CN110297936A - A kind of automatic geographical semantics localization method of video image and system - Google Patents
A kind of automatic geographical semantics localization method of video image and system Download PDFInfo
- Publication number
- CN110297936A CN110297936A CN201910546530.9A CN201910546530A CN110297936A CN 110297936 A CN110297936 A CN 110297936A CN 201910546530 A CN201910546530 A CN 201910546530A CN 110297936 A CN110297936 A CN 110297936A
- Authority
- CN
- China
- Prior art keywords
- image
- video
- information
- location information
- geographical semantics
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/29—Geographical information databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/587—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using geographical or spatial information, e.g. location
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Remote Sensing (AREA)
- Library & Information Science (AREA)
- Processing Or Creating Images (AREA)
Abstract
The present invention provides a kind of automatic geographical semantics localization method of video image and systems.Wherein method is the following steps are included: obtain the first digital video undetermined；According to the first multiple first images of video acquisition to be positioned；First image is filtered with after feature extraction processing, the first image feature information and the second image feature information are obtained；First graphic feature information, the second image feature information are matched with image geographical semantics library, to obtain the final location information of the first image；Successively each final location information of first image is integrated and handled by video time sequencing, and then obtains the first video location information.The present invention can position the spot of video content, and obtain the routing information of entire video according to the video of the street photographs in cities and towns, while third party personnel also being enable to be accurately positioned each position in path in video.
Description
Technical field
The present invention relates to a kind of video image identification and positioning fields, and in particular to a kind of automatic geographical semantics of video image
Localization method and system.
Background technique
With the fast development of computer science and technology, image recognition technology starts to be widely used in every field.Together
When, the arrival of information age and the popular of big data are accumulated with the video image data of magnanimity is made, wherein video
Image data contains extremely abundant and valuable information, is widely used in people's lives work, such as video monitoring
The fields such as social public security, criminal investigation and case detection.
Currently, there is " recognition methods in geographical location and the device " such as number of patent application for CN201410332517.0
This kind of patent, such patent be used for the place weak in building interior or satellite positioning device signal, according to user scene clapped
Video or image carry out precise positioning to it, but this method needs to cooperate satellite positioning device to come to entering building or street
The filming apparatus in road carries out satellite positioning, to obtain position such as place city, cities and towns, street etc. substantially, to make other people
Member can not realize positioning according to video or picture material merely, have limitations.Meanwhile although video image description ten
Divide concrete image lively, however these video images largely still lack geographical semantics information, so can not be to each in video
The accurate coordinate setting of Duan Jinhang and geographical semantics obtain.
It therefore, can be according in cities and towns the present invention provides a kind of automatic geographical semantics localization method of video image and system
Street photographs video, the spot of video content is positioned, and obtain the routing information of entire video, while
Third party personnel are enable to be accurately positioned each position in path in video.Be conducive to Claims Resolution, the personnel safety of insurance
Protection, the formulation in navigation system path.
Summary of the invention
In order to solve the above-mentioned technical problems, the present invention provides a kind of automatic geographical semantics localization methods of video image in being
System.
The automatic geographical semantics localization method of a kind of video image provided by the invention is the following steps are included: acquisition first is undetermined
Digital video；According to the first multiple first images of video acquisition to be positioned；First image is filtered after being handled with feature extraction,
Obtain the first image feature information and the second image feature information；By the first graphic feature information, the second image feature information with
Image geographical semantics library is matched, to obtain the final location information of the first image；It successively will be each by video time sequencing
A final location information of first image is integrated and is handled, and then obtains the first video location information.
Further, the first video location information includes the driving path of the first digital video undetermined, passes through in driving path
The geographical semantics information for each coordinate points crossed.
Further, the first image characteristic information includes road sign, the text on guideboard or graphical information, and described second
Image feature information includes the text or image information of the shop signboard occurred in image.
Further, when being free of the first image feature information in the first acquired image, stop to current first image
The final location information of the first image acquisition.
