CN102231149A - Method for searching visual information of mobile phone based on local characteristics - Google Patents
Method for searching visual information of mobile phone based on local characteristics Download PDFInfo
- Publication number
- CN102231149A CN102231149A CN201010599974.8A CN201010599974A CN102231149A CN 102231149 A CN102231149 A CN 102231149A CN 201010599974 A CN201010599974 A CN 201010599974A CN 102231149 A CN102231149 A CN 102231149A
- Authority
- CN
- China
- Prior art keywords
- visual information
- mobile phone
- sample
- local feature
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a method for searching visual information of a mobile phone based on local characteristics. The method is characterized by comprising the following steps of: selecting and determining a visual information sample in an operation interface of the mobile phone; extracting and re-processing local information of the visual information sample; extracting unchanged regional characteristics and invariant vectors; establishing a characteristic model and/or index and searching; searching one or more results having the highest correlation degree with the characteristic model and an index of the visual information sample; and taking one or more results as visual information target results. According to the method, specific contents of pictures, images and videos can be searched accurately and can also be adjusted manually; the contents are in according with contents of pictures and videos which users really want to search; meanwhile, results and literal contents wanted by a plurality of users can be returned; and the method is easy to realize and operate.
Description
Technical field
The present invention relates to the search technique field, specially refer to a kind of mobile phone visual information searching method based on local feature.
Background technology
Along with broadband network advances and application, service that is provided and application are more and more, wherein search service is one of most important service, search also has number of different types and application, for example except keyword search commonly used, also have picture searching, music searching etc., picture searching can pass through search utility, be provided at picture information relevant on network or database, the particularly internet to the user, at present, in multi-media network, have two class image search engine technology comparatively commonly used: the one, adopt traditional image retrieval technologies based on key word; Next is exactly to adopt the CBIR technology.
Use traditional search engine based on key word to carry out picture search, its principle is the same with the search general information, and difference is the keyword difference of search just, and perhaps the class categories difference comprises picture, image, photo.Traditional image specialized search engine principle of work also is like this, its general course of work roughly was divided into for three steps, at first send the searching machine people that picture in the online website and page info are searched for automatically, and document file page and pictorial information are returned search engine database.The information that search engine is returned is processed then, and extracting keywords, image low-level feature or process classification carry out the index of image information in the database.Accept user's query requests at last by the Web page access, provide user's demand data according to correlativity.The traditional images search engine roughly has four parts to constitute: searching machine people, search controller, image search engine database and user inquiring service interface.Searching machine people has the function of image in the discriminance analysis document in internet roaming, and image is classified.Search controller is the core of whole search engine, coordinates the work of each robot, analyzes the webpage that obtains automatically, and adds the view data after analyzing to the image search engine database.The image search engine database is used to deposit the webpage that collects, the descriptor of extraction, and the ephemeral data that is used for the robot searches state recording, comprises the full content of webpage, perhaps the address of image, title, keyword, summary etc.The user inquiring service interface is used for accepting user's query requests, and the result for retrieval image is returned to the user by the degree of correlation, generally includes keyword subject retrieval and split catalog search method.
And content-based image search engine principle of work is the picture search instrument that is based upon on the CBIR technology, can set up index information for the image of on Web, browsing, can carry out graphical analysis and differentiation, annotated for image, the index information that storage extracts is also set up index database.The CBIR technology is one of retrieval technique of comparatively popular at present and future development trend.It needs user's participation hardly when characteristic matching, and mainly utilizes the feature (as color, texture, shape etc.) of image self to finish, and has stronger objectivity and intuitive.Utilize this characteristics, can be used for extracting the feature of all images file in all kinds of image libraries.In retrieving, the user is general only need to provide an example image, system at first extracts the feature of this example image, then according to the similarity matching algorithm of characteristics of image, compare with corresponding feature in the image library, at last the image similar to exemplary characteristics returned to the user, to reach a complete image retrieval process.
But this method also has deficiency, and is slow as travelling speed, makes mistakes easily, more undesirable for the result of local feature search, that is to say, also has deficiency in the middle of present visual information search, particularly is its service and application in broadband network.
