CN101984420A

CN101984420A - Method and equipment for searching pictures based on word segmentation processing

Info

Publication number: CN101984420A
Application number: CN2010102717985A
Authority: CN
Inventors: 魏星; 郭昶; 聂昆; 王梓旋; 常齐; 吕文先; 高静霞
Original assignee: Beijing Baidu Netcom Science and Technology Co Ltd
Current assignee: Baidu Online Network Technology Beijing Co Ltd; Beijing Baidu Netcom Science and Technology Co Ltd
Priority date: 2010-09-03
Filing date: 2010-09-03
Publication date: 2011-03-09
Anticipated expiration: 2030-09-03
Also published as: CN101984420B

Abstract

The invention provides a method and equipment for searching pictures based on word segmentation processing, and the method comprises the following steps: network equipment acquires a first search sequence input when a user searches the pictures in user equipment; the word segmentation processing is carried out on the first search sequence to obtain one or more second search sequences; matching search is carried out according to the second search sequences to obtain one or more pictures matched with the second search sequences; and one or more pictures are provided to the user equipment. Compared with the prior art, in the invention, the second search sequences are obtained by carrying out the word segmentation processing on the first search sequence, and the pictures are searched according to the second search sequences, and the searched pictures are provided to a user, thereby improving the searching efficiency of the pictures so that the user can fast and correctly acquire the desirable pictures through the pictures searching, thus the satisfaction degree and the experience degree of the user are promoted.

Description

A kind of based on the method and apparatus of tearing speech processing carrying out picture searching open

Technical field

The present invention relates to the picture searching technical field, relate in particular to a kind of based on the technology of tearing speech processing carrying out picture searching open.

Background technology

Along with the development of internet, applications and image processing techniques, increasing people begin to obtain and share the own picture of being liked by network, as the star's who dotes on stage photo, description, living photo, and the various places landscape of taking shines etc.

Yet, because other forms of media resources such as the relative literal of picture, have characteristics such as quantity is many, classification is assorted, bandwidth consumption is big, display requirement height, this picture that makes the user find that search engine returns not is own desired, perhaps need through repeatedly adjusting the picture searching keyword, just can obtain some and meet the picture that expectation requires, thereby have a strong impact on user's experience.Especially, the mobile phone users that day by day increases for quantity since the upstream and downstream bandwidth of portable terminal is limited, screen is less than normal or Presentation Function relatively a little less than, make this particularly outstanding.

Therefore, how to improve the efficient of picture searching, obtain its desirable picture fast and accurately by picture searching, thereby promote user's satisfaction and Experience Degree, become a problem needing solution now badly with the support user.

Summary of the invention

The purpose of this invention is to provide a kind of based on the method and apparatus of tearing speech processing carrying out picture searching open.

According to an aspect of the present invention, provide a kind of and handled the method for carrying out picture searching based on tearing speech open, this method may further comprise the steps:

A. obtain first search sequence of importing when the user carries out picture searching in subscriber equipment;

B. described first search sequence is torn open speech and handle, to obtain one or more second search sequence;

C. carry out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence;

D. described one or more pictures are offered described subscriber equipment.

According to a further aspect in the invention, also provide a kind of and handled the equipment that carries out picture searching based on tearing speech open, this equipment comprises:

Deriving means is used for obtaining first search sequence that the user imports when subscriber equipment carries out picture searching;

Tear the speech device open, be used for that described first search sequence is torn open speech and handle, to obtain one or more second search sequence;

Searcher is used for carrying out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence;

Generator is used for described one or more pictures are offered described subscriber equipment.

Compared with prior art, the present invention handles acquisition second search sequence by first search sequence being torn open speech, according to the second time search sequence carry out picture searching, and will search picture offer the user, improved the efficient of picture searching, make the user to obtain its desirable picture fast and accurately, thereby promoted user's satisfaction and Experience Degree by picture searching.

Description of drawings

By reading the detailed description of doing with reference to the following drawings that non-limiting example is done, it is more obvious that other features, objects and advantages of the present invention will become:

Fig. 1 illustrates and handles the system topological figure carry out picture searching according to one aspect of the invention based on tearing speech open;

Fig. 2 illustrates and handles the equipment synoptic diagram carry out picture searching according to one aspect of the invention based on tearing speech open;

Fig. 3 illustrate according to a further aspect of the present invention handle the method flow diagram carry out picture searching based on tearing speech open;

Fig. 4 illustrate in accordance with a preferred embodiment of the present invention handle the method flow diagram carry out picture searching based on tearing speech open.

Same or analogous Reference numeral is represented same or analogous parts in the accompanying drawing.

Embodiment

Below in conjunction with accompanying drawing the present invention is described in further detail.

Fig. 1 illustrates being used for based on the system topological figure that tears speech processing carrying out picture searching open, comprising the network equipment 1 and a plurality of subscriber equipment 2 that is connected via network with it according to one aspect of the invention.Wherein, the network equipment 1 can be connected via network with a plurality of subscriber equipment 2, and obtains first search sequence of importing when the user carries out picture searching by subscriber equipment 2; Described first search sequence is torn open speech handle, to obtain one or more second search sequence; Carry out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence; Described one or more pictures are offered described subscriber equipment.At this, network includes but not limited to internet, wide area network, Metropolitan Area Network (MAN), LAN (Local Area Network), VPN network, wireless self-organization network (Ad Hoc network) etc.

In addition, the network equipment 1 includes but not limited to network host, single network server, a plurality of webserver collection or based on the set of computers of cloud computing.And subscriber equipment 2 can be any electronic product that can carry out man-machine interaction by modes such as keyboard, mouse, telepilot, touch pad or hand-written equipment with the user, for example computing machine, mobile phone, PDA, palm PC PPC or IPTV etc.Communicating by letter between the network equipment 1 and a plurality of subscriber equipmenies 2 is separate, can be based on the packet data transmission such as ICP/IP protocol, udp protocol etc.

Those skilled in the art will be understood that the above-mentioned network equipment 1, subscriber equipment 2 and connection network and communication mode therebetween only is for example; other network equipments existing or that may occur from now on, subscriber equipment, network or communication mode are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Those skilled in the art also will be understood that, the quantity of the disparate networks element that only illustrates for simplicity's sake among Fig. 1 may be less than the quantity in the real network, but this omission far and away with can not influence to the present invention carry out clear, be disclosed as prerequisite fully.

For simplicity's sake, be that example is described with a subscriber equipment 2 below.Those skilled in the art will be understood that the network equipment 1 is can be simultaneously mutual with a plurality of subscriber equipment 2, and first search sequence of importing when carrying out picture searching according to the user who obtains from different subscriber equipmenies 2; Described first search sequence is torn open speech handle, to obtain one or more second search sequence; Carry out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence; Described one or more pictures are offered different subscriber equipment 2 respectively.

Fig. 2 illustrate according to one aspect of the invention handle the equipment synoptic diagram carry out picture searching based on tearing speech open, this network equipment 1 comprises deriving means 11, tears speech device 12 open, searcher 13 and generator 14.

Wherein, deriving means 11 is used for obtaining first search sequence that the user imports when subscriber equipment carries out picture searching.Particularly, the user by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, input first search sequence is to carry out picture searching in browser or client software; With the keyboard is example, when knocking keyboard in the input frame of user at browser and import, deriving means 11 is by page technology, as JSP, ASP, PHP technology, obtain the keystroke sequence that the user knocks in real time, and be recorded as first search sequence of importing when the user carries out picture searching.Those skilled in the art will be understood that above-mentioned several page technology only for giving an example, and other page technology existing or that may occur from now on also should be included in the protection domain of the present invention as applicable to the present invention, and are contained in this at this with way of reference.