Further, described " by the first graphic feature information, the second image feature information and image geographical semantics library carry out
Match, to obtain the final location information of the first image " it specifically includes: the first graphic feature information and image geographical semantics library are carried out
Matching obtains the first candidate locations collection；When determining that the first candidate locations collection only has uniquity, it is finally fixed to determine that it is the first image
Position information, the final location information of the first image includes the geographical semantics and spatial coordinated information of this；Determine the first candidate
It is when ground point set has multiple subitems, the second image feature information is relevant with the first candidate locations collection to image geographical semantics library
Information matches, and then obtains the second Candidate Set；When the second Candidate Set there are multiple subitems, the default positioning letter of the first image is generated
Breath, and the first image is preset in location information deposit figure geography semantic information library；When the second Candidate Set only has uniquity,
Determine that it is the final location information of the first image.
Further, when being free of the second image feature information or the second Candidate Set in the second acquired image is empty set,
Stop obtaining the second Candidate Set.
Further, described " successively to integrate each final location information of first image by video time sequencing
With processing, and then obtain the first video location information " specifically: by video time sequencing will each first image finally calmly
Position information is integrated；It, will be last using the place of the first video location information obtained at first as the first video path starting point
The place of the first video location information obtained is as the first video path terminal；The first view is obtained in conjunction with image geographical semantics library
Frequency location information.
The present invention also provides a kind of system for realizing the above method, including video acquisition unit, for obtaining the
One digital video undetermined；Image extraction unit, for the first digital video undetermined to be decomposed into multiple first images frame by frame；Feature extraction
Unit is handled for being filtered to the first image with feature extraction, obtains the first image feature information or the second characteristics of image
Information；Data processing unit, for carrying out the first graphic feature information, the second image feature information and image geographical semantics library
Matching, to obtain the final location information of the first image；Generation unit, for successively pressing video time sequencing for each first
The final location information of image is integrated and is handled, and then generates the first video location information；Storage unit is used for storage driving
The software program of above-mentioned each unit.
Wherein, described image feature extraction unit can believe the road sign in the first image, the text on guideboard or figure
Breath carries out feature extraction and obtains the first image feature information, can by the text or image information of the shop signboard occurred in image
Feature extraction is carried out to obtain the second image feature information.
Wherein, it when feature extraction unit can not be according to the first image feature information be obtained in current first image, can control
Processing unit processed the first image corresponding with generation unit stopping processing.
The present invention obtains the first image by the decomposition frame by frame to the first digital video undetermined, then by each first figure
The processing of picture obtains the final location information of the first image, and then combines needed for all the final location informations of the first image obtain
The first video location information；To improve the serious forgiveness of this method, improves the driving path of acquisition, passes through in driving path
The precision degree of the geographical semantics information for each coordinate points crossed.
It is multiple that the present invention also passes through the first graphic feature information, the second image feature information and the progress of image geographical semantics library
Matching, while giving up non-compliant first image, to prevent the matching result of mistake to the first video location information
Influence.The extracted first video location information of the present invention can also be combined with the first image given up to be come, to having given up
The corresponding position of the first image abandoned is relocated, and the geographical semantics letter for each coordinate points passed through in driving path is realized
The precision of breath positions.
Detailed description of the invention
Fig. 1 is a kind of overall flow figure of the automatic geographical semantics localization method of video image of the invention；
Fig. 2 is the flow chart of the step S4 of the automatic geographical semantics localization method of video image of the invention a kind of；
Fig. 3 is the flow chart of the step S5 of the automatic geographical semantics localization method of video image of the invention a kind of；
Fig. 4 is the judgement block diagram of the step S4 of the automatic geographical semantics localization method of video image of the invention a kind of；
Fig. 5 is a kind of structural schematic diagram of the automatic geographical semantics positioning system of video image of the invention.
Specific embodiment
In order to illustrate the embodiments of the present invention more clearly and/or technical solution in the prior art, attached drawing will be compareed below
Illustrate a specific embodiment of the invention.It should be evident that drawings in the following description are only some embodiments of the invention,
For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings
Other attached drawings, and obtain other embodiments.In addition, the relative position belonged between only indicating each component for being related to orientation is closed
System, rather than absolute positional relation.
The scope of application of the video acted on of the invention is the position that cities and towns, tourist attractions etc. have infrastructure, together
When the of the invention video acted on be not suitable for the indoor environment of not common occasion, remote mountain village.Geographical semantics in the present invention
Positioning is exactly exposed place name in the form of verbal description, such as the steamed stuffed bun shop of South Road 62 in the area of wuchang, wuhan.
