CN106885580A - Localization method and device based on shop signboard in electronic map - Google Patents
Localization method and device based on shop signboard in electronic map Download PDFInfo
- Publication number
- CN106885580A CN106885580A CN201510934763.8A CN201510934763A CN106885580A CN 106885580 A CN106885580 A CN 106885580A CN 201510934763 A CN201510934763 A CN 201510934763A CN 106885580 A CN106885580 A CN 106885580A
- Authority
- CN
- China
- Prior art keywords
- shop
- information
- interest
- point
- sequence
- 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
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/26—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 specially adapted for navigation in a road network
- G01C21/34—Route searching; Route guidance
- G01C21/3453—Special cost functions, i.e. other than distance or default speed limit of road segments
- G01C21/3476—Special cost functions, i.e. other than distance or default speed limit of road segments using point of interest [POI] information, e.g. a route passing visible POIs
Abstract
The invention discloses localization method and device based on shop signboard in a kind of electronic map, it is related to mobile terminal positioning technology, methods described includes：The store information in shop near its position of acquisition for mobile terminal；By searching electronic map database, the shop interest point information matched with acquired store information is found；The corresponding positional information of the shop interest point information is included in the electronic map.The store information that the present invention is swept or be manually entered by mobile terminal to the random ring of surrounding environment, matching and lookup store information location information, realize the positive location of user in database, and positioning precision is high, is that user brings preferably positioning experience.
Description
Technical field
The present invention relates to mobile terminal positioning technology, the positioning side based on shop signboard more particularly in electronic map
Method and device.
Background technology
In mobile terminal location application, existing localization method is divided into two kinds of outdoor positioning and indoor positioning.Its
In, global positioning system (Global Positioning System, GPS), telecommunications are passed through outdoor positioning more
Net cellular basestation, or wifi signals, position with reference to inertial navigation system to mobile terminal, indoor many
Mobile terminal is positioned using technologies such as bluetooth, ultra wide bands.The positioning precision of above-mentioned technology is difficult to meet
Between the location requirement of complicated urban environment or indoor environment, especially City Building or interior of building
To GPS, telecommunications network, wifi, bluetooth, the signal blocks of ultra wide band and reflection so that positioning precision is further
Reduce, or even positioning failure, cause inconvenience to the user.
The lookup of point of interest (Point of Interest, POI) is a common technology in electronic map,
Mainly the keyword of user input is found by way of text matches in map, and POI is shown on map
Position.This mode may find the position of multiple POI, it is impossible to the need for meeting positioning.Number
Code can realize unique positioning of POI, but doorplate is generally smaller, be difficult to find under the complex environment of city,
The missing of a large amount of house numbers is there is also in electronic map database, it is impossible to meet the demand of POI lookups.
The content of the invention
It is an object of the invention to provide localization method and device based on shop signboard, energy in a kind of electronic map
Preferably solve that existing mobile terminal location precision is low, Wrong localization, and positioning failure problem.
According to an aspect of the present invention, there is provided electronic map in based on shop signboard localization method, including：
The store information in shop near its position of acquisition for mobile terminal；
By searching electronic map database, the shop point of interest letter matched with acquired store information is found
Breath；
The corresponding positional information of the shop interest point information is included in the electronic map.
Preferably, also include before the store information in shop near its position of acquisition for mobile terminal：
According to path connected relation, the shop point of interest in the electronic map database is ranked up, and
The sequential value of an order for identifying shop point of interest is distributed for each shop point of interest.
Preferably, also include before the store information in shop near its position of acquisition for mobile terminal：
Obtain the signboard image information of each shop point of interest；
Signboard image information to each shop point of interest carries out feature extraction, obtains each shop point of interest
Image feature information；
By the signboard image information and/or image feature information and the electronic map data of each shop point of interest
Shop point of interest in storehouse is associated.
Preferably, the store information include the keyword sequence that is formed by multiple shop titles of arranged in sequence or
The signboard image sequence that person is formed by multiple signboard images of arranged in sequence.
Preferably, it is described by searching electronic map database, find and match with acquired store information
The step of shop interest point information, includes：
In the electronic map database, search consistent with the shop name order in the keyword sequence
Shop interest point sequence, and according to the sequential value of each shop point of interest in the shop interest point sequence, calculate
The length of the shop interest point sequence for being found；
If the length of the shop interest point sequence is in the range of preseting length, it is determined that the shop interest for being found
Point sequence is matched with the keyword sequence.
Preferably, it is described by searching electronic map database, find and match with acquired store information
The step of shop interest point information, includes：
Feature extraction is carried out to each the signboard image in acquired signboard image sequence, each signboard figure is obtained
The image feature information of picture；
In the electronic map database, the image with each signboard image in the signboard image sequence is searched
The consistent shop interest point sequence of characteristic information, and according to each shop interest in the shop interest point sequence
The sequential value of point, calculates the length of found shop interest point sequence；
If the length of the shop interest point sequence is in the range of preseting length, it is determined that the shop interest for being found
Point sequence is matched with the signboard image sequence.
