CN106815272A - A kind of image search method, apparatus and system - Google Patents
A kind of image search method, apparatus and system Download PDFInfo
- Publication number
- CN106815272A CN106815272A CN201510874622.1A CN201510874622A CN106815272A CN 106815272 A CN106815272 A CN 106815272A CN 201510874622 A CN201510874622 A CN 201510874622A CN 106815272 A CN106815272 A CN 106815272A
- Authority
- CN
- China
- Prior art keywords
- target
- retrieved
- image
- candidate target
- retrieval
- 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
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000005540 biological transmission Effects 0.000 claims description 6
- 238000007689 inspection Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 235000013399 edible fruits Nutrition 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000000638 solvent extraction Methods 0.000 description 2
- 230000018199 S phase Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- 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
- G06F16/53—Querying
- G06F16/532—Query formulation, e.g. graphical querying
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Processing Or Creating Images (AREA)
Abstract
A kind of image search method provided in an embodiment of the present invention, apparatus and system, are applied to client, and methods described includes：Image to be retrieved is sent to retrieval server；The information of the candidate target that retrieval server sends is received, wherein, candidate target is that the retrieval server determines the target that rule determines from image to be retrieved according to default candidate target；According to user for the target determine instruction to be retrieved of candidate target, target to be retrieved is determined；The image retrieval instruction for target to be retrieved is sent to retrieval server, to cause that retrieval server carries out image retrieval according to described image search instruction；Receive the image searching result for target to be retrieved of retrieval server feedback.When carrying out image retrieval using scheme provided in an embodiment of the present invention, each target to be retrieved that can be directed in image to be retrieved is retrieved, and reduces influence of the information in the background of image to be retrieved to retrieval result, improves the accuracy rate of image retrieval.
Description
Technical field
The present invention relates to image retrieval technologies field, more particularly to a kind of image search method, device and it is
System.
Background technology
As interconnection picture on network information is increasingly huge, the image that user is desired based on internet carries out image
The demand of retrieval is more and more stronger, therefore, the various image indexing systems based on internet arise at the historic moment.Figure
After occurring as searching system, the image that user is based on internet carries out image retrieval becomes very simple, example
Such as, user can send retrieval image or by client to retrieval service by client to retrieval server
Device sends the information such as URL (Uniform Resource Locator, URL) address of image,
After retrieval server receives these information, according to the image for being received or the corresponding images of above-mentioned URL
Image retrieval is carried out, commodity corresponding with above-mentioned image or similar, and the merchandise news that will be obtained is obtained
Client is fed back to, enables a user to be checked in client the merchandise news of correlation.
In the prior art, there is provided such a image search method：First, client receives user and submits to
Image to be retrieved, and by user submit to image to be retrieved send to retrieval server；Then, the retrieval
Server, according to default image retrieval rule, enters after the image to be retrieved is received to image to be retrieved
Row image retrieval, and retrieval result is fed back into client.
In practical application, user's majority of case only wants to retrieve and certain a part of area in image to be retrieved
The same or analogous picture of content in domain.However, often include various picture materials in image to be retrieved,
Especially the information in the background of image to be retrieved influences whether retrieval result, causes image retrieval accuracy rate
It is low.
The content of the invention
The purpose of the embodiment of the present invention is to provide a kind of image search method, apparatus and system, to improve figure
As the accuracy rate of retrieval.
To reach above-mentioned purpose, the embodiment of the invention discloses a kind of image search method, client is applied to,
Methods described includes：
Image to be retrieved is sent to retrieval server；
The information of the candidate target that the retrieval server sends is received, wherein, the candidate target is described
Retrieval server determines the target that rule determines from the image to be retrieved according to default candidate target；
According to user for the target determine instruction to be retrieved of the candidate target, target to be retrieved is determined；
The image retrieval instruction for the target to be retrieved is sent to the retrieval server, it is described to cause
Retrieval server carries out image retrieval according to described image search instruction；
Receive the image searching result for the target to be retrieved of the retrieval server feedback.
Preferably, being included in the information of the candidate target：The unique mark of the candidate target；
The image retrieval instruction sent to the retrieval server for the target to be retrieved, including：
The image retrieval for sending the unique mark for carrying the target to be retrieved to the retrieval server refers to
Order.
Preferably, the target determine instruction to be retrieved that the candidate target is directed to according to user, it is determined that treating
Searched targets, including：
Based on the information of the candidate target, the candidate target is showed into user；
The target determine instruction to be retrieved of the candidate target is directed to according to user, from the candidate target really
Fixed target to be retrieved.
Preferably, the target determine instruction to be retrieved that the candidate target is directed to according to user, it is determined that treating
Searched targets, including：
Based on the information of the candidate target, the candidate target is showed into user；
Whether comprising the desired candidate target of user in the shown candidate target of judgement；
If it is, determining that the desired candidate target of user is target to be retrieved；Otherwise, according to user described
The selection region of the image described to be retrieved of hand drawn determines target to be retrieved in client.
Preferably, the target determine instruction to be retrieved that the candidate target is directed to according to user, it is determined that treating
Searched targets, including：
Whether the information for judging the candidate target is sky information；
If it is, according to the selection region of user's image described to be retrieved of hand drawn in the client
Determine target to be retrieved.
To reach above-mentioned purpose, the embodiment of the invention also discloses a kind of image search method, retrieval is applied to
Server, methods described includes：
Receive the image to be retrieved that client sends；
Rule is determined according to default candidate target, candidate target is determined from the image to be retrieved；
The information of identified candidate target is sent to the client；
Receive the image retrieval instruction for the target to be retrieved that the client sends；Wherein, it is described
Target to be retrieved is the target determine instruction to be retrieved that the client is directed to the candidate target according to user
The target of determination；
Image retrieval is carried out according to described image search instruction, the image obtained for the target to be retrieved is examined
Hitch is really；
Described image retrieval result is sent to the client.