As shown in Figure 1, 2, 3, the embodiment provides a kind of automatic geographical semantics localization method of video image, packets
It includes step S1-S5:S1, obtain the first digital video undetermined；First digital video undetermined can be any video equipment shooting
The substantially distinguishable video of content.
S2, according to the first multiple first images of video acquisition to be positioned.It is undetermined by first especially by video intercepting software
Digital video is divided into multiple continuous first images frame by frame in chronological order.
S3, the first image is filtered with after feature extraction processing, obtains the first image feature information and the second image
Characteristic information；The step filters the disturbing factors such as people, vehicle, tree by Image filter arithmetic, is then calculated by image characteristics extraction
Method such as obtains the first image feature information including road sign, the text on guideboard or graphical information by SIFT algorithm, then obtains
Take the second image feature information including the text of shop signboard or image information that occur in image.
Wherein, the feature of each image can also be encoded by the hash algorithm of image coding technique, in reducing
Consumption is deposited, such coding is known approaches, can easily be understood by insider.
Meanwhile when being free of the first image feature information in the first acquired image, stop to current first image
The acquisition of the final location information of first image.And then subsequent other first images are started to process, this is used to improve
The working efficiency of the system of method, and corresponding first image can be ignored.
S4, the first graphic feature information, the second image feature information are matched with image geographical semantics library, to obtain
The final location information of first image.
Above-mentioned steps S4 detailed process can refer to Fig. 2 and Fig. 4, as shown in Fig. 2, step S4 is specifically included: S41, by first
Graphic feature information is matched with image geographical semantics library, obtains the first candidate locations collection；S42, the first candidate locations are determined
When collection only has uniquity, the final location information of the first image is determined that it is, the final location information of the first image includes this
Geographical semantics and spatial coordinated information；S43, when determining that the first candidate locations collection there are multiple subitems, the second characteristics of image is believed
With the first candidate locations collection, relevant information matches breath to image geographical semantics library, and then obtains the second Candidate Set；S44,
When the second Candidate Set there are multiple subitems, generates the first image and preset location information, and the first image is preset into location information and is deposited
Enter in figure geography semantic information library；S45, when the second Candidate Set only has uniquity, determine that it is the first image and finally position
Information.
Wherein, when being free of the second image feature information or the second Candidate Set in the second acquired image is empty set, stop
The second Candidate Set is only obtained, at this moment every location information of the first Candidate Set obtained can be stored in image geographical semantics library
In, improve the matching precision of subsequent pictures.
S5, successively by video time sequencing each final location information of first image is integrated and is handled, into
And obtain the first video location information.Wherein, the first video location information include the first digital video undetermined driving path,
The geographical semantics information for each coordinate points passed through in driving path.
As shown in figure 3, step S5 is specifically included: S51, finally positioning each first image by video time sequencing
Information is integrated；S52, using the place of the first video location information obtained at first as the first video path starting point, will most
The place of the first video location information obtained afterwards is as the first video path terminal；S53, it is obtained in conjunction with image geographical semantics library
First video location information.
Wherein, extracted first video location information can also be combined with the first image given up comes, to having given up
The corresponding position of the first image abandoned is relocated, and the geographical semantics letter for each coordinate points passed through in driving path is realized
The precision of breath positions.
A kind of system for realizing the above method is additionally provided in the embodiment of the present invention, the carrier of the system can be energy
The mainframe computer for handling mass data, such as Z/Architecture schema master.It is quick so as to be carried out to upper image data
Processing.Above-mentioned function can also be realized with personal computer using the mode of cloud computing.
As shown in figure 5, a kind of system is present embodiments provided, including video acquisition unit, it is to be positioned for obtaining first
Video；Image extraction unit, for the first digital video undetermined to be decomposed into multiple first images frame by frame；Feature extraction unit is used
It is handled in being filtered to the first image with feature extraction, obtains the first image feature information or the second image feature information；Number
Come according to processing unit for matching the first graphic feature information, the second image feature information with image geographical semantics library
Obtain the final location information of the first image；Generation unit is used to successively press video time sequencing by each first image most
Whole location information is integrated and is handled, and then generates the first video location information；Storage unit, it is above-mentioned each for storage driving
The software program of unit.
Wherein, described image feature extraction unit can believe the road sign in the first image, the text on guideboard or figure
Breath carries out feature extraction and obtains the first image feature information, can by the text or image information of the shop signboard occurred in image
Feature extraction is carried out to obtain the second image feature information.