Preferably, find and match with acquired store information by searching electronic map database described
Shop interest point information the step of before, also include：
The position range information of its position of acquisition for mobile terminal, to be searched in acquired position range
The shop interest point information matched with acquired store information.
According to another aspect of the present invention, there is provided electronic map in based on shop signboard positioner, including：
Information input and processing module, the store information for obtaining shop near mobile terminal position；
Information searching and matching module, for by searching electronic map database, finding and acquired shop
The shop interest point information of information match；
Positioning result display module, for the corresponding positional information of the shop interest point information to be included described
In electronic map.
Preferably, also include：
Interest point data order module, for according to path connected relation, in the electronic map database
Shop point of interest is ranked up, and for each shop point of interest distributes a sequential value, so that described information is searched
Ranking results and sequential value are utilized with matching module, the shop interest for matching is found in electronic map database
Point information.
Preferably, also include：
Location information acquisition module, the position range information for obtaining mobile terminal position, so as to described
Information searching and matching module are searched in acquired position range and matched with acquired store information
Shop interest point information.
Compared with prior art, the beneficial effects of the present invention are：
Combination of the present invention according to shop signboard and signboards several adjacent to one another in electronic map system is carried out
Positive location, that is to say, that mobile phone camera is opened by user and is swept to the random ring of surrounding environment or artificial defeated
The mode for entering, receives the shop name information near user, and matching is believed with the shop title is searched in database
The location of breath information, realizes the positive location of user, and positioning precision is high, is that user brings preferably positioning
Experience.
Brief description of the drawings
Fig. 1 is the localization method theory diagram based on shop signboard in electronic map provided in an embodiment of the present invention；
Fig. 2 is the positioner block diagram based on shop signboard in electronic map provided in an embodiment of the present invention；
Fig. 3 is the positioning device structure figure based on shop signboard and combination provided in an embodiment of the present invention；
Fig. 4 is the pretreatment process figure of Fig. 3 shown devices；
Fig. 5 is the positioning flow figure of Fig. 3 shown devices.
Specific embodiment
Below in conjunction with accompanying drawing to a preferred embodiment of the present invention will be described in detail, it will be appreciated that described below
Preferred embodiment be merely to illustrate and explain the present invention, be not intended to limit the present invention.
Shop signboard：The signboard OR gate head in retail shop shop, comprising retail shop's trade mark or retail shop's title.
Fig. 1 is the localization method theory diagram based on shop signboard in electronic map provided in an embodiment of the present invention,
As shown in figure 1, including：
Step S101：The store information in shop near its position of acquisition for mobile terminal.
The store information includes the keyword sequence formed by multiple shop titles of arranged in sequence or by pressing
The signboard image sequence that multiple signboard images of sequence arrangement are formed.
, it is necessary to according to path connected relation before step S101 is performed, in the electronic map database
Shop point of interest be ranked up, and be used to identify the suitable of shop point of interest for each shop point of interest distributes one
The sequential value of sequence.Wherein, shop point of interest can include：Government organs, railway station subway station, bus station,
Bus stop, gas station of gas station, parking lot, high-speed service area, financial service, commercial mansion, retail firm
Industry, hotel etc. be various the information of the shop front, non-concerning security matters unit.
Further, before step S101 is performed, the signboard of each shop point of interest can also in advance be obtained
Image information, the signboard image information of each shop point of interest can include the signboard in retail shop shop, door head,
Trade mark etc..Then the signboard image information to each shop point of interest carries out feature extraction, obtains each shop
The image feature information of point of interest, finally by the signboard image information and/or characteristics of image of each shop point of interest
Information is associated with the shop point of interest in the electronic map database, so as to the trick associated by later use
Board image information and/or image feature information carry out information matches.
Step S102：By searching electronic map database, find and match with acquired store information
Shop interest point information.
According to the type of acquired store information, following two modes can be selected to search and acquired shop
The shop interest point information of information match：
1st, for being manually entered the mode of store information (such as shop title), in the electronic map data
In storehouse, the lookup shop interest point sequence consistent with the shop name order in the keyword sequence, and according to
The sequential value of each shop point of interest in the shop interest point sequence, calculates found shop interest point sequence
Length, if the length of the shop interest point sequence is in the range of preseting length, it is determined that the shop found
Interest point sequence is matched with the keyword sequence.
2nd, for the mode of input signboard image information (such as shop trade mark) of taking pictures, first to acquisition of taking pictures
Signboard image sequence in each signboard image carry out feature extraction, obtain the characteristics of image of each signboard image
Information, then in the electronic map database, searches and each signboard image in the signboard image sequence
The consistent shop interest point sequence of image feature information, and according to each shop in the shop interest point sequence
The sequential value of point of interest is spread, the length of found shop interest point sequence is calculated, if the shop point of interest sequence
The length of row is in the range of preseting length, it is determined that the shop interest point sequence for being found and the signboard image sequence
Row matching.
In the case where mobile terminal has traditional positioning function, mobile terminal can in advance obtain its position
Position range information, the search procedure of so described step S102 can enter in acquired position range
OK, so as to improve locating speed.