Tear speech device 12 open and be used for described first search sequence is torn open the speech processing, to obtain one or more second search sequence.Particularly, after tearing speech device 12 open first search sequence that provides by deriving means 11 being provided, pass through matching inquiry, for example by at the descriptor split table or tear modes such as carrying out match query in the dictionary open, described first search sequence is torn open speech handle, to obtain one or more second search sequence.For example, when picture searching is carried out in user's input " Fan Bingbing ", deriving means 11 gets access to first search sequence " Fan Bingbing " and this is passed to tears speech device 12 open, tear speech device 12 open according to " Fan Bingbing ", match query in the descriptor split table, inquiring second corresponding search sequence of " Fan Bingbing " is " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc.And for example, picture searching is carried out in user's input " Fan Bingbing ", deriving means 11 gets access to first search sequence " Fan Bingbing " and this is passed to tears speech device 12 open, tear speech device 12 open according to the category " star " under " Fan Bingbing ", in tearing dictionary open, carry out match query, inquire that " star " is pairing to tear speech open and have " stage photo ", " description ", " living photo " etc., the above-mentioned speech of tearing open is combined with first search sequence " Fan Bingbing " and obtains second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " model ice is iced living photo " etc.Those skilled in the art will be understood that the above-mentioned mode of speech processing of tearing open is only for giving an example; other existing or tearing open of may occurring from now on modes that speech handles are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Searcher 13 is used for carrying out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence.Particularly, searcher 13 is according to tearing second search sequence that speech device 12 is provided open, carry out match search, for example carry out online picture searching, carry out picture searching in the picture indices storehouse of this locality or both combine, obtain the one or more pictures that are complementary with described second search sequence according to second search sequence according to second search sequence.For example, tear open speech device 12 tear open according to first search sequence " Fan Bingbing " of user input speech handle obtain second search sequence " model ice ice stage photo " after, searcher 13 carries out match search, and search obtains containing the picture of " model ice ice stage photo " in picture tag, picture text message or picture keyword.Those skilled in the art be will be understood that the mode of above-mentioned match search is only for giving an example; the mode of other match searchs existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.At this, described match search can have multiple mode, for example according to the related text information of second search sequence to picture, as picture file name, picture tag, picture text message or picture keyword etc., carries out match search; Perhaps, determine the example picture of a correspondence according to second search sequence, this example picture and other pictures are carried out content-based match search, and as color, texture, plane space corresponding relation, profile based on image, perhaps other statistical natures carry out match search.Those skilled in the art be will be understood that the mode of above-mentioned picture searching is only for giving an example; the mode of other picture searchings existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.At this, searcher 13 carries out match search to one of tearing open in second search sequence that speech device 12 provided at least, to obtain one or more pictures.

Generator 14 is used for described one or more pictures are offered described subscriber equipment.Particularly, generator 14 is by page technology, and as JSP, ASP, PHP technology, picture and/or its link that searcher 13 search are obtained write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.Preferably, described picture can sort according to certain rule, and the picture after the described ordering is write one or more webpages, and described webpage and/or link are offered subscriber equipment 2.Connect example, after searcher 13 search obtain containing the picture of " model ice ice stage photo " in picture tag, picture text message or picture keyword, generator 14 passes through page technology, as JSP, ASP, PHP technology, described picture and/or its link are write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2, select to browse for the user.Those skilled in the art will be understood that above-mentioned page technology only for giving an example, and other page technology existing or that may occur from now on also should be included in the protection domain of the present invention as applicable to the present invention, and are contained in this at this with way of reference.Those skilled in the art will be understood that also the above-mentioned mode of picture that provides is only for giving an example; other existing or modes that picture is provided that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Preferably, between the device of each in the network equipment 1 be constant work.Particularly, deriving means 11 obtains first search sequence of importing when the user carries out picture searching in subscriber equipment in real time; Tear 12 pairs of described first search sequence of speech device open and tear the speech processing open, to obtain one or more second search sequence; Searcher 13 carries out match search according to described second search sequence again, with the one or more pictures that obtain to be complementary with described second search sequence; Generator 14 further offers described subscriber equipment with described one or more pictures.At this, it will be understood by those skilled in the art that " continuing " is meant that each device requires to carry out obtaining, tear open speech processing, picture searching and corresponding picture being provided of first search sequence according to the mode of operation of setting or adjust in real time respectively, stop to carry out network operations such as picture searchings until this user in a long time by subscriber equipment 2.

Preferably, searcher 13 can also comprise the interactive device (not shown).Described interactive device is used for described one or more second search sequence are offered described subscriber equipment 2, and obtains second search sequence that this user selects from described one or more second search sequence; Searcher 13 carries out match search according to this second user-selected search sequence again, with the one or more pictures that obtain to be complementary with this second search sequence.Particularly, tear speech open when first search sequence of tearing 12 pairs of user's inputs of speech device open and handle, obtain one or more second search sequence, interactive device passes through page technology, as JSP, ASP, PHP technology, described one or more second search sequence are offered described subscriber equipment 2.Preferably, interactive device can be shown to the user with the form of described one or more second search sequence by the drop-down alternate item of the input frame of first search sequence, also can be shown to the user by the form of suspension frame.Those skilled in the art will be understood that the above-mentioned display format that one or more second search sequence are offered subscriber equipment 2 is only for for example; other display formats existing or that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.Afterwards, the user browses by browser software in the subscriber equipment 2 or client software, and by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, described one or more second search sequence of tearing speech open that contain are selected.With the mouse is example, when choosing or clicking certain second search sequence with mouse in the drop-down menu of user at input frame, interactive device is by submitting, receive forms such as request of access to detection cursor focus, reception list, obtain user's selection that this user carries out second search sequence, for subsequent treatment.After this, described searcher 13 also is used for carrying out match search according to the second user-selected search sequence, with the one or more pictures that obtain to be complementary with described second search sequence.For example, when importing first search sequence " Fan Bingbing " in the input frame of user at picture searching, deriving means 11 as JSP, ASP, PHP technology, gets access to described first search sequence " Fan Bingbing " by page technology; Tear speech device 12 open and obtain corresponding second search sequence of this first search sequence " Fan Bingbing " " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " according to the coupling of tabling look-up by tearing the speech processing open again; Interactive device offers subscriber equipment with above-mentioned second search sequence in the drop-down alternate item of this input frame, select for the user; If the user chooses or click selections " model ice ice stage photo ", interactive device is again by detecting cursor focus, receiving list and submit, receive forms such as request of access to, and that obtains this user is chosen as second search sequence " model ice is iced stage photo "; Searcher 13 carries out match search according to the second user-selected search sequence " model ice ice stage photo ", obtains the one or more pictures that are complementary with described second search sequence " model ice ice stage photo "; Generator 14 is again by page technology, and as JSP, ASP, PHP technology, the one or more pictures that will be above-mentioned be complementary with " model is iced stage photo " and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.