Preferably, the information by identified candidate target is sent to the client, including：
Obtain the retrieval frequency values of the candidate target；
According to the height of the retrieval frequency values, the unique mark of the candidate target is generated；
The information that the unique mark of the candidate target will be carried is sent to the client.
To reach above-mentioned purpose, the embodiment of the invention discloses a kind of image retrieving apparatus, client is applied to,
Described device includes：
Retrieval image sending module, for sending image to be retrieved to retrieval server；
Information receiving module, the information for receiving the candidate target that the retrieval server sends, wherein,
The candidate target is that the retrieval server determines rule from described to be retrieved according to default candidate target
The target determined in image；
Searched targets determining module, for determining to refer to for the target to be retrieved of the candidate target according to user
Order, determines target to be retrieved；
Search instruction sending module, for sending the figure for the target to be retrieved to the retrieval server
As search instruction, to cause that the retrieval server carries out image retrieval according to described image search instruction；
Retrieval result receiver module, for receive retrieval server feedback for the target to be retrieved
Image searching result.
Preferably, the search instruction sending module, described for being included in the information of the candidate target
During the unique mark of candidate target, sent to the retrieval server and carry the unique of the target to be retrieved
The image retrieval instruction of mark.
Preferably, the searched targets determining module, including：Candidate target shows submodule and searched targets
Determination sub-module；Wherein,
The candidate target shows submodule, for the information based on the candidate target, by candidate's mesh
Mark shows user；
The searched targets determination sub-module, the target to be retrieved for being directed to the candidate target according to user
Determine instruction, determines target to be retrieved from the candidate target.
Preferably, the searched targets determining module, including：Candidate target displaying submodule, the first candidate
Object judgement submodule, the first searched targets determination sub-module and the second searched targets determination sub-module；Wherein,
The candidate target shows submodule, for the information based on the candidate target, by candidate's mesh
Mark shows user；
The first candidate target judging submodule, for judging whether include use in shown candidate target
The desired candidate target in family；
The first searched targets determination sub-module, for sentencing in the first candidate target judging submodule
Disconnected result is target to be retrieved when being, to determine the desired candidate target of user；
The second searched targets determination sub-module, for sentencing in the first candidate target judging submodule
When disconnected result is no, according to the selection area of user's image described to be retrieved of hand drawn in the client
Domain determines target to be retrieved.
Preferably, the searched targets determining module, including：Second candidate target judging submodule,
Three searched targets determination sub-modules；Wherein,
The second candidate target judging submodule, whether the information for judging the candidate target is empty letter
Breath；
The 3rd searched targets determination sub-module, for sentencing in the second candidate target judging submodule
Disconnected result is when being, according to the selection area of user's image described to be retrieved of hand drawn in the client
Domain determines target to be retrieved.
To reach above-mentioned purpose, the embodiment of the invention also discloses a kind of image retrieving apparatus, retrieval is applied to
Server, described device includes：
Retrieval image receiver module, the image to be retrieved for receiving client transmission；
Candidate target determining module, for determining rule according to default candidate target, from the figure to be retrieved
Determine candidate target as in；
Information sending module, for the information of identified candidate target to be sent to the client；
Search instruction receiver module, for receiving the figure for the target to be retrieved that the client sends
As search instruction；Wherein, the target to be retrieved is that the client is directed to the candidate target according to user
Target determine instruction to be retrieved determine target；
Retrieval result obtains module, for carrying out image retrieval according to described image search instruction, is directed to
The image searching result of the target to be retrieved；
Retrieval result sending module, for described image retrieval result to be sent to the client.
Preferably, described information sending module, including：Retrieval frequency obtains submodule, unique mark generation
Submodule and information sending submodule；Wherein,
The retrieval frequency obtains submodule, the retrieval frequency values for obtaining the candidate target；
The unique mark generates submodule, and for the height according to the retrieval frequency values, generation is described
The unique mark of candidate target；
Described information sending submodule, sends for will carry the information of unique mark of the candidate target
To the client.
To reach above-mentioned purpose, the embodiment of the invention discloses a kind of image indexing system, the system includes：
Using the client and the retrieval server of application said apparatus of said apparatus.
As seen from the above, in technical scheme provided in an embodiment of the present invention, the client in image indexing system
End receives the image to be retrieved that user submits to first, and the image is sent into the retrieval in image indexing system
Server；Then, image retrieval server determines rule according to default candidate target, from image to be retrieved
Middle determination candidate target, and the information of identified candidate target is sent to client；Then, client
Information according to the candidate target for receiving determines target to be retrieved, and is sent for be checked to retrieval server
Rope mesh target image retrieval is instructed；Then, retrieval server carries out image retrieval according to the instruction, obtains pin
The image searching result of searched targets is treated, and image searching result is sent to client.Using the present invention
When the scheme that embodiment is provided carries out image retrieval, each target to be retrieved in image to be retrieved can be directed to
Retrieved, reduced influence of the information in the background of image to be retrieved to retrieval result, improve image
The accuracy rate of retrieval.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to implementing
Example or the accompanying drawing to be used needed for description of the prior art are briefly described, it should be apparent that, describe below
In accompanying drawing be only some embodiments of the present invention, for those of ordinary skill in the art, do not paying
On the premise of going out creative work, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 a are that a kind of flow of image search method for being applied to client provided in an embodiment of the present invention is shown
It is intended to；
Fig. 1 b are the flow of the image search method that another kind provided in an embodiment of the present invention is applied to client
Schematic diagram；
Fig. 1 c are the flow of the image search method that another kind provided in an embodiment of the present invention is applied to client
Schematic diagram；
Fig. 1 d are the flow of the image search method that another kind provided in an embodiment of the present invention is applied to client
Schematic diagram；
Fig. 2 is a kind of flow of image search method for being applied to retrieval server provided in an embodiment of the present invention
Schematic diagram；
Fig. 3 is a kind of structural representation of image retrieving apparatus for being applied to client provided in an embodiment of the present invention
Figure；
Fig. 4 is a kind of structure of the image retrieving apparatus for being applied to retrieval server provided in an embodiment of the present invention
Schematic diagram；
Fig. 5 is a kind of structural representation of image indexing system provided in an embodiment of the present invention；
Fig. 6 is a kind of signaling process schematic diagram of image indexing system provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clearly
Chu, it is fully described by, it is clear that described embodiment is only a part of embodiment of the invention, rather than
Whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art are not making creation
Property work under the premise of the every other embodiment that is obtained, belong to the scope of protection of the invention.