Step S103：The corresponding positional information of the shop interest point information is included in the electronic map.
Further, the display centre of electronic map is moved to the corresponding position of shop interest point information.
It should be noted that when step S102 find it is multiple emerging with acquired store information matches shop
Interesting point information, can re-start positioning, institute with return to step S101 by way of changing store information
State includes adjusting the order of shop title/signboard image, increases shop title/signboard figure to the modification of store information
Picture, deletion shop title/signboard image, change shop title/signboard image etc..
Fig. 2 is the positioner block diagram based on shop signboard in electronic map provided in an embodiment of the present invention, is such as schemed
Shown in 2, including：
Information input and processing module, the store information for obtaining shop near mobile terminal position, bag
Include the keyword sequence or the multiple signboard figures by arranged in sequence formed by multiple shop titles of arranged in sequence
As the signboard image sequence for being formed.
Information searching and matching module, for by searching electronic map database, finding and acquired shop
The shop interest point information of information match, specifically, according to the type of acquired store information, can be with
The different lookup mode of selection, such as when information input and processing module get keyword sequence, information is looked into
Look for and using shop title can carry out information searching with matching module and match, when information input is obtained with processing module
When getting signboard image sequence, information searching can utilize the image feature information of signboard image with matching module
Carry out information searching and matching.
Positioning result display module, for the corresponding positional information of the shop interest point information to be included described
In electronic map, it is possible to further be displayed in the display centre position of electronic map.
On the basis of above-mentioned functions module, described device is further included：
Interest point data order module, for according to path connected relation, in the electronic map database
Shop point of interest is ranked up, and for each shop point of interest distributes a sequential value, so that described information is searched
Ranking results and sequential value are utilized with matching module, the shop interest for matching is found in electronic map database
Point information.
Location information acquisition module, the position range information for obtaining mobile terminal position, so as to described
Information searching and matching module are searched in acquired position range and matched with acquired store information
Shop interest point information.
Fig. 3 is the positioning device structure figure based on shop signboard and combination provided in an embodiment of the present invention, such as Fig. 3
It is shown, including：
Initialization module, for emptying the text sequence in memory module, image sequence, characteristics of image and lookup
As a result.
Location information acquisition module, the coarse position information for obtaining current mobile terminal present position, specifically
The mode of acquisition includes：Gps system, telecommunications network cellular basestation, wifi signals, bluetooth, ultra wide band, inertia
One or more in the technologies such as navigation system of combination.
POI data order module, for being ranked up to the POI data in electronic map database, specific row
Sequential mode is, unified according to incremental (or successively decreasing) from left to right for the POI on every road in electronic map
Order assign sequential value.
First judge module, the mode for judging user input shop front information, specific input mode includes two kinds,
One kind is to be manually entered shop front title (or shop title), and another kind is to obtain shop front signboard image by taking pictures
Information.If it is determined that user needs to be manually entered as the shop front title of keyword, then call keyword sequence defeated
Enter module, if user needs acquisition shop front signboard image information of taking pictures, call photo sequence input module.
Keyword sequence input module, for provide Text Entry and move left and right, text insertion button, connect
Receive and show the shop front title of user input.Specifically, user can be input into the shop front title of the shop front to be positioned,
And continue to be input into according to the shop front order seen, button is inserted with text by moving left and right, input in order is faced
The shop front title of the nearly shop front, and by each keyword in the keyword sequence of input according to being incremented by from left to right
The order of (or successively decreasing) assigns sequential value.
String searching module, for searching the POI being consistent with the keyword sequence of input in electronic map database
Sequence, specifically, is searched in the range of the rough position that location information acquisition module is obtained first, if
The POI sequences being consistent are found, then calls the second judge module, if not finding the POI sequences being consistent, entered
One step expands hunting zone until whole urban area.Wherein, text sequence search procedure is specifically, successively time
The keyword sequence of input is gone through, and searches the POI being consistent with current shop title in the keyword sequence of input,
If finding the POI being consistent, the follow-up shop title in the keyword sequence of input is compared according to sequential value order
With the follow-up POI of the POI that is consistent for finding, comparison process can cross over, i.e., not yet terminate when at road ending
During comparison, then proceed to compare with the connected relation of road according to road.If in the keyword sequence of input
Each keyword found consistent POI titles, and be sequentially connected consistent POI for being found
Scope of the length of POI sequences obtained from title less than setting, then it is assumed that found the crucial word order with input
The POI sequences that row are consistent.
Signboard image sequence input module, for provide mobile terminal camera function and move left and right, image insert
Enter button, receive the shop signboard image information that user photographs.If signboard extraction module is from shop signboard figure
As in only extract a signboard information, then by move left and right with image insert button, directly by signboard information
It is added in photo sequence, if extracting multiple signboard information, according to the signboard order for positioning from left to right
With being incremented by (or successively decreasing) for each signboard information assigns sequential value, and button is inserted with image by moving left and right,
It is inserted into photo sequence.