Preferably, the described speech device 12 of tearing open also is used for described first search sequence being torn open speech handling, to obtain one or more second search sequence according to tearing dictionary open.Particularly, after tearing speech device 12 open first search sequence that provides by deriving means 11 being provided, in tearing dictionary open, first search sequence is carried out matching inquiry, described first search sequence is torn open speech handle, to obtain one or more second search sequence.For example, picture searching is carried out in user input " Fan Bingbing ", and deriving means 11 gets access to first search sequence " Fan Bingbing " and this is passed to tears speech device 12 open.Tear speech device 12 open and carry out matching inquiry according to " Fan Bingbing " in tearing dictionary open, inquiring second corresponding search sequence of " Fan Bingbing " is " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc.Searcher 13 carries out match search according to described second search sequence, obtains the one or more pictures that are complementary with described second search sequence.Generator 14 is again by page technology, and as JSP, ASP, PHP technology, one or more pictures that will be complementary with described second search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.More preferably, tearing open after speech device 12 inquiry obtains second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " model ice is iced living photo " etc., interactive device can offer subscriber equipment 2 with forms such as the drop-down alternate item of described second search sequence by the input frame of first search sequence, suspension frames, selects for the user.Then, the user by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, described one or more second search sequence of tearing speech open that contain are selected.Interactive device passes through to obtain user's selection that this user carries out second search sequence to detect cursor focus, to receive forms such as list is submitted to, reception request of access.Searcher 13 carries out match search according to the second user-selected search sequence again, obtains the one or more pictures that are complementary with described second search sequence.Generator 14 passes through page technology, as JSP, ASP, PHP technology, one or more pictures that will be complementary with the second user-selected search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.And for example, picture searching is carried out in user input " Fan Bingbing ", and deriving means 11 gets access to first search sequence " Fan Bingbing " and this is passed to tears speech device 12 open.Tear speech device 12 open and in tearing dictionary open, carry out matching inquiry, inquire that " star " is pairing to tear speech open " stage photo ", " description ", " living photo " etc. are arranged according to the category " star " under " Fan Bingbing "; Then, tear speech device 12 open and the more above-mentioned speech of tearing open is combined with first search sequence " Fan Bingbing ", obtain second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc.Searcher 13 carries out match search according to described second search sequence, obtains the one or more pictures that are complementary with described second search sequence.Generator 14 is again by page technology, and as JSP, ASP, PHP technology, one or more pictures that will be complementary with described second search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.More preferably, to tear speech " stage photo ", " description ", " living photo " open and combine tearing speech device 12 open with first search sequence " Fan Bingbing ", obtain after second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " the model ice ice living photo " etc., interactive device can offer subscriber equipment 2 with forms such as the drop-down alternate item of described second search sequence by the input frame of first search sequence, suspension frames, selects for the user.Then, the user by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, described one or more second search sequence of tearing speech open that contain are selected.Interactive device passes through to obtain user's selection that this user carries out second search sequence to detect cursor focus, to receive forms such as list is submitted to, reception request of access.Searcher 13 carries out match search according to the second user-selected search sequence again, obtains the one or more pictures that are complementary with described second search sequence.Generator 14 passes through page technology, as JSP, ASP, PHP technology, one or more pictures that will be complementary with the second user-selected search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.Those skilled in the art will be understood that above-mentioned basis tears dictionary open and tear mode that speech handles open only for for example; other bases existing or that may occur are from now on torn dictionary open and are torn mode that speech handles open as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Preferably, the network equipment 1 can also comprise the first updating device (not shown), is used for according to the statistical study of the employed picture searching keyword of a large number of users, sets up or upgrades the described dictionary of tearing open.Particularly, first updating device is according to the statistical study of the employed picture searching keyword of user in a period of time, as analyzing the picture searching keyword that obtains " Fan Bingbing " by statistics " Fan Bingbing description " (accounting for 33%) arranged, " model ice ice stage photo " (accounting for 31%), " model ice ice living photo " (accounting for 28%), " model ice is iced other " (accounting for 8%), and then first search sequence of determining to relate to " Fan Bingbing " can corresponding second search sequence " Fan Bingbing description ", " model ice ice stage photo ", " model ice ice living photo " etc., set up or upgrade this and tear dictionary open, correspondingly add in the dictionary or revise " Fan Bingbing " pairing speech of tearing open and handle record as tearing open at this.At this, first updating device both can independently have been finished the statistical study to the employed picture searching keyword of a large number of users, also can accept from other products or the third party device statistic analysis result to the employed picture searching keyword of a large number of users.For example, when first updating device is independently finished statistical study to the employed picture searching keyword of a large number of users, the employed picture searching keyword of user in first updating device recording storage a period of time, then these picture searching keywords are carried out statistical study, for example draw the shared number percent of various picture searching keywords of corresponding same descriptor, get one or more second search sequence of the bigger picture retrieval keyword of wherein shared number percent as this descriptor correspondence, getting the picture searching keyword that frequency of usage surpasses certain threshold value for another example is one of second search sequence of corresponding descriptor.For example, when first user uses picture searching keyword " model ice ice stage photo ", first updating device can be noted it, and the frequency of usage of described picture searching keyword " model ice ice stage photo " added up, when its frequency that is used surpasses preset threshold (for example 1000 times), this picture searching keyword is deposited in and tears dictionary open as tearing speech open.Thereafter, when picture searching was carried out in arbitrary user's input " Fan Bingbing ", the network equipment 1 can carry out match search as second search sequence with this picture searching keyword " model ice ice stage photo ", perhaps offers the user and selects for it.For another example, when a large number of users uses the form of " so-and-so star's stage photo " to carry out picture searching as first search sequence, first updating device can be with wherein picture searching keyword " stage photo " as tearing speech open, deposit in and tear dictionary open, when arbitrary user imports first search sequence that comprises " so-and-so star " printed words once more and carries out picture searching, the network equipment 1 can be torn this open speech " stage photo " and combine as second search sequence with " so-and-so star " and carry out match search, perhaps offers the user and selects for it.Preferably, the user can import first search sequence in search pictures, can obtain the used picture searching keyword of search pictures by it being carried out pre-service or tearing the speech processing open.For example, input first search sequence in the input frame of user at picture searching " model ice ice atlas sheet " is because the hunting zone of search engine itself is exactly a picture, therefore, remove " picture " and this time want information, the picture searching keyword that obtains described first search sequence is " Fan Bingbing ".And for example, when the user imported first search sequence " model ice ice stage photo ", " model ice ice stage photo " promptly became the picture searching keyword.Those skilled in the art be will be understood that it only is for example that the mode of dictionary is torn in above-mentioned renewal open; the mode that dictionary is torn in other renewals existing or that may occur from now on open is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

More preferably, the described first updating device (not shown) also is used for the statistical study selected according to the employed picture searching keyword of a large number of users and pairing picture thereof, sets up or upgrades the described dictionary of tearing open.Particularly, first updating device can also be according to for a large number of users employed picture searching keyword and final statistical study of selecting the picture browsed thereof when the search pictures, for example correct corresponding picture searching keyword or derive corresponding picture searching keyword according to the selected picture of browsing according to the selected picture of browsing, set up or upgrade this and tear dictionary open, for example tear open and correspondingly add in the dictionary or revise the pairing speech of tearing open of one or more descriptor and handle record at this.For example, when the user utilizes " Fan Bingbing " to carry out picture searching, according to the descriptors such as label, text message or keyword in this user-selected picture, the picture that this user selects to browse is " stage photo " of " Fan Bingbing ", and the picture retrieval keyword of then revising this user is " model ice ice stage photo "; And,, upgrade this and tear dictionary open according to statistical study to the employed picture searching keyword of a large number of users according to last embodiment (basic identical because of embodiment, as to repeat no more, and be contained in this by reference) at this.And for example, in the picture of relevant " Fan Bingbing " that offer the user, by extraction to picture tag, text message or keyword, what determine a large number of users selection is the picture of " stage photo " of relevant " Fan Bingbing ", first updating device can deposit " model ice ice stage photo " in and tear dictionary open as second search sequence of correspondence " Fan Bingbing ", to upgrade the described dictionary of tearing open.Those skilled in the art be will be understood that it only is for example that the mode of dictionary is torn in above-mentioned renewal open; the mode that dictionary is torn in other renewals existing or that may occur from now on open is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.At this, first updating device both can independently have been finished the statistical study that the employed picture searching keyword of a large number of users and pairing picture thereof are selected, and also can accept the statistic analysis result of the employed picture searching keyword of a large number of users and pairing picture thereof being selected from other products or third party device.For example, when first updating device is independently finished the statistical study that the employed picture searching keyword of a large number of users and pairing picture thereof are selected, employed picture searching keyword of user and pairing picture thereof are selected in first updating device recording storage a period of time, then statistical study selected to carry out in these picture searching keywords and pairing picture thereof, the various picture searching keywords that for example draw corresponding same descriptor (comprise the picture searching keyword after correcting according to the picture of selecting to browse, picture searching keyword according to the picture derivation of selecting to browse) shared number percent, get one or more second search sequence of the bigger picture retrieval keyword of wherein shared number percent, get the picture searching keyword that frequency of usage surpasses certain threshold value for another example and (comprise the picture searching keyword after correcting according to the picture of selecting to browse as this descriptor correspondence, picture searching keyword according to the picture derivation of selecting to browse) be one of second search sequence of corresponding descriptor.

Preferably, described searcher 13 also is used for according to described second search sequence, and in conjunction with this user's user related information, carries out match search, with the one or more pictures that obtain to be complementary with described second search sequence and described user related information.Particularly, searcher 13 is torn second search sequence that speech device 12 provides open receiving, or during user-selected second search sequence that passes over by interactive device, in conjunction with the user related information network equipment 1 record or that this user sets, carry out match search, with the one or more pictures that obtain to be complementary with described second search sequence and described user related information.For example, if the preference options that the user sets is an automobile, then when second search sequence is " jaguar ", searcher 13 is according to second search sequence " jaguar ", and, carry out match search in conjunction with this user's preference options " automobile ", obtain the picture of relevant " jaguar automobile ", generator 14 offers subscriber equipment with the picture of described relevant " jaguar automobile " again, selects to browse for the user.And for example, if the preference options that the user sets is an animal, then when second search sequence is " jaguar ", searcher 13 is according to second search sequence " jaguar ", and, carry out match search in conjunction with this user's preference options " animal ", obtain the picture of relevant " jaguar animal ", generator 14 offers subscriber equipment with the picture of described relevant " jaguar automobile " again, selects to browse for the user.