The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be said that
Specific implementation of the invention is only limited to these instructions.For those of ordinary skill in the art to which the present invention belongs, exist
Under the premise of not departing from present inventive concept, a number of simple deductions or replacements can also be made, all shall be regarded as belonging to of the invention
Protection scope.
Claims (10)
1. a kind of automatic geographical semantics localization method of video image, which is characterized in that can position, wrap to video shoot location
It includes following steps: obtaining the first digital video undetermined；According to the first multiple first images of video acquisition to be positioned；To the first image into
Row filtering obtains the first image feature information and the second image feature information with after feature extraction processing；By the first graphic feature
Information, the second image feature information are matched with image geographical semantics library, to obtain the final location information of the first image；Successively
Each final location information of first image is integrated and handled by video time sequencing, and then it is fixed to obtain the first video
Position information.
2. a kind of automatic geographical semantics localization method of video image as described in claim 1, it is characterised in that: first view
Frequency location information includes the geographical semantics letter of the driving path of the first digital video undetermined, each coordinate points passed through in driving path
Breath.
3. a kind of automatic geographical semantics localization method of video image as described in claim 1, it is characterised in that: first figure
As characteristic information includes road sign, the text on guideboard or graphical information, second image feature information includes going out in image
The text or image information of existing shop signboard.
4. a kind of automatic geographical semantics localization method of video image as described in claim 1, it is characterised in that: when acquired
When being free of the first image feature information in the first image, stop obtaining to the final location information of the first image of current first image
It takes.
5. a kind of automatic geographical semantics localization method of video image as described in claim 1, it is characterised in that: described " by the
One graphic feature information, the second image feature information are matched with image geographical semantics library, finally fixed to obtain the first image
Position information " specifically includes: the first graphic feature information being matched with image geographical semantics library, obtains the first candidate locations
Collection；When determining that the first candidate locations collection only has uniquity, the final location information of the first image is determined that it is, the first image is most
Whole location information includes the geographical semantics and spatial coordinated information of this；It, will when determining that the first candidate locations collection has multiple subitems
With the first candidate locations collection, relevant information matches second image feature information to image geographical semantics library, and then obtains the
Two Candidate Sets；When the second Candidate Set there are multiple subitems, generates the first image and preset location information, and the first image is presetting
In the information deposit figure geography semantic information library of position；When the second Candidate Set only has uniquity, it is final to determine that it is the first image
Location information.
6. a kind of automatic geographical semantics localization method of video image as claimed in claim 5, it is characterised in that: when acquired
When without the second image feature information or the second Candidate Set being empty set in the second image, stop obtaining the second Candidate Set.
7. a kind of automatic geographical semantics localization method of video image as described in claim 1, which is characterized in that it is described " successively
Each final location information of first image is integrated and handled by video time sequencing, and then it is fixed to obtain the first video
Position information " specifically: integrate each final location information of first image by video time sequencing；It will obtain at first
The first video location information place as the first video path starting point, by the ground of the first video location information finally obtained
Point is used as the first video path terminal；The first video location information is obtained in conjunction with image geographical semantics library.
8. a kind of system of the automatic geographical semantics positioning of video image, it is characterised in that: including video acquisition unit, for obtaining
First digital video undetermined；Image extraction unit, for the first digital video undetermined to be decomposed into multiple first images frame by frame；Feature mentions
Unit is taken, is handled for being filtered to the first image with feature extraction, the first image feature information is obtained or the second image is special
Reference breath；Data processing unit, for by the first graphic feature information, the second image feature information and image geographical semantics library into
Row matching, to obtain the final location information of the first image；Generation unit, for successively by video time sequencing by each the
The final location information of one image is integrated and is handled, and then generates the first video location information；Storage unit is driven for storing
Move the software program of above-mentioned each unit.
9. a kind of system of the automatic geographical semantics positioning of video image as claimed in claim 8, it is characterised in that: described image
Feature extraction unit can carry out feature extraction to the road sign in the first image, the text on guideboard or graphical information to obtain the
The text of the shop signboard occurred in image or image information can be carried out feature extraction to obtain second by one image feature information
Image feature information.