Summary of the invention
Purpose of the present invention is exactly in order to overcome the above problems, and a kind of method that just can accurately not carry out the visual information search by literal fast is provided.
The present invention realizes by such method:
A kind of mobile phone visual information searching method based on local feature, main points be, selects and determine the visual information sample at the operation interface of mobile phone; The visual information sample is carried out local message to be extracted; The local message that is extracted is handled again; Extract the invariant region feature, and generate the invariant vector of visual information; Set up characteristic model with or index; This characteristic model or index are transmitted in network or the network data base line search of going forward side by side; Search with the highest one or more results of the characteristic model of this visual information sample and the index degree of correlation as the visual information objective result; The above-mentioned objective result that searches is sorted and be illustrated in the user mobile phone interface.
Further, described processing again to local message is based on the characteristic processing and the detection method of the invariant technology of metric space, is image to be carried out the image local feature that balance, rotation, convergent-divergent, affined transformation maintain the invariance handle.
Further, selecting and determine the visual information sample at the operation interface of mobile phone, is to obtain wherein part or all from visual interface, as the visual information sample.
Further, the described some of choosing is at least a as foundation with in color characteristic, textural characteristics, the shape facility information in picture, the characteristics of image from the visual information sample, and as the visual information sample.
Further, the described processing again comprises at least and adds the wherein at least a contents processing of making an uproar, be out of shape, shear, insert content, rotation, convergent-divergent and brightness increase and decrease, obfuscation etc.
Further, with the target visual information extraction related text information that is obtained, and send to user's visual interface, the user also can select not show.
Another kind method is: in preceding described processing again to local message, be that one of them is adjusted at least to local message features such as the color characteristic in the image information, textural characteristics, shape facilities.
Further, the adjustment of above-mentioned visual information local feature is based on existing template or sample carry out; Or carry out manually freely adjusting and handling, described adjustment and processing are to be similar to manipulation of image data such as combination, desalination, reinforcement, splicing, hollow out, blanking, filling, increase.
Further, as the visual information sample, or at least one part is as the visual information sample with the information after adjust handling.
Equally, will extract visual information again in the target visual information that be obtained,, search for once more, target is approached one by one as the reference visual information sample of searching for once more, satisfied up to the user.
Like this, the present invention has following advantage and effect:
This mobile phone visual information searching method based on local feature, can accurately search at the particular content of picture, image and video, and can manually adjust, the visual information content of searching for out based on this mode, need not to be associated, and want really that with the user picture video content of searching for is consistent, can return the result that a plurality of users want simultaneously with the content of the key word of picture, video, and word content, realize easily and operate.
Description of drawings
Fig. 1 is the theory diagram of mobile phone visual information searching method based on local feature proposed by the invention.
Embodiment
Below in conjunction with accompanying drawing system and method for the present invention is further detailed.
For realizing purpose of the present invention, it is that such method realizes that the present invention adopts:
What this method involved in the present invention related at present embodiment is mobile phone, but also can be the information equipment that computer, intelligent television etc. have display screen and operating system software, promptly can be used for multiple device and handheld device.
Below by embodiment method of the present invention is described.
Embodiment one:
Smart mobile phone is very common now, this mobile phone not only can be retrieved and be inquired about by document, also can retrieve and simple the processing picture, for example the user need find more information about this person by a width of cloth personage image or picture, so just can select a width of cloth this person head portrait at the operation interface of mobile phone, and determine as the visual information sample, this sample may be lack of standardization, perhaps and not in full conformity with user's requirement, can carry out local message to this sample and extract, as the first feature of selecting a personage and extracting face from image or picture; The local message that is extracted is handled again, just the face feature is handled again again; Become extraction invariant region feature, and generated the invariant vector of visual information; And with this set up the face characteristic model with or index; After this face characteristic model or index generate, can be transmitted in network or the network data base line search of going forward side by side, this database comprises local data base or network data base, in database, search with the highest one or more results of the characteristic model of this visual information sample and the index degree of correlation as the visual information objective result; At last the above-mentioned objective result that searches is sorted and be illustrated in the user mobile phone interface, so just finished the process of a basic search pictures or image.