Preferably, the network equipment 1 can also comprise the collator (not shown), is used for the user related information according to this user, and described one or more pictures are sorted; Wherein, described generator 14 also is used for the described one or more pictures after the ordering are offered described subscriber equipment.Particularly, carry out according to second search sequence after match search obtains one or more pictures at searcher 13, collator is according to user related information, for example, according to label, text message or the keyword of the picture that extracts and the matching degree of user related information, described one or more pictures are sorted; Generator 14, writes one or more webpages with described one or more pictures and/or its link after the ordering, and these webpages and/or its link is offered subscriber equipment 2 as JSP, ASP, PHP technology again by page technology.Connect example, if the preference options that the user sets is an automobile, then when second search sequence was " jaguar ", searcher 13 carried out match search according to second search sequence " jaguar ", obtained the picture of relevant " jaguar "; The preference options that collator is set according to the user, i.e. " automobile ", the picture of prioritization relevant " jaguar automobile "; Generator 14 offers subscriber equipment with the picture of described relevant " jaguar automobile " in proper order according to this prioritization again, selects to browse for the user.And for example, if the preference options that the user sets is an animal, then when second search sequence was " jaguar ", searcher 13 carried out match search according to second search sequence " jaguar ", obtained the picture of relevant " jaguar "; The preference options that collator is set according to the user, i.e. " animal ", the picture of prioritization relevant " jaguar animal "; Generator 14 offers subscriber equipment with the picture of described relevant " jaguar animal " in proper order according to this prioritization again, selects to browse for the user.Those skilled in the art be will be understood that the mode of above-mentioned picture ordering is only for giving an example; the mode of other pictures existing or that may occur from now on orderings is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

More preferably, described collator (not shown) also is used for the user related information according to this user, and described one or more second search sequence are sorted; Subsequently, described one or more second search sequence after searcher 13 will sort offer described subscriber equipment, and obtain second search sequence that this user selects from described one or more second search sequence, carry out match search according to second search sequence that described user selects, with the one or more pictures that obtain to be complementary with described second search sequence.Particularly, tear one or more second search sequence that the speech processing obtains open for tearing speech device 12 open, collator can sort to described one or more second search sequence according to this user's relevant information; Drop-down alternate item or the form such as suspension frame of described one or more search sequence after searcher 13 will sort again by the input frame of for example first search sequence offers described subscriber equipment; Then, searcher 13 by to detect cursor focus, to receive forms such as list is submitted to, reception request of access, obtains second search sequence that this user selects again from described one or more second search sequence.Preferably, the drop-down alternate item of this input frame or suspension frame can also be according to the adaptive adjustment of page size, for example, for screen constrained portable terminal, drop-down menu or suspension frame can diminish adaptively, and a demonstration is positioned at part second search sequence in the prostatitis of described second search sequence after the ordering.Those skilled in the art be will be understood that the mode of above-mentioned second search sequence ordering is only for giving an example; the mode of other orderings existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Preferably, described user related information includes but not limited to following at least one: a) described user's personal attribute; B) described user's input historical record; C) preference options of described user's setting; D) address information of described subscriber equipment.The network equipment 1 can be according to personal attribute's (for example user's age, sex, occupation, education background etc.) of user, the interested picture of match search user possibility, or know the user by inference may interested second search sequence or picture, described second search sequence or picture are stood out, and the confession user selects or browses.For example,, then when picture retrieval is carried out in its input " jaguar ", the picture of relevant " jaguar automobile " is stood out, be shown to the user if the network equipment 1 this user of record is the male sex.The network equipment 1 can also be according to the historical record of user input, and the match search user may interested picture, or know the user by inference may interested second search sequence or picture, described second search sequence or picture are stood out, for user's selection or browse.For example, the user once imported " A Fanda " and carried out picture searching, and so, when this user imported " film " again and carries out picture searching, the figure sector-meeting relevant with " A Fanda " stood out, and is shown to this user.The network equipment 1 can also be according to the preference options (for example " film ", " TV ", " stage photo ", " description ", " living photo " etc.) of user's setting, the interested picture of match search user possibility, or know the user by inference may interested second search sequence or picture, so that described second search sequence or picture are stood out, the confession user selects or browses.For example, if the preference that the user sets is a still, then when user's input " Fan Bingbing ", in second search sequence, standing out in " model ice ice stage photo apple ", " model ice ice stage photo besieged city in October " etc., selects for the user, in the picture that returns, pictures such as relevant " model ice ice stage photo apple ", " model ice ice stage photo besieged city in October " are stood out, and select to browse for the user; If user's preference is the TV play stage photo, then when user's input " Fan Bingbing ", in second search sequence, " model ice ice stage photo also pearl sound of laughing ", " model ice ice stage photo kermes snow " etc. are stood out, select for the user, in the picture that returns, relevant " model ice ice stage photo also pearl sound of laughing ", " model ice ice stage photo kermes snow " wait picture to stand out, and browse for user's selection.The network equipment 1 can also be according to the address information (for example user device IP address, GPS information etc.) of subscriber equipment, the interested picture of match search user possibility, or know the user by inference may interested second search sequence or picture, so that described second search sequence or picture are stood out, the confession user selects or browses.For example, the network equipment 1 detects the current IP address of subscriber equipment 2 in Shanghai, first search sequence of obtaining for deriving means 11 " Nanjing Road " then, tear speech device 12 open and can tear the speech processing open, obtain for example " Nanjing Road discount store ", second search sequence such as " Nanjing Road star's activities ", searcher 13 can be searched for according to above-mentioned second search sequence, perhaps interactive device is stood out above-mentioned second search sequence, select for the user, collator can be stood out above-mentioned relevant picture, offer subscriber equipment by generator 14, browse selection for the user.And for example, the current IP address that detects subscriber equipment 2 when the network equipment 1 is not during in Shanghai, can know the travel information that it wants to understand the Nanjing Road, Shanghai by inference, tear speech device 12 open and can tear the speech processing open, obtain for example " Nanjing Road map ", second search sequence such as " Nanjing Road characteristic snacks ", searcher 13 can be searched for according to above-mentioned second search sequence, perhaps interactive device is stood out above-mentioned second search sequence, select for the user, collator can be stood out above-mentioned relevant picture, offer subscriber equipment by generator 14, browse selection for the user.Those skilled in the art be will be understood that above-mentioned user related information only for giving an example, and other user related informations existing or that may occur from now on also should be included in the protection domain of the present invention as applicable to the present invention, and are contained in this at this with way of reference.

Preferably, this equipment also comprises the second updating device (not shown), is used for the selection described one or more picture done according to described second search sequence and/or this user, upgrades described user related information.For example, second updating device can upgrade this user's user related informations such as input historical record according to user-selected second search sequence that interactive device provided.For example, when the user in " Fan Bingbing " corresponding second search sequence or picture, selected " Fan Bingbing description " or belonged to the picture of " Fan Bingbing description ", so, second updating device then correspondingly upgrades uses relevant information, as increasing historical record such as " Fan Bingbing description " input.

Those skilled in the art will be understood that above-mentioned first updating device and second updating device only are example, and in practice, they can be two independently modules, also can be integrated in the module.

Preferably, the described speech device 12 of tearing open also is used for described first search sequence is carried out pre-service; Tear pretreated described first search sequence open speech and handle, to obtain one or more second search sequence.Particularly, in first search sequence of user's input, may comprise some less important information, as " ", " picture " etc., perhaps described first search sequence is a long sentence, therefore, tearing speech device 12 open need carry out pre-service to these first search sequence, handles with convenient speech that this is torn open.Then, tear speech device 12 open according to pretreated first search sequence, by matching inquiry, for example by at the descriptor split table or tear modes such as carrying out match query in the dictionary open, described first search sequence is torn open speech handle, to obtain one or more second search sequence.

More preferably, described pre-service comprises following at least one: a) synonym is replaced; B) classification is handled; C) word segmentation processing.For example, when the user imports first search sequence " little S ", tear speech device 12 open and can carry out pre-service to this, " little S " synonym is replaced with " Xu Xidi ", handle further this is torn open speech.And for example, first search sequence " Fan Bingbing " of user's input belongs to " performer ", and " performer's picture " can be split as " performer's stage photo ", " performer's description ", " performer's living photo " etc., earlier this is classified to handle and further tear speech again open and handle, promptly obtain second search sequence " Fan Bingbing description ", " model ice ice stage photo " and " model is iced living photo " etc. of its correspondence according to first search sequence " Fan Bingbing ".For another example, in first search sequence of user input, may comprise some less important information (for example " "), can carry out word segmentation processing to this first search sequence, for example earlier these less important information are removed, perhaps long sentence is made pauses in reading unpunctuated ancient writings, remove and repeat vocabulary, therefrom extract descriptor again, carry out word segmentation processing, handle further this is torn open speech.Those skilled in the art be will be understood that above-mentioned first search sequence is carried out pretreated mode only for for example; other pretreated modes existing or that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Fig. 3 illustrate according to a further aspect of the present invention handle the method flow diagram carry out picture searching based on tearing speech open.