Signboard extraction module, for extracting signboard data from the photo for shooting, specifically, passing through rectangular extraction
Algorithm, such as Hough transformation and Color Picking, extract to the signboard region in image.
Image characteristics extraction module, for extracting in signboard region and electronic map database in POI image datas
Characteristics of image, specific extracting method is to carry out dimensionality reduction and denoising to image using wavelet transformation, extracts its chi
Degree invariant features conversion (Scale-invariant feature transform, SIFT) point feature, while carrying out
Principal component analysis (Principal Component Analysis, PCA) dimensionality reduction.
Image sequence searching modul, for searching the signboard image sequence phase with input in electronic map database
The POI sequences of symbol, specifically, are carried out in the range of the rough position that location information acquisition module is obtained first
Search, if finding the POI sequences being consistent, the second judge module is called, if not finding what is be consistent
POI sequences, then further expand hunting zone until whole urban area.Wherein, image sequence search procedure
To travel through each the signboard image in the signboard image sequence of input successively, and search and current shop signboard
The characteristics of image of the POI signboards that characteristics of image is consistent, can select the matching based on point of interest and calculates during characteristic matching
Method, is matched using the nearest neighbor method based on Kd trees.If the POI signboards being consistent are found, according to sequence
The follow-up shop signboard image that train value order is compared in the image sequence of input is follow-up with the POI that is consistent for finding
The characteristics of image of POI signboards, comparison process can be crossed over, i.e., when not yet terminating to compare at road ending,
Then proceed to compare with the connected relation of road according to road.If each in the image sequence according to input
Width signboard image all have found consistent POI signboard images, and be sequentially connected the length of POI in sequence and be less than
The scope of setting, then it is assumed that found the POI sequences being consistent with the image sequence of input.
Memory module, it is special for storing keyword sequence input module, signboard image sequence input module, image
Levy text sequence, image sequence, the image in extraction module, String searching module and image sequence searching modul
Feature and lookup result.
Whether the second judge module, the lookup result for judging text sequence or image sequence is unique.If
Lookup result is unique, then positioning result display module is called, if there is multiple lookup results, then by prompting mould
Block is pointed out user.
Nformation alert module, for after the second judge module is judged to have multiple queries result, electronically
Multiple queries result is shown on figure, and the follow-up POI titles for having found POI sequences are shown in Text Entry
Pointed out as input.
3rd judge module, for judging whether user changes the shop front title being input into text box, if user
Continue the shop front title being input into text box, modification or deletion have been input into, then call keyword sequence to be input into
Module is processed；If user completes text input, positioning result display module is called.
4th judge module, for judging whether user changes the shop front signboard of shooting, if user continues to shoot
Or the shop front signboard that modification has shot, then call signboard extraction module to be processed；If user deletes defeated
The shop front signboard image for entering, then call image sequence searching modul to search again；If user completes photograph taking,
Then call positioning result display module.
Positioning result display module, it is specific aobvious for the positioning result that display system on the electronic map finds
The mode of showing is to highlight first shop front title of user input, or user with symbol on the electronic map
First position of shop front signboard for shooting, and the display centre of map is moved to the position.
Wherein, the String searching module and image sequence searching modul shown in Fig. 3 realize the information shown in Fig. 2
Search the function with matching module.Keyword sequence input module shown in Fig. 3, signboard image sequence input module,
Signboard extraction module, image characteristics extraction module, memory module realize the information input and treatment shown in Fig. 2
The function of module.
The workflow of Fig. 3 shown devices is broadly divided into the three below stage：
Pretreatment stage：The shop interest point data in electronic map database is ranked up first；Then will
The shop signboard view data associated with shop point of interest in electronic map database is pre-processed, and is extracted wherein
Can be used for the characteristic information of images match, if not gathering shop signboard image on the spot in electronic map database
Data, then extract in the trademark image registered or generally acknowledge from shop and can be used for the characteristic information of images match.Finally
The characteristic information of extraction is associated with shop point of interest, and is stored in electronic map database.
Input phase：The store information in shop near user is obtained by way of taking pictures or being manually entered, for example
The shop signboard image of shooting and/or the shop title being manually entered；If mobile terminal has location equipment, also
The rough position of mobile terminal can be obtained by location equipment.
The lookup stage：Judge the shop interest point information of the store information and electronic map database memory storage for obtaining
Whether (such as characteristic information of interest point name information and/or interest dot image) is consistent, and search procedure can lead to
The coarse localization information for crossing mobile terminal location equipment reduces the scope for searching point of interest.If find meeting condition
Point of interest, then return user nearby shop POI positional information；If do not found, it is required that user's weight
New input store information.
The above-mentioned store information information such as including signboard information, shop title.
The three above stage is further described below by way of Fig. 4 and Fig. 5.
Fig. 4 is the pretreatment process figure of Fig. 3 shown devices, as shown in figure 4, step includes：
Step S201：Pretreatment starts.
Step S202：POI data sorts.