In existing image retrieval technologies, often client directly transmits the image to be retrieved that user submits to
Image retrieval is carried out to retrieval server, so there is drawback.On the one hand, in an image to be retrieved
Often there are multiple targets simultaneously；On the other hand, the background in image to be retrieved is also often extremely complex
's.However, in practical application, user's majority of case only wants to a certain in retrieval and image to be retrieved
The same or analogous picture of content in subregion.It is readily appreciated that, image retrieval server is treated for this
When inspection image carries out image retrieval, also user in the image can not expected that the content of retrieval also one is acted as
For searched targets are retrieved, it is incomplete to be so easily caused retrieval result, causes image retrieval accuracy rate low.
For case above, a kind of image search method, apparatus and system are the embodiment of the invention provides, under
Face first carries out lower detailed description to image search method provided in an embodiment of the present invention.
Fig. 1 a are that a kind of flow of image search method for being applied to client provided in an embodiment of the present invention is shown
It is intended to, the method may comprise steps of：
Step S101：Image to be retrieved is sent to retrieval server.
" image to be retrieved " mentioned here is the image that user is submitted in client by various modes,
At least one target is usually contained in the prospect of the image, wherein, this " target " may be employed to conduct
The picture material of image retrieval condition.For example, by including that the monitoring camera of traffic block port shoots
In the image of several vehicles, each vehicle can be referred to as a target, the car door of each vehicle
A target can be referred to as, this needs to determine according to specific image retrieval demand in practical application.
Step S102：Receive the information of the candidate target that retrieval server sends.
Wherein, candidate target is that retrieval server determines rule from image to be retrieved according to default candidate target
The target of middle determination.
Specifically, default candidate target determines that rule can be according to the foreground target of image and dividing for background
Rule determined by algorithm is cut, for example, the difference based on foreground target Yu the gray value of background, by setting
Suitable gray threshold is separated from image to be retrieved by foreground target, determines candidate target.
It should be noted that the candidate target based on foreground target and the gray value of background described herein determines
Mode is only citing, and the present invention is not limited the foreground target of image with the partitioning algorithm of background, any
Possible implementation can apply to the present invention, and those skilled in the art are needed according to practical application
In concrete condition reasonably set.
Here can include candidate target in image to be retrieved in mentioned " information of candidate target "
Positional information (coordinate of such as candidate target in image to be retrieved, etc.), further, the information
In can also include candidate target unique mark, for example, obtaining each candidate target according to retrieval server
When the unique mark being sequentially generated, etc..It should be noted that the embodiment of the present invention and need not to wait
The information of target is selected to be defined, those skilled in the art need the concrete condition in practical application
Reasonably set；The present invention is carried out also without the concrete mode of the unique mark to generating candidate target
Limit, any possible implementation can apply to the present invention.
Step S103：According to user for the target determine instruction to be retrieved of the candidate target, determine to be retrieved
Target.
Client can determine target to be retrieved by way of user selects in candidate target.Citing and
Speech, in the case of including positional information of the candidate target in image to be retrieved in the information of candidate target,
User can determine target to be retrieved according to positional information from candidate target, and client receives the choosing of user
After selecting operational order, it is determined that selected candidate target is target to be retrieved.It is, of course, also possible to there is other roots
The mode of target to be retrieved is determined according to positional information, for example, user can locally preset one in client
Regional extent, then whether the information of the received candidate target of client judgement is positioned at the default region model
In enclosing, if it is judged that it is to confirm as target to be retrieved to be.It should be noted that being only here
For example, those skilled in the art need the concrete condition in practical application reasonably to be set
Put.
Step S104：The image retrieval instruction for target to be retrieved is sent to retrieval server, to cause inspection
Rope server carries out image retrieval according to image retrieval instruction.
In a kind of concrete implementation mode of the embodiment of the present invention, when including time in the information of candidate target
When selecting positional information of the target in image to be retrieved, client can send carrying and need to retrieval server
The image retrieval instruction of the positional information of searched targets.
In another concrete implementation mode of the embodiment of the present invention, include when in the information of candidate target
During the unique mark of candidate target, client can send to retrieval server and carry target to be retrieved only
The image retrieval instruction of one mark.
It should be noted that both the above mode is only citing, certainly, the embodiment of the present invention is to retrieval
Can also there are other modes when sending the image retrieval instruction for target to be retrieved in server, for example,
The image that the positional information and unique mark that carry target to be retrieved can be sent to retrieval server is examined
Suo Zhiling, etc., those skilled in the art need the concrete condition in practical application to be configured.
Step S105：Receive the image searching result for target to be retrieved of retrieval server feedback.
As seen from the above, when carrying out image retrieval using scheme provided in an embodiment of the present invention, can be directed to and treat
Each target to be retrieved in retrieval image is retrieved, and reduces the information pair in the background of image to be retrieved
The influence of retrieval result, improves the accuracy rate of image retrieval.
Referring to Fig. 1 b, on the mode embodiment basis shown in Fig. 1 a, the embodiment of the invention provides another
Plant the schematic flow sheet of the image search method for being applied to client.
Wherein, step S103 may comprise steps of：
Step S1031：Information based on candidate target, user is showed by candidate target.