Wherein, the network equipment 1 includes but not limited to network host, single network server, a plurality of webserver collection or based on the set of computers of cloud computing.And subscriber equipment 2 can be any electronic product that can carry out man-machine interaction by modes such as keyboard, mouse, telepilot, touch pad or hand-written equipment with the user, for example computing machine, mobile phone, PDA, palm PC PPC or IPTV etc.Communicating by letter between the network equipment 1 and a plurality of subscriber equipmenies 2 is separate, can be based on the packet data transmission such as ICP/IP protocol, udp protocol etc.

Those skilled in the art will be understood that the above-mentioned network equipment 1, subscriber equipment 2 and communication mode therebetween are only for giving an example; other network equipments existing or that may occur from now on, subscriber equipment or communication mode are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Particularly, in step S1, the network equipment 1 obtains first search sequence of importing when the user carries out picture searching in subscriber equipment.Particularly, the user by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, input first search sequence is to carry out picture searching in browser or client software; With the keyboard is example, when knocking keyboard in the input frame of user at browser and import, the network equipment 1 is by page technology, as JSP, ASP, PHP technology, obtain the keystroke sequence that the user knocks in real time, and be recorded as first search sequence of importing when the user carries out picture searching.Those skilled in the art will be understood that above-mentioned several page technology only for giving an example, and other page technology existing or that may occur from now on also should be included in the protection domain of the present invention as applicable to the present invention, and are contained in this at this with way of reference.

In step S2,1 pair of described first search sequence of the network equipment is torn speech open and is handled, to obtain one or more second search sequence.Particularly, after the network equipment 1 obtains first search sequence of user input, by matching inquiry, for example by at the descriptor split table or tear modes such as carrying out match query in the dictionary open, described first search sequence is torn open speech handle, to obtain one or more second search sequence.For example, when picture searching is carried out in user's input " Fan Bingbing ", the network equipment 1 gets access to first search sequence " Fan Bingbing ", and according to described first search sequence " Fan Bingbing ", match query in the descriptor split table, inquiring second corresponding search sequence of " Fan Bingbing " is " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc.And for example, picture searching is carried out in user's input " Fan Bingbing ", the network equipment 1 gets access to first search sequence " Fan Bingbing ", and according to the category " star " under described first search sequence " Fan Bingbing ", in tearing dictionary open, carry out match query, inquire that " star " is pairing to tear speech open and have " stage photo ", " description ", " living photo " etc., the above-mentioned speech of tearing open is combined with first search sequence " Fan Bingbing " and obtains second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " model ice is iced living photo " etc.Those skilled in the art will be understood that the above-mentioned mode of speech processing of tearing open is only for giving an example; other existing or tearing open of may occurring from now on modes that speech handles are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

In step S3, the network equipment 1 carries out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence.Particularly, the network equipment 1 is handled second search sequence that obtains according to tearing speech open, carry out match search, for example carry out online picture searching, carry out picture searching in the picture indices storehouse of this locality or both combine, obtain the one or more pictures that are complementary with described second search sequence according to second search sequence according to second search sequence.For example, the network equipment 1 tear open according to first search sequence " Fan Bingbing " of user input speech handle obtain second search sequence " model ice ice stage photo " after, the network equipment 1 carries out match search, and search obtains containing the picture of " model ice ice stage photo " in picture tag, picture text message or picture keyword.Those skilled in the art be will be understood that the mode of above-mentioned match search is only for giving an example; the mode of other match searchs existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.At this, described match search can have multiple mode, for example according to the related text information of second search sequence to picture, as picture file name, picture tag, picture text message or picture keyword etc., carries out match search; Perhaps, determine the example picture of a correspondence according to second search sequence, this example picture and other pictures are carried out content-based match search, and as color, texture, plane space corresponding relation, profile based on image, perhaps other statistical natures carry out match search.Those skilled in the art be will be understood that the mode of above-mentioned picture searching is only for giving an example; the mode of other picture searchings existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.At this, the network equipment 1 one of is handled in second search sequence obtain and to be carried out match search tearing speech open at least, to obtain one or more pictures.

In step S4, the network equipment 1 offers described subscriber equipment with described one or more pictures.Particularly, the network equipment 1 is by page technology, and as JSP, ASP, PHP technology, picture and/or its link that search is obtained write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.Preferably, described picture can sort according to certain rule, and the picture after the described ordering is write one or more webpages, and described webpage and/or link are offered subscriber equipment 2.Connect example, after the network equipment 1 search obtains containing the picture of " model ice ice stage photo " in picture tag, picture text message or picture keyword, the network equipment 1 passes through page technology, as JSP, ASP, PHP technology, described picture and/or its link are write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2, select to browse for the user.Those skilled in the art will be understood that above-mentioned page technology only for giving an example, and other page technology existing or that may occur from now on also should be included in the protection domain of the present invention as applicable to the present invention, and are contained in this at this with way of reference.Those skilled in the art will be understood that also the above-mentioned mode of picture that provides is only for giving an example; other existing or modes that picture is provided that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Preferably, between the step S1-S4 be constant work.Particularly, the network equipment obtains first search sequence of importing when the user carries out picture searching in subscriber equipment in real time in step S1; In step S2, described first search sequence is torn open speech and handle, to obtain one or more second search sequence; In step S3, carry out match search according to described second search sequence again, with the one or more pictures that obtain to be complementary with described second search sequence; In step S4, the network equipment 1 further offers described subscriber equipment with described one or more pictures.At this, it will be understood by those skilled in the art that " continuing " is meant that each step requires to carry out obtaining, tear open speech processing, picture searching and corresponding picture being provided of first search sequence according to the mode of operation of setting or adjust in real time respectively, stops by network operations such as subscriber equipment 2 execution picture searchings in a long time until this user.

Fig. 4 illustrate in accordance with a preferred embodiment of the present invention handle the method flow diagram carry out picture searching based on tearing speech open, wherein, step S1 ', S2 ', S303 ' and S4 ' and corresponding step S1 shown in Figure 3, S2, S3 are identical with S4, so locate to repeat no more, and mode by reference is contained in this.

The network equipment 1 offers described subscriber equipment 2 with described one or more second search sequence in step S301 ', and obtains second search sequence that this user selects from described one or more second search sequence in step S302 '; Then, in step S303 ', the network equipment 1 carries out match search according to this second user-selected search sequence again, with the one or more pictures that obtain to be complementary with this second search sequence.Particularly, tear speech open when first search sequence of 1 pair of user's input of the network equipment and handle, obtain one or more second search sequence, the network equipment 1 passes through page technology, as JSP, ASP, PHP technology, described one or more second search sequence are offered described subscriber equipment 2.Preferably, the network equipment 1 can be shown to the user with the form of described one or more second search sequence by the drop-down alternate item of the input frame of first search sequence, also can be shown to the user by the form of suspension frame.Those skilled in the art will be understood that the above-mentioned display format that one or more second search sequence are offered subscriber equipment 2 is only for for example; other display formats existing or that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.Afterwards, the user browses by browser software in the subscriber equipment 2 or client software, and by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, described one or more second search sequence of tearing speech open that contain are selected.With the mouse is example, when choosing or clicking certain second search sequence with mouse in the drop-down menu of user at input frame, the network equipment 1 is by submitting, receive forms such as request of access to detection cursor focus, reception list, obtain user's selection that this user carries out second search sequence, for subsequent treatment.After this, the network equipment 1 can also carry out match search according to the second user-selected search sequence, with the one or more pictures that obtain to be complementary with described second search sequence.For example, input first search sequence " Fan Bingbing " in the input frame of user at picture searching, the network equipment 1 as JSP, ASP, PHP technology, gets access to described first search sequence " Fan Bingbing " by page technology; Then, the network equipment 1 obtains corresponding second search sequence of this first search sequence " Fan Bingbing " " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " according to the coupling of tabling look-up by tearing the speech processing open; In the drop-down alternate item of this input frame, above-mentioned second search sequence is offered subscriber equipment, select for the user; If the user chooses or click selections " model ice ice stage photo ", the network equipment 1 is again by detecting cursor focus, receiving list and submit, receive forms such as request of access to, and that obtains this user is chosen as second search sequence " model ice is iced stage photo "; The network equipment 1 carries out match search according to the second user-selected search sequence " model ice ice stage photo ", obtains the one or more pictures that are complementary with described second search sequence " model ice ice stage photo "; Again by page technology, as JSP, ASP, PHP technology, the one or more pictures that will be above-mentioned be complementary with " model is iced stage photo " and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.