10. a kind of system of the automatic geographical semantics positioning of video image as claimed in claim 8, it is characterised in that: work as feature
When extraction unit can not be according to the first image feature information be obtained in current first image, processing unit and generation unit can be controlled
Stop handling corresponding first image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910546530.9A CN110297936A (en) | 2019-06-24 | 2019-06-24 | A kind of automatic geographical semantics localization method of video image and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910546530.9A CN110297936A (en) | 2019-06-24 | 2019-06-24 | A kind of automatic geographical semantics localization method of video image and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110297936A true CN110297936A (en) | 2019-10-01 |
Family
ID=68028667
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910546530.9A Pending CN110297936A (en) | 2019-06-24 | 2019-06-24 | A kind of automatic geographical semantics localization method of video image and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110297936A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826568A (en) * | 2019-11-08 | 2020-02-21 | 青岛合启立智能科技有限公司 | Indoor positioning method and device based on road signs and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105246149A (en) * | 2014-07-11 | 2016-01-13 | 北京中星微电子有限公司 | Identifying method of geographic location and device |
| CN106032990A (en) * | 2015-03-21 | 2016-10-19 | 吴红平 | Working method of real scene navigation system |
| CN106227732A (en) * | 2016-07-08 | 2016-12-14 | 增城市城乡规划测绘院 | A kind of method of real-time acquisition mobile video photographed scene position |
| CN108020231A (en) * | 2016-10-28 | 2018-05-11 | 大辅科技（北京）有限公司 | A kind of map system and air navigation aid based on video |
-
2019
- 2019-06-24 CN CN201910546530.9A patent/CN110297936A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105246149A (en) * | 2014-07-11 | 2016-01-13 | 北京中星微电子有限公司 | Identifying method of geographic location and device |
| CN106032990A (en) * | 2015-03-21 | 2016-10-19 | 吴红平 | Working method of real scene navigation system |
| CN106227732A (en) * | 2016-07-08 | 2016-12-14 | 增城市城乡规划测绘院 | A kind of method of real-time acquisition mobile video photographed scene position |
| CN108020231A (en) * | 2016-10-28 | 2018-05-11 | 大辅科技（北京）有限公司 | A kind of map system and air navigation aid based on video |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110826568A (en) * | 2019-11-08 | 2020-02-21 | 青岛合启立智能科技有限公司 | Indoor positioning method and device based on road signs and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2010328337B2 (en) | Matching an approximately located query image against a reference image set | |
| CN104106260B (en) | Control based on geographical map | |
| CN106339428A (en) | Identity identification method and device for suspects based on large video data | |
| EP2975555A1 (en) | Method and apparatus for displaying point of interest | |
| CN103699677B (en) | A kind of criminal's whereabouts mapping system and method based on face recognition technology | |
| CN107084740B (en) | Navigation method and device | |
| CN110659391A (en) | Video detection method and device | |
| US9836826B1 (en) | System and method for providing live imagery associated with map locations | |
| US9529803B2 (en) | Image modification | |
| US20100134486A1 (en) | Automated Display and Manipulation of Photos and Video Within Geographic Software | |
| CN110245563A (en) | Refitted car recognition methods and Related product | |
| EP4224338A1 (en) | Data labeling method, apparatus, and system, device, and storage medium | |
| WO2023160722A1 (en) | Interactive target object searching method and system and storage medium | |
| CN114549595A (en) | Data processing method and device, electronic equipment and storage medium | |
| CN110297936A (en) | A kind of automatic geographical semantics localization method of video image and system | |
| EP4083815A1 (en) | Image search method and apparatus | |
| CN109767645A (en) | A kind of parking planning householder method and system based on AR glasses | |
| CN108827326A (en) | A kind of acquisition method and its acquisition device of the navigation map based on big data | |
| CN112528265A (en) | Identity recognition method, device, equipment and medium based on online conference | |
| CN112257638A (en) | Image comparison method, system, equipment and computer readable storage medium | |
| CN111083446A (en) | Tracking and positioning method, system and equipment combining visual signal and wireless signal | |
| CN111274435A (en) | Video backtracking method and device, electronic equipment and readable storage medium | |
| CN105472332B (en) | Analysis method based on location technology and video technique and its analysis system | |
| CN109117765A (en) | Video investigation device and method | |
| JP6704134B1 (en) | Information processing apparatus and method |
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: 20191001 |
|
| RJ01 | Rejection of invention patent application after publication |