Specifically, can be according to the positional information of candidate target, such as candidate target institute in image to be retrieved
In the area coordinate in region, determined determined by retrieval server from the local image to be retrieved of client
Each candidate target, and each candidate target is showed into user.
Step S1032：The target determine instruction to be retrieved of candidate target is directed to according to user, from candidate target
It is middle to determine target to be retrieved.
Specifically, show each candidate target of user to be typically what can be clicked on, when user uses mouse etc.
When cursor position is moved into a certain candidate target for being shown, the candidate target can be highlighted, and user is only
The candidate target highlighted with mouse click is needed to may be selected by out the desired candidate target of user, client's termination
To be retrieved target determine instruction of the user for the desired candidate target is received, so as to from each candidate target
Determine target to be retrieved.
As seen from the above, when carrying out image retrieval using scheme provided in an embodiment of the present invention, can be directed to and treat
Each target to be retrieved in retrieval image is retrieved, and reduces the information pair in the background of image to be retrieved
The influence of retrieval result, improves the accuracy rate of image retrieval.In addition, the technical scheme that the present embodiment is provided
In, select oneself desired candidate's mesh in each candidate target that user can determine from retrieval server
Mark, so can more accurately retrieve the image searching result for meeting user's requirement, further increase inspection
The accuracy rate of hitch fruit.
Referring to Fig. 1 c, on the embodiment of the method basis shown in Fig. 1 b, the embodiment of the invention provides another
Plant the schematic flow sheet of the image search method for being applied to client.
Wherein, step S103 may comprise steps of：
Step S1031：Information based on candidate target, user is showed by candidate target.
Step S1033：Judge in shown candidate target whether comprising the desired candidate target of user, such as
Fruit is to perform step S1034, otherwise performs step S1035.
After to user's displaying candidate target, in the candidate target that can be shown according to the operation judges of user
Whether the desired candidate target of user is included.For example, can by judge user whether by mouse or
Whether person's touch-screen chooses the mode of one or more candidate targets, judges included in shown candidate target
The desired candidate target of user, specifically, when client receives user for shown candidate target
After target determine instruction to be retrieved, comprising the desired candidate target of user in the shown candidate target of judgement；
Furthermore it is also possible to by way of judging whether that receiving user clicks on " manually selecting " button, judge institute
Whether the desired candidate target of user is included in the candidate target of displaying, specifically, when client receives use
After " manually selecting " button is clicked at family, it is possible to determine that desired not comprising user in the candidate target for being shown
Candidate target.
It should be noted that the application is to be illustrated as a example by above-mentioned, judge to be shown in practical application
Candidate target in whether be not limited to that the application is not comprising the mode of the desired candidate target of user
This is defined.
Step S1034：Determine that the desired candidate target of user is target to be retrieved.
Step S1035：According to user in the client the image to be retrieved of hand drawn selection region determine
Target to be retrieved.
As seen from the above, compared with the embodiment of the method shown in Fig. 1 b, as the shown candidate of judgement
Not comprising in the case of the desired candidate target of user in target, the present embodiment additionally provides one kind and is existed by user
Hand drawn selection region determines mesh calibration method to be retrieved in client, this way it is ensured that retrieval service
The retrieval result of device is desired with user consistent as far as possible, improves the accuracy rate of retrieval result.
Referring to Fig. 1 d, on the basis of mode embodiment shown in Fig. 1 a, another kind is the embodiment of the invention provides
It is applied to the schematic flow sheet of the image search method of client.
Wherein, step S103 may comprise steps of：
Step S106：Whether the information for judging candidate target is sky information.
Specifically, retrieval server is after the image to be retrieved for receiving client transmission, according to default time
The image processing algorithm included in Object selection rule is selected to process image to be retrieved, it is readily appreciated that,
For in the processing result image of image to be retrieved whether comprising candidate target and comprising several candidate targets,
This is to rely on the specific algorithm in default candidate target selection rule, therefore, in default candidate's mesh
When mark selection rule sets not fogging Chu improper, to be retrieved, it is more likely that appearance cannot be detected
The situation of candidate target, that is to say, that the information of candidate target is empty information.
Step S107：According to user, the selection region of the image to be retrieved of hand drawn determines to treat in the client
Searched targets.
As seen from the above, compared with the embodiment of the method shown in Fig. 1 a, Fig. 1 b and Fig. 1 c, work as judgement
Under whether the information of candidate target is sky information state, the present embodiment additionally provides one kind by user in client
Middle hand drawn selection region determines mesh calibration method to be retrieved, this way it is ensured that for each to be retrieved
Image obtains corresponding retrieval result as far as possible, and the retrieval result as far as possible with desired by user
It is consistent, improve the accuracy rate of retrieval result.
Fig. 2 is the schematic flow sheet of another image search method provided in an embodiment of the present invention, is applied to inspection
Rope server, the method may comprise steps of：
Step S201：Receive the image to be retrieved that client sends.
Step S202：Rule is determined according to default candidate target, candidate target is determined from image to be retrieved.
Specifically, default candidate target determines that rule can be according to the foreground target of image and dividing for background
Rule determined by algorithm is cut, for example, the difference based on foreground target Yu the gray value of background, by setting
Suitable gray threshold is separated from image to be retrieved by foreground target, determines candidate target.
It should be noted that the candidate target based on foreground target and the gray value of background described herein determines
Mode is only citing, and the present invention is not limited the foreground target of image with the partitioning algorithm of background, any
Possible implementation can apply to the present invention, and those skilled in the art are needed according to practical application
In concrete condition reasonably set.
Step S203：The information of identified candidate target is sent to client.
In a kind of specific implementation of the embodiment of the present invention, the information of identified candidate target is sent
To client, may comprise steps of：
(1) the retrieval frequency values of candidate target are obtained.