Preferably, in step S2, the network equipment 1 can also be torn speech open to described first search sequence and handle, to obtain one or more second search sequence according to tearing dictionary open.Particularly, after the network equipment 1 obtains first search sequence, in tearing dictionary open, first search sequence is carried out matching inquiry, described first search sequence is torn open speech handle, to obtain one or more second search sequence.For example, picture searching is carried out in user's input " Fan Bingbing ", and the network equipment 1 gets access to first search sequence " Fan Bingbing ".The network equipment 1 carries out matching inquiry according to " Fan Bingbing " in tearing dictionary open, inquiring second corresponding search sequence of " Fan Bingbing " is " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc.The network equipment 1 carries out match search according to described second search sequence, obtains the one or more pictures that are complementary with described second search sequence.The network equipment 1 is again by page technology, and as JSP, ASP, PHP technology, one or more pictures that will be complementary with described second search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.More preferably, after the network equipment 1 inquiry obtains second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc., forms such as the drop-down alternate item of described second search sequence by the input frame of first search sequence, suspension frame can be offered subscriber equipment 2, select for the user.Then, the user by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, described one or more second search sequence of tearing speech open that contain are selected.The network equipment 1 passes through to obtain user's selection that this user carries out second search sequence to detect cursor focus, to receive forms such as list is submitted to, reception request of access.The network equipment 1 carries out match search according to the second user-selected search sequence again, obtains the one or more pictures that are complementary with described second search sequence.The network equipment 1 passes through page technology, as JSP, ASP, PHP technology, one or more pictures that will be complementary with the second user-selected search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.And for example, picture searching is carried out in user's input " Fan Bingbing ", and the network equipment 1 gets access to first search sequence " Fan Bingbing ".The network equipment 1 carries out matching inquiry according to the category " star " under " Fan Bingbing " in tearing dictionary open, inquire that " star " is pairing to tear speech open " stage photo ", " description ", " living photo " etc. are arranged; Then, the network equipment 1 combines the above-mentioned speech of tearing open again with first search sequence " Fan Bingbing ", obtains second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " model ice ice living photo " etc.The network equipment 1 carries out match search according to described second search sequence, obtains the one or more pictures that are complementary with described second search sequence.The network equipment 1 is again by page technology, and as JSP, ASP, PHP technology, one or more pictures that will be complementary with described second search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.More preferably, to tear speech " stage photo ", " description ", " living photo " open at the network equipment 1 combines with first search sequence " Fan Bingbing ", obtain after second search sequence " model ice ice stage photo ", " Fan Bingbing description ", " the model ice ice living photo " etc., the network equipment 1 can offer subscriber equipment 2 with forms such as the drop-down alternate item of described second search sequence by the input frame of first search sequence, suspension frames, selects for the user.Then, the user by with the interactive mode of subscriber equipment 2, include but not limited to keyboard, mouse, telepilot, touch pad or hand-written equipment, described one or more second search sequence of tearing speech open that contain are selected.The network equipment 1 passes through to obtain user's selection that this user carries out second search sequence to detect cursor focus, to receive forms such as list is submitted to, reception request of access.The network equipment 1 carries out match search according to the second user-selected search sequence again, obtains the one or more pictures that are complementary with described second search sequence.The network equipment 1 passes through page technology, as JSP, ASP, PHP technology, one or more pictures that will be complementary with the second user-selected search sequence and/or its link and write one or more webpages, and these webpages and/or its link are offered subscriber equipment 2.Those skilled in the art will be understood that above-mentioned basis tears dictionary open and tear mode that speech handles open only for for example; other bases existing or that may occur are from now on torn dictionary open and are torn mode that speech handles open as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Preferably, in step S5 (not shown), the network equipment 1 can also be according to the statistical study to the employed picture searching keyword of a large number of users, sets up or upgrades the described dictionary of tearing open.Particularly, the network equipment 1 is according to the statistical study of the employed picture searching keyword of user in a period of time, as analyzing the picture searching keyword that obtains " Fan Bingbing " by statistics " Fan Bingbing description " (accounting for 33%) arranged, " model ice ice stage photo " (accounting for 31%), " model ice ice living photo " (accounting for 28%), " model ice is iced other " (accounting for 8%), and then first search sequence of determining to relate to " Fan Bingbing " can corresponding second search sequence " Fan Bingbing description ", " model ice ice stage photo ", " model ice ice living photo " etc., upgrade this and tear dictionary open, correspondingly add in the dictionary or the pairing speech processing record of tearing open of modification " Fan Bingbing " as tearing open at this.At this, the network equipment 1 both can independently have been finished the statistical study to the employed picture searching keyword of a large number of users, also can accept from other products or the third party device statistic analysis result to the employed picture searching keyword of a large number of users.For example, when the network equipment 1 is independently finished statistical study to the employed picture searching keyword of a large number of users, the employed picture searching keyword of user in 1 recording storage a period of time of the network equipment, then these picture searching keywords are carried out statistical study, for example draw the shared number percent of various picture searching keywords of corresponding same descriptor, get one or more second search sequence of the bigger picture retrieval keyword of wherein shared number percent as this descriptor correspondence, getting the picture searching keyword that frequency of usage surpasses certain threshold value for another example is one of second search sequence of corresponding descriptor.For example, when first user uses picture searching keyword " model ice ice stage photo ", the network equipment 1 can be noted it, and the access times of described picture searching keyword " model ice ice stage photo " are added up, when its number of times that is used surpasses preset threshold (for example 1000 times), this picture searching keyword is deposited in and tears dictionary open as tearing speech open.Thereafter, when picture searching was carried out in arbitrary user's input " Fan Bingbing ", the network equipment 1 can carry out match search as second search sequence with this picture searching keyword " model ice ice stage photo ", perhaps offers the user and selects for it.For another example, when a large number of users uses the form of " so-and-so star's stage photo " to carry out picture searching as first search sequence, the network equipment 1 can be with wherein picture searching keyword " stage photo " as tearing speech open, deposit in and tear dictionary open, when arbitrary user imports first search sequence that comprises " so-and-so star " printed words once more and carries out picture searching, the network equipment 1 can be torn this open speech " stage photo " and combine as second search sequence with " so-and-so star " and carry out match search, perhaps offers the user and selects for it.Preferably, the user can import first search sequence in search pictures, can obtain the used picture searching keyword of search pictures by it being carried out pre-service or tearing the speech processing open.For example, input first search sequence in the input frame of user at picture searching " model ice ice atlas sheet " is because the hunting zone of search engine itself is exactly a picture, therefore, remove " picture " and this time want information, the picture searching keyword that obtains described first search sequence is " Fan Bingbing ".And for example, when the user imported first search sequence " model ice ice stage photo ", " model ice ice stage photo " promptly became the picture searching keyword.Those skilled in the art be will be understood that it only is for example that the mode of dictionary is torn in above-mentioned renewal open; the mode that dictionary is torn in other renewals existing or that may occur from now on open is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