POI data order module is called to enter shop interest point data stored in electronic map database
Row sequence.For the shop POI on every road, assigned according to the order for being incremented by (or successively decreasing) from left to right
Sequential value.
Step S203：Extract characteristics of image.
The shop signboard view data that will be associated with shop point of interest in electronic map database is pre-processed, and is carried
Take and wherein can be used for the characteristic information of images match, the characteristic information that then will be extracted is related to shop point of interest
Connection, and store in electronic map database.
If it should be noted that do not gather shop signboard view data on the spot in electronic map database,
Extracted in the trademark image registered or generally acknowledge from shop and can be used for the characteristic information of images match.
It should be noted that step S203 can also be performed before step S202.
Step S204：Pretreatment terminates.
Fig. 5 is the positioning flow figure of Fig. 3 shown devices, as shown in figure 5, step includes：
Step S301：Start positioning flow.
Step S302：Initialization.
Initialization module is called, text sequence in memory module, image sequence, characteristics of image and lookup is emptied
As a result, so as to start once new position enquiring.
Step S303：Obtain location information.
The coarse position information of current mobile terminal is obtained by location information acquisition module, acquisition modes include,
In the technologies such as GPS, telecommunications network cellular basestation, wifi signals, bluetooth, ultra wide band, inertial navigation system one
Plant or various combinations.
The step is optional step, for reducing the hunting zone positioned according to shop signboard, so as to improve
Search speed.
Step S304：The first judge module is called to judge whether user prepares to be manually entered store information (for example
Shop front title), if so, then performing step S305 and step S306 successively, otherwise, illustrate that user prepares logical
Acquisition store information (such as shop front signboard information) of taking pictures is crossed, step S307 to step S310 is now performed successively.
It is essentially all from the shop POI sequences in electronic map database to be manually entered mode and shooting style
Searched in data and input store information identical POI sequences.
In actual implementation process, can be only with one kind of both lookup modes, it would however also be possible to employ two kinds of lookups
The combination of mode.
Step S305：If it is determined that user needs to be manually entered store information, that is, it is input into as the door of keyword
Face title, then by provide text box inputting interface and move left and right, text insert button, receive and show
The shop front title of user input, keyword is referred to as by shop front name, is then stored.
User can be input into the shop front title for needing positioning, and it is follow-up to continue input according to the shop front order seen
Shop front title, during being manually entered, can insert button by moving left and right with text, be put in order by the shop front
Input closes on the shop front title of the shop front.Mobile terminal receives user and is put in order according to the shop front shop front title of input
After sequence, using shop front title sequence as keyword sequence, and to each keyword in keyword sequence according to
It is incremented by the sequential value that (or successively decreasing) assigns a shop front order for recording user input from left to right.For example
Mobile terminal receives the keyword sequence " KFC " with shop front sequence consensus on the spot, " the traffic silver of user input
OK ", " China Securities ", when so being searched in electronic map database and matched, it is possible to increase retrieval
Speed and accuracy rate.
Step S306：String searching is carried out, i.e., the key with user input is searched in electronic map database
The POI sequences that word sequence is consistent, then perform step S311.
Searched in the range of the rough position that location information acquisition module is obtained first, if finding what is be consistent
POI sequences, then call the second judge module；
If not finding the POI sequences being consistent, further expand hunting zone until whole urban area.
String searching process is to travel through keyword sequence successively, search be input into keyword sequence in it is current
The POI that shop title is consistent.If finding the POI being consistent, according to the key that the order of sequential value will be input into
Follow-up shop title in word sequence is compared with the follow-up POI of the POI that is consistent for finding, and comparison process can be with
Across, for example when not yet terminate at road ending compare when, according to road and the connected relation of road continue into
Row is compared.For example, the keyword sequence of input is " KFC ", " Bank of Communications ", " China Securities ", electricity
In POI sequences in sub- map data base, " KFC ", " blog article bookstore ", " Bank of Communications " these three points
On New Port East Road, " the rich good purchase of property ", " U.S. hair salon ", " China Securities " these three point on New Port West Road,
And New Port East Road is connected with New Port West Road, herein it will be understood that have found the POI sequences being consistent.
Wherein, traversal keyword and find the process of the POI being consistent and be, first will be in hunting zone it is all
POI is ranked up from first character to the last character, then searched in hunting zone in current shop
One word, if searched, continues search for second word, until all search is completed；If do not searched for
Arrive, then return to the POI for not finding and being consistent.
If each keyword in the keyword sequence of input has found consistent POI titles, and according to
The secondary length for connecting POI sequences obtained from the consistent POI titles for being found then is recognized less than the scope of setting
To have found the POI sequences being consistent with the keyword sequence of input.For example, the keyword sequence of input is " to agree
De Ji ", " Bank of Communications ", " China Securities ".POI sequences in map data base are " KFCs ", " win
Document shop ", " Bank of Communications ", " the rich good purchase of property ", " U.S. hair salon ", " China Securities ", due to the three of input
Individual keyword " KFC ", " Bank of Communications ", " China Securities " are not only included in map data base, and
Order in POI sequences is also consistent, and POI sequence lengths are less than the scope (such as 10) of setting, now really
Surely the POI sequences being consistent be have found.If again for example, the POI sequences in electronic map database are " Ken De
Base ", " China Securities ", " Bank of Communications ", or only " China Securities ", " Bank of Communications ", sequentially differ
Corresponding keyword is caused or lacked, is now defined as not finding the POI sequences being consistent.