Described here processing again to local message, be based on the characteristic processing and the detection method of the invariant technology of metric space, this is a comparatively objectively method of a kind of Search Results, can carry out the image local feature that balance, rotation, convergent-divergent, affined transformation maintain the invariance to image and handle, and can keep original picture characteristics constant substantially.
In addition, it is emphasized that, select and determine the visual information sample at the operation interface of mobile phone, be from visual interface, to obtain wherein part or all, as the visual information sample, particularly select the part of entire image, picture to be used as sample, have more practicality, search is got up also more quick.
The described here some of choosing is wherein at least a as foundation with in color characteristic, textural characteristics, the shape facility information in picture, the characteristics of image from the visual information sample, and as the visual information sample.
To handling again of image also can be to comprise at least adding the wherein at least a contents processing of making an uproar, be out of shape, shear, insert content, rotation, convergent-divergent and brightness increase and decrease, obfuscation etc., this is important measures of taking at the difference between the target sample, like this, even target image seems and sample image between differ greatly, but, just can find target image as long as the key feature difference is little.
Last the important point, exactly in the target visual information that is obtained, can extract corresponding therewith related text information, and send to user's visual interface, the user also can select not show or hide, this is just at present with in the near future, for visual informations such as picture, images, Xiang Guan word content necessarily has much therewith, and all have close getting in touch with image, picture, therefore, when finding target image, picture, also Word message can be displayed, provide more detailed content to the user.
Embodiment two:
This embodiment and last embodiment are basic identical, difference is the processing again to described local message, be to the color characteristic in the image information, textural characteristics, at least one of them adjusts local message features such as shape facility, the adjustment of above-mentioned visual information local feature is based on existing template or sample carry out, the gang form that one two width of cloth picture is for example arranged, and the splicing template of certain part in two pictures, and to filtering profile of a requirement etc., in these templates, have some specific new elements, as the color characteristic template, the textural characteristics template, shape feature template etc., just had more new element as picture or image, come down on original basis, to have generated new picture through template; Another one way or can carry out manually freely adjusting and handling, described adjustment and processing are to be similar to manipulation of image data such as combination, desalination, reinforcement, splicing, hollow out, blanking, filling, increase.That is to say that the adjustment dynamics of this mode is bigger, may be two different pictures or image are made up or to splice, and has generated similar visual informations such as a brand-new image, picture.
Like this, the information after handling will adjusted again is as the visual information sample, or at least one part is as the visual information sample.
Simultaneously, in two above-mentioned embodiment, can will extract visual information again in the target visual information that be obtained,, search for once more, target is approached one by one as the reference visual information sample of searching for once more, satisfied up to the user.
By above-mentioned method and step, promptly can be accurately at the particular content of visual information, and the visual information of searching for out based on this mode, need not to be associated with the content of the key word of visual information, and want really that with the user visual information true content of searching for is consistent, simultaneously can return the result that a plurality of users want, realize easily and operate.
Though a preferred embodiment of the present invention only has been described here, its meaning be not limit the scope of the invention, applicability and configuration.On the contrary, the detailed description to embodiment can make those skilled in the art be implemented.Will be understood that not departing under the definite spirit and scope of the invention situation of appended claims, can make suitable change and modification some details.
Claims (10)
1. mobile phone visual information searching method based on local feature is characterized in that:
Select and determine the visual information sample at the operation interface of mobile phone;
The visual information sample is carried out local message to be extracted;
The local message that is extracted is handled again;
Extract the invariant region feature, and generate the invariant vector of visual information;
Set up characteristic model with or index;
This characteristic model or index are transmitted in network or the network data base line search of going forward side by side;
Search with the highest one or more results of the characteristic model of this visual information sample and the index degree of correlation as the visual information objective result;
The above-mentioned objective result that searches is sorted and be illustrated in the user mobile phone interface.
2. the mobile phone visual information searching method based on local feature according to claim 1, it is characterized in that: described processing again local message, being based on the characteristic processing and the detection method of the invariant technology of metric space, is image to be carried out the image local feature that balance, rotation, convergent-divergent, affined transformation maintain the invariance handle.