More preferably, in step S5 (not shown), the network equipment 1 can also be according to the statistical study that the employed picture searching keyword of a large number of users and pairing picture thereof are selected, and sets up or upgrades the described dictionary of tearing open.Particularly, the network equipment 1 can also be according to the statistical study of the picture of finally selecting for a large number of users to browse, for example correct corresponding picture searching keyword or derive corresponding picture searching keyword according to the selected picture of browsing according to the selected picture of browsing, upgrade this and tear dictionary open, for example tear open and correspondingly add in the dictionary or revise the pairing speech of tearing open of one or more descriptor and handle record at this.For example, when the user utilizes " Fan Bingbing " to carry out picture searching, according to the descriptors such as label, text message or keyword in this this user-selected picture, the picture that this user selects to browse is " stage photo " of " Fan Bingbing ", and the picture retrieval keyword of then revising this user is " model ice ice stage photo "; And,, upgrade this and tear dictionary open according to statistical study to the employed picture searching keyword of a large number of users according to last embodiment (basic identical because of embodiment, as to repeat no more, and be contained in this by reference) at this.And for example, in the picture of relevant " Fan Bingbing " that offer the user, by extraction to picture tag, text message or keyword, what determine a large number of users selection is the picture of " stage photo " of relevant " Fan Bingbing ", the network equipment 1 can deposit " model ice ice stage photo " in and tear dictionary open as second search sequence of correspondence " Fan Bingbing ", to upgrade the described dictionary of tearing open.Those skilled in the art be will be understood that it only is for example that the mode of dictionary is torn in above-mentioned renewal open; the mode that dictionary is torn in other renewals existing or that may occur from now on open is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.At this, the network equipment 1 both can independently have been finished the statistical study that the employed picture searching keyword of a large number of users and pairing picture thereof are selected, and also can accept the statistic analysis result of the employed picture searching keyword of a large number of users and pairing picture thereof being selected from other products or third party device.For example, when the network equipment 1 is independently finished the statistical study that the employed picture searching keyword of a large number of users and pairing picture thereof are selected, employed picture searching keyword of user and pairing picture thereof are selected in 1 recording storage a period of time of the network equipment, then statistical study selected to carry out in these picture searching keywords and pairing picture thereof, the various picture searching keywords that for example draw corresponding same descriptor (comprise the picture searching keyword after correcting according to the picture of selecting to browse, picture searching keyword according to the picture derivation of selecting to browse) shared number percent, get one or more second search sequence of the bigger picture retrieval keyword of wherein shared number percent, get the picture searching keyword that frequency of usage surpasses certain threshold value for another example and (comprise the picture searching keyword after correcting according to the picture of selecting to browse as this descriptor correspondence, picture searching keyword according to the picture derivation of selecting to browse) be one of second search sequence of corresponding descriptor.

Preferably, in step S3, the network equipment 1 can also be according to described second search sequence, and in conjunction with this user's user related information, carry out match search, with the one or more pictures that obtain to be complementary with described second search sequence and described user related information.Particularly, the network equipment 1 is according to handling second search sequence that obtains through tearing speech open, or by user-selected when tearing speech open and handle second search sequence that obtains, in conjunction with the user related information network equipment 1 record or that this user sets, carry out match search, with the one or more pictures that obtain to be complementary with described second search sequence and described user related information.For example, if the preference options that the user sets is an automobile, then when second search sequence is " jaguar ", the network equipment 1 is according to second search sequence " jaguar ", and, carry out match search in conjunction with this user's preference options " automobile ", obtain the picture of relevant " jaguar automobile ", the network equipment 1 offers subscriber equipment with the picture of described relevant " jaguar automobile " again, selects to browse for the user.And for example, if the preference options that the user sets is an animal, then when second search sequence is " jaguar ", the network equipment 1 is according to second search sequence " jaguar ", and, carry out match search in conjunction with this user's preference options " animal ", obtain the picture of relevant " jaguar animal ", the network equipment 1 offers subscriber equipment with the picture of described relevant " jaguar automobile " again, selects to browse for the user.

Preferably, in step S6 (not shown), the network equipment 1 can also sort to described one or more pictures according to this user's user related information; After this, in step S4, the network equipment 1 can also offer described subscriber equipment with the described one or more pictures after the ordering.Particularly, carry out according to second search sequence after match search obtains one or more pictures at the network equipment 1, the network equipment 1 is according to user related information, for example, according to label, text message or the keyword of the picture that extracts and the matching degree of user related information, described one or more pictures are sorted; The network equipment 1, writes one or more webpages with described one or more pictures and/or its link after the ordering, and these webpages and/or its link is offered subscriber equipment 2 as JSP, ASP, PHP technology again by page technology.Connect example, if the preference options that the user sets is an automobile, then when second search sequence was " jaguar ", the network equipment 1 carried out match search according to second search sequence " jaguar ", obtained the picture of relevant " jaguar "; The preference options that the network equipment 1 is set according to the user, i.e. " automobile ", the picture of prioritization relevant " jaguar automobile "; Picture with described relevant " jaguar automobile " offers subscriber equipment again, selects to browse for the user.And for example, if the preference options that the user sets is an animal, then when second search sequence was " jaguar ", the network equipment 1 carried out match search according to second search sequence " jaguar ", obtained the picture of relevant " jaguar "; The preference options that the network equipment 1 is set according to the user, i.e. " automobile ", the picture of prioritization relevant " jaguar animal "; In proper order the picture of described relevant " jaguar animal " is offered subscriber equipment according to this prioritization again, select to browse for the user.Those skilled in the art be will be understood that the mode of above-mentioned picture ordering is only for giving an example; the mode of other pictures existing or that may occur from now on orderings is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

More preferably, in step S6 (not shown), the network equipment 1 can also sort to described one or more second search sequence according to this user's user related information; After this, in step S3, the network equipment 1 can also offer described subscriber equipment with described one or more second search sequence after the ordering, and obtain second search sequence that this user selects from described one or more second search sequence, carry out match search according to second search sequence that described user selects, with the one or more pictures that obtain to be complementary with described second search sequence.Particularly, for handling one or more second search sequence that obtain through tearing speech open, the network equipment 1 can sort to described one or more second search sequence according to this user's relevant information; Drop-down alternate item or the form such as suspension frame of described one or more search sequence after the network equipment 1 will sort again by the input frame of for example first search sequence offers described subscriber equipment 2; Then, the network equipment 1 by to detect cursor focus, to receive forms such as list is submitted to, reception request of access, obtains second search sequence that this user selects again from described one or more second search sequence.Preferably, the drop-down alternate item of this input frame or suspension frame can also be according to the adaptive adjustment of page size, for example, for screen constrained portable terminal, drop-down menu or suspension frame can diminish adaptively, and a demonstration is positioned at part second search sequence in the prostatitis of described second search sequence after the ordering.Those skilled in the art be will be understood that the mode of above-mentioned second search sequence ordering is only for giving an example; the mode of other orderings existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

Preferably, described user related information includes but not limited to following at least one: a) described user's personal attribute; B) described user's input historical record; C) preference options of described user's setting; D) address information of described subscriber equipment.The network equipment 1 can be according to user's the interested picture of personal attribute's (for example user's age, sex, occupation, education background etc.) match search user possibility, or know the user by inference may interested second search sequence or picture, described second search sequence or picture are stood out, and the confession user selects or browses.For example,, then when picture retrieval is carried out in its input " jaguar ", the picture of relevant " jaguar automobile " is stood out, be shown to the user if the network equipment 1 this user of record is the male sex.The network equipment 1 can also be according to the historical record of user's input, the match search user is the picture of interest just, or know possible interested second search sequence of user or picture by inference, so that described second search sequence or picture are stood out, the confession user selects or browses.For example, the user once imported " A Fanda " and carried out picture searching, and so, when this user imported " film " again and carries out picture searching, the figure sector-meeting relevant with " A Fanda " stood out, and is shown to this user.The network equipment 1 can also be according to the preference options (for example " film ", " TV ", " stage photo ", " description ", " living photo " etc.) of user's setting, the interested picture of match search user possibility, or know the user by inference may interested second search sequence or picture, so that described second search sequence or picture are stood out, the confession user selects or browses.For example, if the preference that the user sets is a still, then when user's input " Fan Bingbing ", in second search sequence, standing out in " model ice ice stage photo apple ", " model ice ice stage photo besieged city in October " etc., selects for the user, in the picture that returns, pictures such as relevant " model ice ice stage photo apple ", " model ice ice stage photo besieged city in October " are stood out, and select to browse for the user; If user's preference is the TV play stage photo, then when user's input " Fan Bingbing ", in second search sequence, " model ice ice stage photo also pearl sound of laughing ", " model ice ice stage photo kermes snow " etc. are stood out, select for the user, in the picture that returns, relevant " model ice ice stage photo also pearl sound of laughing ", " model ice ice stage photo kermes snow " wait picture to stand out, and browse for user's selection.The network equipment 1 can also be according to the address information (for example user device IP address, GPS information etc.) of subscriber equipment, the interested picture of match search user possibility, or know the user by inference may interested second search sequence or picture, so that described second search sequence or picture are stood out, the confession user selects or browses.For example, the network equipment 1 detects the current IP address of subscriber equipment 2 in Shanghai, then for first search sequence of obtaining " Nanjing Road ", the network equipment 1 can be torn speech open and handle, obtain for example second search sequence such as " Nanjing Road discount store ", " Nanjing Road star's activity ", the network equipment 1 can be searched for according to above-mentioned second search sequence, perhaps the network equipment 1 is stood out above-mentioned second search sequence, select for the user, the network equipment 1 can be stood out above-mentioned relevant picture, offer subscriber equipment 2, browse selection for the user.And for example, the current IP address that detects subscriber equipment 2 when the network equipment 1 is not during in Shanghai, can know the travel information that it wants to understand the Nanjing Road, Shanghai by inference, the network equipment 1 can be torn speech open and handle, obtain for example second search sequence such as " Nanjing Road map ", " Nanjing Road characteristic snack ", search for according to above-mentioned second search sequence again, perhaps the network equipment 1 is stood out above-mentioned second search sequence, select for the user, the network equipment 1 can be stood out above-mentioned relevant picture, offer subscriber equipment, browse selection for the user.Those skilled in the art be will be understood that above-mentioned user related information only for giving an example, and other user related informations existing or that may occur from now on also should be included in the protection domain of the present invention as applicable to the present invention, and are contained in this at this with way of reference.