In actual applications, it is directed in the retrieval of keyword, each website often can be by counting certain
Individual retrieval frequency keyword preferential recommendation higher facilitates user search to user.Likewise, for image
For retrieval, it is also possible to which the retrieval frequency to searched targets is counted, and the retrieval higher by frequency is retrieved
Target priority recommends user, to facilitate user search, so, when the mark of candidate target is generated, can
Generated with according to retrieval frequency.
Specifically, retrieval server can retrieve note to the history of the candidate target of each in image retrieval procedure
Record is counted, so as to obtain the retrieval frequency values of candidate target.It should be noted that the present invention is not right
The specific acquisition pattern for retrieving frequency values is defined.
(2) according to the height of retrieval frequency values, the unique mark of candidate target is generated.
It should be noted that candidate target can be generated according to retrieval frequency values order from high to low
Unique mark, can also generate unique mark of candidate target according to retrieval frequency values order from low to high
Know, the present invention need not be defined to the specific generating mode of the unique mark of candidate target.
(3) information that will carry the unique mark of candidate target is sent to client.
It should be noted that the retrieval frequency values generation candidate target according to candidate target enumerated here
The mode of unique mark is only the one kind in numerous achievable modes, and the embodiment of the present invention is not to unique mark
Generating mode be defined, any possible implementation can apply to the present invention, in the art
Technical staff needs the concrete condition in practical application reasonably to be set.
Step S204：Receive the image retrieval instruction for target to be retrieved that client sends.
Wherein, target to be retrieved is the target determine instruction to be retrieved that client is directed to candidate target according to user
The target of determination.
Step S205：Image retrieval is carried out according to image retrieval instruction, the image for target to be retrieved is obtained
Retrieval result.
Step S206：Image searching result is sent to client.
As seen from the above, when carrying out image retrieval using scheme provided in an embodiment of the present invention, can be directed to and treat
Each target to be retrieved in retrieval image is retrieved, and reduces the information pair in the background of image to be retrieved
The influence of retrieval result, improves the accuracy rate of image retrieval.
Lower detailed description is carried out to image retrieving apparatus provided in an embodiment of the present invention again below.
Fig. 3 is a kind of structural representation of image retrieving apparatus provided in an embodiment of the present invention, is applied to client
End, the device can include with lower module：
Retrieval image sending module 310, for sending image to be retrieved to retrieval server.
Information receiving module 320, the information of the candidate target for receiving retrieval server transmission.
Wherein, candidate target is that retrieval server determines rule from image to be retrieved according to default candidate target
The target of middle determination.
Searched targets determining module 330, for being determined for the target to be retrieved of the candidate target according to user
Instruction, determines target to be retrieved.
Search instruction sending module 340, for sending the image inspection for target to be retrieved to retrieval server
Suo Zhiling, to cause that retrieval server carries out image retrieval according to image retrieval instruction.
Retrieval result receiver module 350, the figure for target to be retrieved for receiving retrieval server feedback
As retrieval result.
Specifically, searched targets determining module 330, for including candidate target in the information of candidate target
During positional information in image to be retrieved, according to positional information, target to be retrieved is determined from candidate target；
Specifically, search instruction sending module 340, for including candidate target in the information of candidate target
During positional information in image to be retrieved, sent to retrieval server and carry the position of target to be retrieved and believe
The image retrieval instruction of breath.
Specifically, search instruction sending module 340, for including candidate target in the information of candidate target
Unique mark when, the image retrieval of unique mark for carrying target to be retrieved is sent to retrieval server and is referred to
Order.
Specifically, searched targets determining module 330, can include：Candidate target shows submodule and retrieval
Target determination sub-module；Wherein,
Candidate target shows submodule, for the information based on candidate target, candidate target is showed into user；
Searched targets determination sub-module, for determining to refer to for the target to be retrieved of candidate target according to user
Order, determines target to be retrieved from candidate target.
Specifically, searched targets determining module 330, including：Candidate target displaying submodule, the first candidate
Object judgement submodule, the first searched targets determination sub-module and the second searched targets determination sub-module；Wherein,
Candidate target shows submodule, for the information based on candidate target, candidate target is showed into user；
First candidate target judging submodule, for whether judging in shown candidate target comprising user's phase
The candidate target of prestige；
First searched targets determination sub-module, for being yes in the judged result of candidate target judging submodule
When, determine that the desired candidate target of user is target to be retrieved；
Second searched targets determination sub-module, for being no in the judged result of candidate target judging submodule
When, the selection region according to user's image to be retrieved of hand drawn in the client determines target to be retrieved.
Specifically, searched targets determining module 330, can include：Second candidate target judging submodule,
3rd searched targets determination sub-module；Wherein,
Second candidate target judging submodule, whether the information for judging candidate target is sky information；
3rd searched targets determination sub-module, for being in the judged result of the second candidate target judging submodule
When being, the selection region according to user's image to be retrieved of hand drawn in the client determines mesh to be retrieved
Mark.
As seen from the above, when carrying out image retrieval using scheme provided in an embodiment of the present invention, can be directed to and treat
Each target to be retrieved in retrieval image is retrieved, and reduces the information pair in the background of image to be retrieved
The influence of retrieval result, improves the accuracy rate of image retrieval.
Fig. 4 is the structural representation of another image search method provided in an embodiment of the present invention, is applied to inspection
Rope server, the device can include with lower module：
Retrieval image receiver module 410, the image to be retrieved for receiving client transmission.
Candidate target determining module 420, for determining rule according to default candidate target, from figure to be retrieved
Determine candidate target as in.
Information sending module 430, for the information of identified candidate target to be sent to client.
Search instruction receiver module 440, the image inspection for target to be retrieved for receiving client transmission
Suo Zhiling.
Wherein, target to be retrieved is that client determines to refer to according to user for the target to be retrieved of the candidate target
Make the target for determining；
Retrieval result obtains module 450, for carrying out image retrieval according to image retrieval instruction, is directed to
The image searching result of target to be retrieved.