3. the mobile phone visual information searching method based on local feature according to claim 1, it is characterized in that: select and determine the visual information sample at the operation interface of mobile phone, be from visual interface, to obtain wherein part or all, as the visual information sample.
4. the mobile phone visual information searching method based on local feature according to claim 3, it is characterized in that: the described some of choosing, be at least a with in color characteristic, textural characteristics, the shape facility information in picture, the characteristics of image from the visual information sample as foundation, and as the visual information sample.
5. the mobile phone visual information searching method based on local feature according to claim 1 is characterized in that: the described processing again comprises at least and adds the wherein at least a contents processing of making an uproar, be out of shape, shear, insert content, rotation, convergent-divergent and brightness increase and decrease, obfuscation etc.
6. the mobile phone visual information searching method based on local feature according to claim 1 is characterized in that: with the target visual information extraction related text information that is obtained, and send to user's visual interface, the user also can select not show.
7. the mobile phone visual information searching method based on local feature according to claim 1, it is characterized in that: described processing again to local message is that one of them is adjusted at least to local message features such as the color characteristic in the image information, textural characteristics, shape facilities.
8. the mobile phone visual information searching method based on local feature according to claim 7 is characterized in that: the adjustment of above-mentioned visual information local feature is based on existing template or sample carry out; Or carry out manually freely adjusting and handling, described adjustment and processing are to be similar to manipulation of image data such as combination, desalination, reinforcement, splicing, hollow out, blanking, filling, increase.
9. the mobile phone visual information searching method based on local feature according to claim 7 is characterized in that: the information after handling will adjusted is as the visual information sample, or at least one part is as the visual information sample.
10. the mobile phone visual information searching method based on local feature according to claim 1, it is characterized in that: will extract visual information again in the target visual information that be obtained, as the reference visual information sample of searching for once more, search for once more, target is approached one by one, satisfied up to the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010599974.8A CN102231149A (en) | 2010-12-22 | 2010-12-22 | Method for searching visual information of mobile phone based on local characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010599974.8A CN102231149A (en) | 2010-12-22 | 2010-12-22 | Method for searching visual information of mobile phone based on local characteristics |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102231149A true CN102231149A (en) | 2011-11-02 |
Family
ID=44843713
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010599974.8A Pending CN102231149A (en) | 2010-12-22 | 2010-12-22 | Method for searching visual information of mobile phone based on local characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102231149A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103064981A (en) * | 2013-01-18 | 2013-04-24 | 浪潮电子信息产业股份有限公司 | Method for searching images on basis of cloud computing |
| CN103164504A (en) * | 2012-09-07 | 2013-06-19 | 深圳市金立通信设备有限公司 | Smartphone refined picture searching system and method |
| CN103488658A (en) * | 2012-06-15 | 2014-01-01 | 李和平 | Content researching method and system based on response of characteristic value information of image frame data |
| CN104317827A (en) * | 2014-10-09 | 2015-01-28 | 深圳码隆科技有限公司 | Picture navigation method of commodity |
| WO2015109992A1 (en) * | 2014-01-22 | 2015-07-30 | Tencent Technology (Shenzhen) Company Limited | Information retrieval method, apparatus and system |
| CN105138633A (en) * | 2015-08-21 | 2015-12-09 | 成都秋雷科技有限责任公司 | Webpage retrieval method |
| CN116012924A (en) * | 2023-01-30 | 2023-04-25 | 人民网股份有限公司 | Face gallery construction method and device and computing equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1723455A (en) * | 2002-11-15 | 2006-01-18 | 皇家飞利浦电子股份有限公司 | Content retrieval based on semantic association |