Step S307：If it is determined that user needs to obtain store information by shooting style, by mobile terminal
Camera function and move left and right, image insertion button, receive and the shop signboard that photographs of memory mobile terminal
Image information.
If signboard extraction module only extracts a signboard information from the signboard image information of shop, by a left side
Move right and insert button with image, directly shop signboard image information is added in photo sequence；If from shop
Multiple signboard information are extracted in paving signboard image information, then according to the signboard order of positioning according to passing from left to right
Increase (or successively decreasing) and assign sequential value, and button is inserted with image by moving left and right, be inserted into photo sequence.
" moving to left " " moving to right " two buttons are for example designed, to the position of the shop signboard image in photo sequence
Move, be related to " left insertion " " right insertion " two buttons that shop signboard image is inserted in photo sequence.
Step S308：Signboard data are extracted from the photo (i.e. shop signboard image) for shooting.
Specifically, by rectangular extraction algorithm, such as Hough transformation and Color Picking, to shop signboard image
In signboard region positioned, i.e., the position of signboard is oriented from whole shop signboard image, obtain signboard
Data.
Step S309：Extract characteristics of image.
Extract the image in the POI image datas in the signboard data and electronic map database of captured image
Feature, specifically can carry out dimensionality reduction and denoising using wavelet transformation to image, extract its SIFT point feature, together
Shi Jinhang PCA dimensionality reductions.
Characteristics of image be the computer that goes out from signboard extracting data can unique identification go out view data or shop
The characteristic information of signboard image.
Step S310：Image is searched, and step S311 is gone to after lookup terminates.
Search the POI sequences being consistent with the signboard image sequence of input in electronic map database.First in positioning
Searched in the range of the rough position that data obtaining module is obtained, if finding the POI sequences being consistent, adjusted
Use the second judge module；If not finding the POI sequences being consistent, further expand hunting zone until whole
Individual urban area.Image sequence search procedure is to travel through the image sequence of input successively, and is searched and input
The POI signboard characteristics of image that current shop signboard characteristics of image is consistent in image sequence.Feature matching method is optional
With the matching algorithm based on point of interest, matched using the nearest neighbor method based on Kd trees.
If finding the POI being consistent, recruited according to the follow-up shop in the image sequence that sequential value order will be input into
Board image is compared with the characteristics of image of the follow-up POI signboards of the POI that is consistent for finding, and comparison process can be across
More, for example when not yet terminate at road ending compare when, then according to road and the connected relation of road continue into
Row is compared.
If each keyword in keyword sequence according to input has found consistent POI titles,
And the length of POI sequences obtained from the consistent POI titles for being found is sequentially connected less than the scope for setting,
Then think to have found the POI sequences being consistent with the keyword sequence of input.
If each width signboard image in image sequence according to input has all found consistent POI signboards
Image, and the length of POI sequences obtained from the consistent POI signboard images for being found is sequentially connected less than setting
Fixed scope, then it is assumed that found the POI sequences being consistent with the image sequence of input.
Step S311：Judge whether the lookup result of text sequence or image sequence is unique, if lookup result
Uniquely, then step S315 is performed, otherwise, illustrates, in the presence of multiple lookup results, now to perform step S312,
User is pointed out.
Step S312：Multiple queries result is shown on the electronic map, and is shown in Text Entry and looked for
Follow-up POI to POI sequences is referred to as being input into prompting, and the input of follow-up shop title is carried out for user,
Improve location efficiency.If do not pointed out, user can also complete positioning, but can experience lf being influenced.
The step is optional step, and user both can further be input into shop title or image completes to be accurately positioned,
Position fixing process can also be terminated, final result is directly chosen on map.
After step S312, step S313 or step S314 can be performed.
Step S313：Judge whether user changes the shop front title being input into text box, i.e. keyword, if
User continues the shop front title being input into text box, modification or deletion have been input into, then perform step S305,
Otherwise, illustrate that user has completed text input, now perform step S315.
Step S314：Judge whether user changes the shop front signboard of shooting, if what user's modification had been taken
Shop front signboard, then perform step S316, otherwise, illustrates that user need not change the shop front signboard figure being input into
Picture, i.e. user have completed photograph taking, now perform step S315.
Step S315：The positioning result that display system finds on the electronic map.
Display mode is, on the electronic map with symbol highlight user input first shop front title or
First position of shop front signboard that user shoots, and the display centre of electronic map is moved to the position.
Step S316：Judge whether to need to take pictures again, if so, then taken pictures and performed step S307,
Otherwise perform step S310.
Step S317：Terminate positioning.