Retrieval result sending module 460, for image searching result to be sent to client.
Specifically, information sending module 430, can include：Retrieval frequency obtains submodule, unique mark
Generation submodule and information sending submodule；Wherein,
Retrieval frequency obtains submodule, the retrieval frequency values for obtaining candidate target；
Unique mark generates submodule, and for the height according to retrieval frequency values, generation candidate target is only
One mark；
Information sending submodule, sends to client for will carry the information of unique mark of candidate target
End.
As seen from the above, when carrying out image retrieval using scheme provided in an embodiment of the present invention, can be directed to and treat
Each target to be retrieved in retrieval image is retrieved, and reduces the information pair in the background of image to be retrieved
The influence of retrieval result, improves the accuracy rate of image retrieval.
Referring to Fig. 5, Fig. 5 is a kind of structural representation of image indexing system provided in an embodiment of the present invention, should
System can include：Using the client of the image retrieving apparatus of previous embodiment offer and using foregoing implementation
The retrieval server of the image retrieving apparatus that example is provided.
First, lower detailed description is carried out to the detailed process of image retrieval with reference to Fig. 6.
Client 510 sends image to be retrieved (S601) to retrieval server 520, and retrieval server 520 is received
To after image to be retrieved, rule is determined according to default candidate target, candidate's mesh is determined from image to be retrieved
Mark (S602), retrieval server 520 after candidate target is determined, by the information of identified candidate target
Send to client 510 (S603), client 510 receives the letter of the candidate target that retrieval server 520 sends
Breath, and target to be retrieved is determined for the target determine instruction to be retrieved of the candidate target according to user
(S604), then, client 510 sends the image retrieval for target to be retrieved to retrieval server 520
Instruction (S605)；After retrieval server 520 receives image retrieval instruction, instructed according to the image retrieval
Carry out image retrieval, obtain the image searching result (S606) for target to be retrieved, and by image retrieval
Result is sent to client 510 (S607).
In a specific embodiment of the present invention, can be included in the information of candidate target：Candidate target
Positional information in image to be retrieved, user can determine from candidate target to be retrieved according to positional information
Target, after client 510 receives the selection operation instruction of user, it is determined that selected candidate target is to treat
Searched targets, the image retrieval for sending the positional information for carrying target to be retrieved to retrieval server 520 refers to
Order.
In another specific implementation of the invention, included in the information of candidate target：Candidate target
Unique mark；The image retrieval for sending the unique mark for carrying target to be retrieved to retrieval server 520 refers to
Order.
In a kind of specific embodiment of the invention, client 510 is based on the information of candidate target, by candidate
Intended display is to user；The target determine instruction to be retrieved of candidate target is directed to according to user, from candidate target
It is middle to determine target to be retrieved.
In another specific embodiment of the invention, it is in the shown candidate target of the judgement of client 510
It is no comprising the desired candidate target of user；If it is, determining that the desired candidate target of user is target to be retrieved；
Otherwise, according to user in client 510 selection region of the image to be retrieved of hand drawn determine it is to be retrieved
Target.
In another specific embodiment of the invention, client 510 judge candidate target information whether be
Empty information；If it is, according to the selection region of user's image to be retrieved of hand drawn in client 510
Determine target to be retrieved.
In another specific embodiment of the invention, retrieval server 520 obtains the retrieval of candidate target frequently
Rate score；According to the height of retrieval frequency values, the unique mark of candidate target is generated；Candidate will be carried
The information of the unique mark of target is sent to client 510.
As seen from the above, when carrying out image retrieval using scheme provided in an embodiment of the present invention, can be directed to and treat
Each target to be retrieved in retrieval image is retrieved, and reduces the information pair in the background of image to be retrieved
The influence of retrieval result, improves the accuracy rate of image retrieval.
For system or device embodiment, because it is substantially similar to embodiment of the method, so description
Fairly simple, the relevent part can refer to the partial explaination of embodiments of method.
It should be noted that herein, such as first and second or the like relational terms be used merely to by
One entity or operation make a distinction with another entity or operation, and not necessarily require or imply these
There is any this actual relation or order between entity or operation.And, term " including ", " bag
Containing " or any other variant thereof is intended to cover non-exclusive inclusion, so that including a series of key elements
Process, method, article or equipment not only include those key elements, but also including being not expressly set out
Other key elements, or it is this process, method, article or the intrinsic key element of equipment also to include.
In the case of there is no more limitations, the key element limited by sentence "including a ...", it is not excluded that including
Also there is other identical element in the process of the key element, method, article or equipment.
One of ordinary skill in the art will appreciate that all or part of step in realizing above method implementation method
Rapid to can be by program to instruct the hardware of correlation to complete, described program can be stored in computer can
In reading storage medium, the storage medium for obtaining designated herein, such as：ROM/RAM, magnetic disc, CD etc..
Presently preferred embodiments of the present invention is the foregoing is only, protection model of the invention is not intended to limit
Enclose.All any modification, equivalent substitution and improvements made within the spirit and principles in the present invention etc., wrap
Containing within the scope of the present invention.
Claims (15)
1. a kind of image search method, it is characterised in that be applied to client, methods described includes：
Image to be retrieved is sent to retrieval server；
The information of the candidate target that the retrieval server sends is received, wherein, the candidate target is described
Retrieval server determines the target that rule determines from the image to be retrieved according to default candidate target；
According to user for the target determine instruction to be retrieved of the candidate target, target to be retrieved is determined；
The image retrieval instruction for the target to be retrieved is sent to the retrieval server, it is described to cause
Retrieval server carries out image retrieval according to described image search instruction；
Receive the image searching result for the target to be retrieved of the retrieval server feedback.