| JP2010009332A (en) * | 2008-06-27 | 2010-01-14 | Denso It Laboratory Inc | Image-retrieving database device, image-retrieving database management method and program |
| CN101777064A (en) * | 2009-01-12 | 2010-07-14 | 鸿富锦精密工业（深圳）有限公司 | Image searching system and method |
-
2010
- 2010-12-22 CN CN201010599974.8A patent/CN102231149A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1723455A (en) * | 2002-11-15 | 2006-01-18 | 皇家飞利浦电子股份有限公司 | Content retrieval based on semantic association |
| JP2010009332A (en) * | 2008-06-27 | 2010-01-14 | Denso It Laboratory Inc | Image-retrieving database device, image-retrieving database management method and program |
| CN101777064A (en) * | 2009-01-12 | 2010-07-14 | 鸿富锦精密工业（深圳）有限公司 | Image searching system and method |
Non-Patent Citations (1)
| Title |
|---|
| 《计算机应用研究》 20080229 吴锐航等 基于SIFT特征的图像检索 第478-481页 1-10 第25卷, 第2期 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103488658A (en) * | 2012-06-15 | 2014-01-01 | 李和平 | Content researching method and system based on response of characteristic value information of image frame data |
| CN103164504A (en) * | 2012-09-07 | 2013-06-19 | 深圳市金立通信设备有限公司 | Smartphone refined picture searching system and method |
| CN103064981A (en) * | 2013-01-18 | 2013-04-24 | 浪潮电子信息产业股份有限公司 | Method for searching images on basis of cloud computing |
| WO2015109992A1 (en) * | 2014-01-22 | 2015-07-30 | Tencent Technology (Shenzhen) Company Limited | Information retrieval method, apparatus and system |
| CN104317827A (en) * | 2014-10-09 | 2015-01-28 | 深圳码隆科技有限公司 | Picture navigation method of commodity |
| CN104317827B (en) * | 2014-10-09 | 2017-06-30 | 深圳码隆科技有限公司 | A kind of picture air navigation aid of commodity |
| CN105138633A (en) * | 2015-08-21 | 2015-12-09 | 成都秋雷科技有限责任公司 | Webpage retrieval method |
| CN116012924A (en) * | 2023-01-30 | 2023-04-25 | 人民网股份有限公司 | Face gallery construction method and device and computing equipment |
| CN116012924B (en) * | 2023-01-30 | 2023-06-27 | 人民网股份有限公司 | Face gallery construction method and device and computing equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| AU2009357597B2 (en) | Methods and apparatuses for facilitating content-based image retrieval | |
| KR101289085B1 (en) | Images searching system based on object and method thereof | |
| Girod et al. | Mobile visual search: Architectures, technologies, and the emerging MPEG standard | |
| CN107562939B (en) | Vertical domain news recommendation method and device and readable storage medium | |
| CN102231149A (en) | Method for searching visual information of mobile phone based on local characteristics | |
| US20140254942A1 (en) | Systems and methods for obtaining information based on an image | |
| US20080270378A1 (en) | Method, Apparatus and Computer Program Product for Determining Relevance and/or Ambiguity in a Search System | |
| CN107229741B (en) | Information searching method, device, equipment and storage medium | |
| CN109902198A (en) | A kind of method, apparatus and application system to scheme to search figure | |
| JP2002288219A (en) | Information search system and method | |
| Li et al. | Interactive multimodal visual search on mobile device | |
| US11709881B2 (en) | Visual menu | |
| US9305215B2 (en) | Apparatus, method and computer readable recording medium for analyzing video using image captured from video | |
| CN102236714A (en) | Extensible markup language (XML)-based interactive application multimedia information retrieval method | |
| CN103268345A (en) | Method and device for retrieving film and television data | |
| US20100228751A1 (en) | Method and system for retrieving ucc image based on region of interest | |
| WO2010071617A1 (en) | Method and apparatus for performing image processing | |
| Lokoč et al. | VIRET tool meets NasNet | |
| CN110765305A (en) | Medium information pushing system and visual feature-based image-text retrieval method thereof | |
| JP5772908B2 (en) | Information processing apparatus, information processing system, control method thereof, and program | |
| JP2007102362A (en) | Automatic classification category forming device, automatic digital image content classification device and digital image content management system | |
| CN102184259A (en) | System for searching teaching video contents and method thereof | |
| KR20150101846A (en) | Image classification service system based on a sketch user equipment, service equipment, service method based on sketch and computer readable medium having computer program recorded therefor | |
| Çalışır et al. | Mobile multi-view object image search | |
| KR101896177B1 (en) | An Image Searching System Providing Multiplex Result |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20111102 |