Being manually entered shop front title and obtaining shop front signboard information by taking pictures in the present invention is two kinds arranged side by side
Shop front information input mode.Two kinds are searched format surface difference, and one kind is user's manual data shop front title, separately
A kind of is the photo array shop front signboard by system by being input into.Follow-up flow is all from electronic map database
In POI sequence datas in find and be input into shop front sequence identical POI sequences.Therefore of the invention specific
Implementation process in can also only with both lookup modes one of which.
In sum, the present invention has following technique effect：
1st, the present invention is used as positioning target by using shop signboard, improves positioning precision；
2nd, the present invention is in database by searching shop signboard image sequence, improves recognition efficiency and accurate
Property；
3rd, the present invention is input into store information by way of shooting shop signboard, improves the convenience of operation.
Although being described in detail to the present invention above, the invention is not restricted to this, the art technology
Personnel can carry out various modifications with principle of the invention.Therefore, all modifications made according to the principle of the invention,
All should be understood to fall into protection scope of the present invention.
Claims (10)
1. the localization method of shop signboard is based in electronic map, it is characterised in that including：
The store information in shop near its position of acquisition for mobile terminal；
By searching electronic map database, the shop point of interest letter matched with acquired store information is found
Breath；
The corresponding positional information of the shop interest point information is included in the electronic map.
2. method according to claim 1, it is characterised in that in its position of acquisition for mobile terminal
Before the store information in neighbouring shop, also include：
According to path connected relation, the shop point of interest in the electronic map database is ranked up, and
The sequential value of an order for identifying shop point of interest is distributed for each shop point of interest.
3. method according to claim 1 and 2, it is characterised in that in acquisition for mobile terminal where it
Position nearby before the store information in shop, also includes：
Obtain the signboard image information of each shop point of interest；
Signboard image information to each shop point of interest carries out feature extraction, obtains each shop point of interest
Image feature information；
By the signboard image information and/or image feature information and the electronic map data of each shop point of interest
Shop point of interest in storehouse is associated.
4. method according to claim 3, it is characterised in that the store information is included by sequentially arranging
The keyword sequence of multiple shop titles formation of row or the trick formed by multiple signboard images of arranged in sequence
Board image sequence.
5. method according to claim 4, it is characterised in that described by searching electronic map data
Storehouse, includes the step of find the shop interest point information matched with acquired store information：
In the electronic map database, search consistent with the shop name order in the keyword sequence
Shop interest point sequence, and according to the sequential value of each shop point of interest in the shop interest point sequence, calculate
The length of the shop interest point sequence for being found；
If the length of the shop interest point sequence is in the range of preseting length, it is determined that the shop interest for being found
Point sequence is matched with the keyword sequence.
6. method according to claim 4, it is characterised in that described by searching electronic map data
Storehouse, includes the step of find the shop interest point information matched with acquired store information：
Feature extraction is carried out to each the signboard image in acquired signboard image sequence, each signboard figure is obtained
The image feature information of picture；
In the electronic map database, the image with each signboard image in the signboard image sequence is searched
The consistent shop interest point sequence of characteristic information, and according to each shop interest in the shop interest point sequence
The sequential value of point, calculates the length of found shop interest point sequence；
If the length of the shop interest point sequence is in the range of preseting length, it is determined that the shop interest for being found
Point sequence is matched with the signboard image sequence.
7. the method according to claim 5 or 6, it is characterised in that described by searching electronically
Chart database, the step of find the shop interest point information matched with acquired store information before, also wrap
Include：
The position range information of its position of acquisition for mobile terminal, to be searched in acquired position range
The shop interest point information matched with acquired store information.
8. the positioner of shop signboard is based in electronic map, it is characterised in that including：
Information input and processing module, the store information for obtaining shop near mobile terminal position；
Information searching and matching module, for by searching electronic map database, finding and acquired shop
The shop interest point information of information match；
Positioning result display module, for the corresponding positional information of the shop interest point information to be included described
In electronic map.
9. device according to claim 8, it is characterised in that also include：
Interest point data order module, for according to path connected relation, in the electronic map database
Shop point of interest is ranked up, and for each shop point of interest distributes a sequential value, so that described information is searched
Ranking results and sequential value are utilized with matching module, the shop interest for matching is found in electronic map database
Point information.