Preferably, in step S7 (not shown), described user related information is upgraded in the selection that the network equipment 1 can also be done described one or more picture according to described second search sequence and/or this user.For example, the network equipment 1 can upgrade this user's user related informations such as input historical record according to user-selected second search sequence.For example, when the user in " Fan Bingbing " corresponding second search sequence or picture, selected " Fan Bingbing description " or belonged to the picture of " Fan Bingbing description ", so, 1 of the network equipment correspondingly upgrades uses relevant information, as increasing historical record such as " Fan Bingbing description " input.

Preferably, in step S2, the network equipment 1 can also carry out pre-service to described first search sequence; Tear pretreated described first search sequence open speech and handle, to obtain one or more second search sequence.Particularly, in first search sequence of user's input, may comprise some less important information, as " ", " picture " etc., perhaps described first search sequence is a long sentence, therefore, the network equipment 1 need carry out pre-service to these first search sequence, handles with convenient speech that this is torn open.Then, the network equipment 1 is according to pretreated first search sequence, by matching inquiry, for example by at the descriptor split table or tear modes such as carrying out match query in the dictionary open, described first search sequence is torn open speech handle, to obtain one or more second search sequence.

More preferably, described pre-service comprises following at least one: a) synonym is replaced; B) classification is handled; C) word segmentation processing.For example, when the user imported first search sequence " little S ", the network equipment 1 can carry out pre-service to this, " little S " synonym is replaced with " Xu Xidi ", handled further this is torn open speech.And for example, first search sequence " Fan Bingbing " of user's input belongs to " performer ", and " performer's picture " can be split as " performer's stage photo ", " performer's description ", " performer's living photo " etc., earlier this is classified to handle and further tear speech again open and handle, promptly obtain second search sequence " Fan Bingbing description ", " model ice ice stage photo " and " model is iced living photo " etc. of its correspondence according to first search sequence " Fan Bingbing ".For another example, in first search sequence of user input, may comprise some less important information (for example " "), can carry out word segmentation processing to this first search sequence, for example earlier these less important information are removed, perhaps long sentence is made pauses in reading unpunctuated ancient writings, remove and repeat vocabulary, therefrom extract descriptor again, handle further this is torn open speech.Those skilled in the art be will be understood that above-mentioned first search sequence is carried out pretreated mode only for for example; other pretreated modes existing or that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference at this.

To those skilled in the art, obviously the invention is not restricted to the details of above-mentioned one exemplary embodiment, and under the situation that does not deviate from spirit of the present invention or essential characteristic, can realize the present invention with other concrete form.Therefore, no matter from which point, all should regard embodiment as exemplary, and be nonrestrictive, scope of the present invention is limited by claims rather than above-mentioned explanation, therefore is intended to be included in the present invention dropping on the implication that is equal to important document of claim and all changes in the scope.Any Reference numeral in the claim should be considered as limit related claim.In addition, obviously other unit or step do not got rid of in " comprising " speech, and odd number is not got rid of plural number.A plurality of unit of stating in the device claim or device also can be realized by software or hardware by a unit or device.The first, the second word such as grade is used for representing title, and does not represent any specific order.

Claims

1. handle the method carry out picture searching based on tearing speech open for one kind, this method may further comprise the steps:

A obtains first search sequence of importing when the user carries out picture searching in subscriber equipment;

B tears speech open to described first search sequence and handles, to obtain one or more second search sequence;

C carries out match search according to described second search sequence, with the one or more pictures that obtain to be complementary with described second search sequence;

D offers described subscriber equipment with described one or more pictures.

2. method according to claim 1, wherein, described step c also comprises:

-described one or more second search sequence are offered described subscriber equipment;

-obtain second search sequence that this user selects from described one or more second search sequence;

-carry out match search according to second search sequence of described user's selection, with the one or more pictures that obtain to be complementary with described second search sequence.

3. method according to claim 1 and 2, wherein, described step b also comprises:

-according to tearing dictionary open, described first search sequence is torn open speech handle, to obtain one or more second search sequence.

4. method according to claim 3, wherein, this method also comprises:

E is according to the statistical study of the employed picture searching keyword of a large number of users, sets up or upgrades the described dictionary of tearing open.

5. method according to claim 4, wherein, described step e also comprises:

-according to the statistical study that the employed picture searching keyword of a large number of users and pairing picture thereof are selected, set up or upgrade the described dictionary of tearing open.

6. according to each described method in the claim 1 to 5, wherein, described step c also comprises:

-according to described second search sequence, and, carry out match search in conjunction with this user's user related information, with the one or more pictures that obtain to be complementary with described second search sequence and described user related information.

7. according to each described method in the claim 1 to 6, wherein, this method also comprises:

F sorts to described one or more pictures according to this user's user related information;

Wherein, described steps d also comprises:

-described one or more pictures after will sorting offer described subscriber equipment.

8. method according to claim 7, wherein, described step f also comprises:

-according to this user's user related information, described one or more second search sequence are sorted;

Wherein, described step c also comprises:

-described one or more second search sequence after will sorting offer described subscriber equipment;

-obtain this user second search sequence of selecting from described one or more second search sequence after the ordering;

9. according to each described method in the claim 6 to 8, wherein, described user related information comprises following at least one:

-described user's personal attribute;

-described user's input historical record;

The preference options of-described user's setting;

The address information of-described subscriber equipment.

10. according to each described method in the claim 1 to 9, wherein, this method also comprises:

G upgrades described user related information according to the selection that described second search sequence and/or this user do described one or more picture.

11. according to each described method in the claim 1 to 10, wherein, described step b also comprises:

-described first search sequence is carried out pre-service;

-tear pretreated described first search sequence open speech to handle, to obtain one or more second search sequence.

12. method according to claim 11, wherein, described pre-service comprises following at least one:

-synonym is replaced;

-classification is handled;

-word segmentation processing.

13. handle the equipment carry out picture searching based on tearing speech open for one kind, this equipment comprises:

14. equipment according to claim 13, wherein, this searcher also comprises:

Interactive device is used for:

Wherein, described searcher also is used for:

15. according to claim 13 or 14 described equipment, wherein, the described speech device of tearing open also is used for:

16. equipment according to claim 15, wherein, this equipment also comprises:

First updating device is used for according to the statistical study of the employed picture searching keyword of a large number of users, sets up or upgrades the described dictionary of tearing open.

17. method according to claim 16, wherein, described first updating device also is used for:

18. according to each described equipment in the claim 13 to 17, wherein, described searcher also is used for:

19. according to each described equipment in the claim 13 to 18, wherein, this equipment also comprises:

Collator is used for the user related information according to this user, and described one or more pictures are sorted;

Wherein, described generator also is used for:

20. equipment according to claim 19, wherein, described collator also is used for:

Wherein, described searcher also is used for:

21. according to claim 18 or 20 described equipment, wherein, described user related information comprises following at least one:

-described user's personal attribute;

-described user's input historical record;

The preference options of-described user's setting;

The address information of-described subscriber equipment.

22. according to each described equipment in the claim 13 to 21, wherein, this equipment also comprises:

Second updating device is used for the selection described one or more picture done according to described second search sequence and/or this user, upgrades described user related information.

23. according to each described equipment in the claim 13 to 22, wherein, the described speech device of tearing open also is used for:

-described first search sequence is carried out pre-service;

24. equipment according to claim 23, wherein, described pre-service comprises following at least one:

-synonym is replaced;

-classification is handled;

-word segmentation processing.

25. according to each described equipment in the claim 13 to 24, wherein, described equipment comprises network host, single network server, a plurality of webserver collection or based on the set of computers of cloud computing.