2. method according to claim 1, it is characterised in that
Included in the information of the candidate target：The unique mark of the candidate target；
The image retrieval instruction sent to the retrieval server for the target to be retrieved, including：
The image retrieval for sending the unique mark for carrying the target to be retrieved to the retrieval server refers to
Order.
3. method according to claim 1, it is characterised in that described that the candidate is directed to according to user
The target determine instruction to be retrieved of target, determines target to be retrieved, including：
Based on the information of the candidate target, the candidate target is showed into user；
The target determine instruction to be retrieved of the candidate target is directed to according to user, from the candidate target really
Fixed target to be retrieved.
4. method according to claim 1, it is characterised in that described that the candidate is directed to according to user
The target determine instruction to be retrieved of target, determines target to be retrieved, including：
Based on the information of the candidate target, the candidate target is showed into user；
Whether comprising the desired candidate target of user in the shown candidate target of judgement；
If it is, determining that the desired candidate target of user is target to be retrieved；Otherwise, according to user described
The selection region of the image described to be retrieved of hand drawn determines target to be retrieved in client.
5. method according to claim 1, it is characterised in that described that the candidate is directed to according to user
The target determine instruction to be retrieved of target, determines target to be retrieved, including：
Whether the information for judging the candidate target is sky information；
If it is, according to the selection region of user's image described to be retrieved of hand drawn in the client
Determine target to be retrieved.
6. a kind of image search method, it is characterised in that be applied to retrieval server, methods described includes：
Receive the image to be retrieved that client sends；
Rule is determined according to default candidate target, candidate target is determined from the image to be retrieved；
The information of identified candidate target is sent to the client；
Receive the image retrieval instruction for the target to be retrieved that the client sends；Wherein, it is described
Target to be retrieved is the target determine instruction to be retrieved that the client is directed to the candidate target according to user
The target of determination；
Image retrieval is carried out according to described image search instruction, the image obtained for the target to be retrieved is examined
Hitch is really；
Described image retrieval result is sent to the client.
7. method according to claim 6, it is characterised in that described by identified candidate target
Information is sent to the client, including：
Obtain the retrieval frequency values of the candidate target；
According to the height of the retrieval frequency values, the unique mark of the candidate target is generated；
The information that the unique mark of the candidate target will be carried is sent to the client.
8. a kind of image retrieving apparatus, it is characterised in that be applied to client, described device includes：
Retrieval image sending module, for sending image to be retrieved to retrieval server；
Information receiving module, the information for receiving the candidate target that the retrieval server sends, wherein,
The candidate target is that the retrieval server determines rule from described to be retrieved according to default candidate target
The target determined in image；
Searched targets determining module, for determining to refer to for the target to be retrieved of the candidate target according to user
Order, determines target to be retrieved；
Search instruction sending module, for sending the figure for the target to be retrieved to the retrieval server
As search instruction, to cause that the retrieval server carries out image retrieval according to described image search instruction；
Retrieval result receiver module, for receive retrieval server feedback for the target to be retrieved
Image searching result.
9. device according to claim 8, it is characterised in that
The search instruction sending module, for including the candidate target in the information of the candidate target
Unique mark when, to the retrieval server send carry the target to be retrieved unique mark figure
As search instruction.
10. device according to claim 8, it is characterised in that the searched targets determining module,
Including：Candidate target shows submodule and searched targets determination sub-module；Wherein,
The candidate target shows submodule, for the information based on the candidate target, by candidate's mesh
Mark shows user；
The searched targets determination sub-module, the target to be retrieved for being directed to the candidate target according to user
Determine instruction, determines target to be retrieved from the candidate target.
11. devices according to claim 8, it is characterised in that the searched targets determining module,
Including：Candidate target displaying submodule, the first candidate target judging submodule, the first searched targets determine son
Module and the second searched targets determination sub-module；Wherein,
The candidate target shows submodule, for the information based on the candidate target, by candidate's mesh
Mark shows user；
The first candidate target judging submodule, for judging whether include use in shown candidate target
The desired candidate target in family；
The first searched targets determination sub-module, for sentencing in the first candidate target judging submodule
Disconnected result is target to be retrieved when being, to determine the desired candidate target of user；
The second searched targets determination sub-module, for sentencing in the first candidate target judging submodule
When disconnected result is no, according to the selection area of user's image described to be retrieved of hand drawn in the client
Domain determines target to be retrieved.
12. devices according to claim 8, it is characterised in that the searched targets determining module,
Including：Second candidate target judging submodule, the 3rd searched targets determination sub-module；Wherein,
The second candidate target judging submodule, whether the information for judging the candidate target is empty letter
Breath；
The 3rd searched targets determination sub-module, for sentencing in the second candidate target judging submodule
Disconnected result is when being, according to the selection area of user's image described to be retrieved of hand drawn in the client
Domain determines target to be retrieved.
13. a kind of image retrieving apparatus, it is characterised in that be applied to retrieval server, described device includes：
Retrieval image receiver module, the image to be retrieved for receiving client transmission；
Candidate target determining module, for determining rule according to default candidate target, from the figure to be retrieved
Determine candidate target as in；
Information sending module, for the information of identified candidate target to be sent to the client；
Search instruction receiver module, for receiving the figure for the target to be retrieved that the client sends
As search instruction；Wherein, the target to be retrieved is that the client is directed to the candidate target according to user
Target determine instruction to be retrieved determine target；
Retrieval result obtains module, for carrying out image retrieval according to described image search instruction, is directed to
The image searching result of the target to be retrieved；
Retrieval result sending module, for described image retrieval result to be sent to the client.
14. devices according to claim 13, it is characterised in that described information sending module, including：
Retrieval frequency obtains submodule, unique mark generation submodule and information sending submodule；Wherein,
The retrieval frequency obtains submodule, the retrieval frequency values for obtaining the candidate target；
The unique mark generates submodule, and for the height according to the retrieval frequency values, generation is described
The unique mark of candidate target；
Described information sending submodule, sends for will carry the information of unique mark of the candidate target
To the client.
