10 | 廊坊和易生活网络科技股份有限公司 | Intelligent vehicle target real-time detection and positioning method based on bionic vision |
| CN115376051A (en) * | 2022-10-25 | 2022-11-22 | 杭州华橙软件技术有限公司 | Key frame management method and device, SLAM method and electronic equipment |
| CN115376051B (en) * | 2022-10-25 | 2023-03-24 | 杭州华橙软件技术有限公司 | Key frame management method and device, SLAM method and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111724439B (en) | Visual positioning method and device under dynamic scene | |
| CN110095111A (en) | A kind of construction method of map scene, building system and relevant apparatus | |
| CN110276826A (en) | A kind of construction method and system of electric network operation environmental map | |
| CN111735446B (en) | Laser and visual positioning fusion method and device | |
| CN110264563A (en) | A kind of Octree based on ORBSLAM2 builds drawing method | |
| CN111951397A (en) | Method, device and storage medium for multi-machine cooperative construction of three-dimensional point cloud map | |
| CN109815847B (en) | Visual SLAM method based on semantic constraint | |
| CN112686877B (en) | Binocular camera-based three-dimensional house damage model construction and measurement method and system | |
| AU2014203125A1 (en) | Automated graph local constellation (GLC) method of correspondence search for registration of 2-D and 3-D data | |
| WO2022241803A1 (en) | Attention mechanism-based system and method for detecting feature in target, and storage medium | |
| CN111797836A (en) | Extraterrestrial celestial body patrolling device obstacle segmentation method based on deep learning | |
| CN114419028A (en) | Transmission line insulator defect duplication removing method and device integrating space multiple visual angles | |
| CN109447902A (en) | A kind of image split-joint method, device, storage medium and equipment | |
| CN114387658A (en) | Image target attribute detection method, device, equipment and storage medium | |
| CN112802104B (en) | Loop detection method based on RGB-D camera | |
| CN111738085B (en) | System construction method and device for realizing automatic driving simultaneous positioning and mapping | |
| CN116433580A (en) | Dam crack detection method based on intelligent unmanned aerial vehicle | |
| CN110345919A (en) | Space junk detection method based on three-dimensional space vector and two-dimensional plane coordinate | |
| CN105930813B (en) | A method of detection composes a piece of writing this under any natural scene | |
| CN110930519B (en) | Semantic ORB-SLAM sensing method and device based on environment understanding | |
| CN112419384B (en) | Digital earth surface model house DLG automatic extraction method based on artificial intelligence | |
| CN113673484A (en) | Road condition identification and decision-making method in unmanned driving scene | |
| CN114387351A (en) | Monocular vision calibration method and computer readable storage medium | |
| KR102178835B1 (en) | Apparatus and method for data management for reconstruct in 3d object surface | |
| Sok et al. | Visually aided feature extraction from 3D range data |
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 |
Application publication date: 20190806 |
|
| RJ01 | Rejection of invention patent application after publication |