10. device according to claim 8, it is characterised in that also include：
Location information acquisition module, the position range information for obtaining mobile terminal position, so as to described
Information searching and matching module are searched in acquired position range and matched with acquired store information
Shop interest point information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510934763.8A CN106885580A (en) | 2015-12-15 | 2015-12-15 | Localization method and device based on shop signboard in electronic map |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510934763.8A CN106885580A (en) | 2015-12-15 | 2015-12-15 | Localization method and device based on shop signboard in electronic map |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106885580A true CN106885580A (en) | 2017-06-23 |
Family
ID=59174165
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510934763.8A Pending CN106885580A (en) | 2015-12-15 | 2015-12-15 | Localization method and device based on shop signboard in electronic map |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106885580A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108022096A (en) * | 2017-12-27 | 2018-05-11 | 珠海市魅族科技有限公司 | A kind of method of mobile payment, device, terminal and readable storage medium storing program for executing |
| CN109582880A (en) * | 2018-12-04 | 2019-04-05 | 百度在线网络技术（北京）有限公司 | Interest point information processing method, device, terminal and storage medium |
| CN111340015A (en) * | 2020-02-25 | 2020-06-26 | 北京百度网讯科技有限公司 | Positioning method and device |
| CN112070833A (en) * | 2020-08-07 | 2020-12-11 | 北京三快在线科技有限公司 | Signboard map generation method and device and path planning method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101945327A (en) * | 2010-09-02 | 2011-01-12 | 郑茂 | Wireless positioning method and system based on digital image identification and retrieve |
| CN103674017A (en) * | 2013-12-20 | 2014-03-26 | 广东瑞图万方科技股份有限公司 | Indoor electronic map generation system, indoor navigation method and system |
| CN104613961A (en) * | 2015-02-12 | 2015-05-13 | 广东欧珀移动通信有限公司 | Shopping mall store navigation method and shopping mall store navigation device |
| CN104678421A (en) * | 2013-11-26 | 2015-06-03 | 财团法人资讯工业策进会 | Positioning control method |
| CN104977011A (en) * | 2014-04-09 | 2015-10-14 | 广东融讯信息科技有限公司 | Positioning method and positioning device based on street-photographing image in electronic map |
-
2015
- 2015-12-15 CN CN201510934763.8A patent/CN106885580A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101945327A (en) * | 2010-09-02 | 2011-01-12 | 郑茂 | Wireless positioning method and system based on digital image identification and retrieve |
| CN104678421A (en) * | 2013-11-26 | 2015-06-03 | 财团法人资讯工业策进会 | Positioning control method |
| CN103674017A (en) * | 2013-12-20 | 2014-03-26 | 广东瑞图万方科技股份有限公司 | Indoor electronic map generation system, indoor navigation method and system |
| CN104977011A (en) * | 2014-04-09 | 2015-10-14 | 广东融讯信息科技有限公司 | Positioning method and positioning device based on street-photographing image in electronic map |
| CN104613961A (en) * | 2015-02-12 | 2015-05-13 | 广东欧珀移动通信有限公司 | Shopping mall store navigation method and shopping mall store navigation device |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108022096A (en) * | 2017-12-27 | 2018-05-11 | 珠海市魅族科技有限公司 | A kind of method of mobile payment, device, terminal and readable storage medium storing program for executing |
| CN109582880A (en) * | 2018-12-04 | 2019-04-05 | 百度在线网络技术（北京）有限公司 | Interest point information processing method, device, terminal and storage medium |
| CN111340015A (en) * | 2020-02-25 | 2020-06-26 | 北京百度网讯科技有限公司 | Positioning method and device |
| CN111340015B (en) * | 2020-02-25 | 2023-10-20 | 北京百度网讯科技有限公司 | Positioning method and device |
| CN112070833A (en) * | 2020-08-07 | 2020-12-11 | 北京三快在线科技有限公司 | Signboard map generation method and device and path planning method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104936283B (en) | Indoor orientation method, server and system | |
| KR101289085B1 (en) | Images searching system based on object and method thereof | |
| CN106548164A (en) | The relevance recognition methods of facial image and mobile device | |
| CN103995889A (en) | Method and device for classifying pictures | |
| CN105517679B (en) | Determination of the geographic location of a user | |
| CN106255968A (en) | Natural language picture search | |
| US20100135527A1 (en) | Image recognition algorithm, method of identifying a target image using same, and method of selecting data for transmission to a portable electronic device | |
| CN106885580A (en) | Localization method and device based on shop signboard in electronic map | |
| CN104850563B (en) | Destination image compare retrieval device, destination image compares searching system and destination image compares search method | |
| CN110134829A (en) | Video locating method and device, storage medium and electronic device | |
| CN103745498B (en) | A kind of method for rapidly positioning based on image | |
| CN105824928A (en) | Mobile terminal, server, content-based image recognition searching method and system | |
| CN104977011A (en) | Positioning method and positioning device based on street-photographing image in electronic map | |
| JP6168355B2 (en) | Information processing system, information processing method, communication terminal, control method thereof, and control program | |
| CN103679674A (en) | Method and system for splicing images of unmanned aircrafts in real time | |
| CN103778261A (en) | Self-guided tour method based on mobile cloud computing image recognition | |
| WO2016199662A1 (en) | Image information processing system | |
| CN105279496A (en) | Human face recognition method and apparatus | |
| CN109038826A (en) | Substation equipment patrol system and inspection method based on Internet of Things and AR technology | |
| CN105159959A (en) | Image file processing method and system | |
| CN104520848A (en) | Searching for events by attendants | |
| CN105760844A (en) | Video stream data processing method, apparatus and system | |
| KR20190124436A (en) | Method for searching building based on image and apparatus for the same | |
| CN107679128A (en) | Information display method and device, electronic equipment and storage medium | |
| CN108509545A (en) | A kind of comment processing method and system of article |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170623 |