15. a kind of image indexing systems, it is characterised in that the system includes：
Using the client of claim 8-12 any one described devices and using claim any one of 13-14
The retrieval server of described device.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510874622.1A CN106815272A (en) | 2015-12-02 | 2015-12-02 | A kind of image search method, apparatus and system |
| CN202311305431.4A CN117370590A (en) | 2015-12-02 | 2015-12-02 | Image retrieval method, device and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510874622.1A CN106815272A (en) | 2015-12-02 | 2015-12-02 | A kind of image search method, apparatus and system |
Related Child Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311305431.4A Division CN117370590A (en) | 2015-12-02 | 2015-12-02 | Image retrieval method, device and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106815272A true CN106815272A (en) | 2017-06-09 |
Family
ID=59106324
Family Applications (2)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510874622.1A Pending CN106815272A (en) | 2015-12-02 | 2015-12-02 | A kind of image search method, apparatus and system |
| CN202311305431.4A Pending CN117370590A (en) | 2015-12-02 | 2015-12-02 | Image retrieval method, device and system |
Family Applications After (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311305431.4A Pending CN117370590A (en) | 2015-12-02 | 2015-12-02 | Image retrieval method, device and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (2) | CN106815272A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111209331A (en) * | 2020-01-06 | 2020-05-29 | 北京旷视科技有限公司 | Target object retrieval method and device and electronic equipment |
| CN111522969A (en) * | 2020-03-31 | 2020-08-11 | 北京旷视科技有限公司 | Image retrieval method, image retrieval device, computer equipment and storage medium |
| CN111814008A (en) * | 2019-04-11 | 2020-10-23 | 杭州海康威视数字技术股份有限公司 | Data processing method and device, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1998052119A1 (en) * | 1997-05-16 | 1998-11-19 | The Trustees Of Columbia University In The City Of New York | Method and system for image retrieval |
| CN1570969A (en) * | 2003-07-23 | 2005-01-26 | 西北工业大学 | An image retrieval method based on marked interest point |
| CN102687140A (en) * | 2009-12-30 | 2012-09-19 | 诺基亚公司 | Methods and apparatuses for facilitating content-based image retrieval |
-
2015
- 2015-12-02 CN CN201510874622.1A patent/CN106815272A/en active Pending
- 2015-12-02 CN CN202311305431.4A patent/CN117370590A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1998052119A1 (en) * | 1997-05-16 | 1998-11-19 | The Trustees Of Columbia University In The City Of New York | Method and system for image retrieval |
| CN1570969A (en) * | 2003-07-23 | 2005-01-26 | 西北工业大学 | An image retrieval method based on marked interest point |
| CN102687140A (en) * | 2009-12-30 | 2012-09-19 | 诺基亚公司 | Methods and apparatuses for facilitating content-based image retrieval |
Non-Patent Citations (1)
| Title |
|---|
| 罗小燕: "基于区域特征的图像检索技术研究", 《万方学位论文数据库》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111814008A (en) * | 2019-04-11 | 2020-10-23 | 杭州海康威视数字技术股份有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN111814008B (en) * | 2019-04-11 | 2023-10-10 | 杭州海康威视数字技术股份有限公司 | Data processing method, device, electronic equipment and storage medium |
| CN111209331A (en) * | 2020-01-06 | 2020-05-29 | 北京旷视科技有限公司 | Target object retrieval method and device and electronic equipment |
| CN111209331B (en) * | 2020-01-06 | 2023-06-16 | 北京旷视科技有限公司 | Target object retrieval method and device and electronic equipment |
| CN111522969A (en) * | 2020-03-31 | 2020-08-11 | 北京旷视科技有限公司 | Image retrieval method, image retrieval device, computer equipment and storage medium |
| WO2021196551A1 (en) * | 2020-03-31 | 2021-10-07 | 北京旷视科技有限公司 | Image retrieval method and apparatus, computer device, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117370590A (en) | 2024-01-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10353947B2 (en) | Relevancy evaluation for image search results | |
| CN102959542B (en) | For the method and apparatus of managing video content | |
| US20080016046A1 (en) | Query categorizer | |
| US20140218400A1 (en) | Method for Providing Real Estate Data on an Interactive Map | |
| CN107679211A (en) | Method and apparatus for pushed information | |
| WO2013192101A1 (en) | Ranking search results based on click through rates | |
| EP2264623A1 (en) | Recommendation information generation apparatus and recommendation information generation method | |
| CN107633023B (en) | Image duplicate removal method and device | |
| CN104346370A (en) | Method and device for image searching and image text information acquiring | |
| CN105718558B (en) | The display methods and device of picture in the page | |
| CN106610988B (en) | Webpage recommendation method and recommendation device | |
| CN108345601A (en) | Search result ordering method and device | |
| CN110298716A (en) | Information-pushing method and device | |
| CN106815272A (en) | A kind of image search method, apparatus and system | |
| CN104991930A (en) | Webpage information sharing method and apparatus | |
| CN110473043B (en) | Article recommendation method and device based on user behaviors | |
| CN109885651A (en) | A kind of question pushing method and device | |
| CN105786810B (en) | The method for building up and device of classification mapping relations | |
| CN103955480A (en) | Method and equipment for determining target object information corresponding to user | |
| CN110929058A (en) | Trademark picture retrieval method and device, storage medium and electronic device | |
| CN111259274B (en) | Information processing method, device, equipment and information display device | |
| JP2011253240A (en) | Information display program, information display program recording computer-readable recording medium, information display method, information display device and information service system | |
| US20130238624A1 (en) | Search system and operating method thereof | |
| US20160210335A1 (en) | Server and service searching method of the server | |
| CN111488434A (en) | Recommendation method and device for input associative words, storage medium and electronic equipment |
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: 20170609 |