CN101571875A - Realization method of image searching system based on image recognition - Google Patents
Realization method of image searching system based on image recognition Download PDFInfo
- Publication number
- CN101571875A CN101571875A CNA2009101071569A CN200910107156A CN101571875A CN 101571875 A CN101571875 A CN 101571875A CN A2009101071569 A CNA2009101071569 A CN A2009101071569A CN 200910107156 A CN200910107156 A CN 200910107156A CN 101571875 A CN101571875 A CN 101571875A
- Authority
- CN
- China
- Prior art keywords
- image
- module
- index
- subimage
- text
- 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
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a realization method of an image searching system based on image recognition; wherein the realization method comprises the following steps: carrying out preprocessing steps such as format transformation on images in adding images; carrying out recursion multilevel cutting to obtain all potential characters, portraits, objects and object subimage blocks based on uniform space, contrast grade, similarity in color and texture; obtaining marking texts of position and color; carrying out texture and material recognition, character recognition, outline object recognition and text marking on the subimage blocks; carrying out combination and word segmentation on the marking results; establishing index for the word segmentation result and related subimage block information and storing related data of the subimage blocks; in the situation of searching based on images, first identifying images and marking texts; then carrying out word segmentation on the text to obtain index data of a keyword sequence; combining and sequencing the index data of the keyword sequence to obtain related data of the subimage blocks and the images matching with the result set, and returning to users. With the system and method of the invention, the users can input characters and images to retrieve the content of the images.
Description
Technical field:
The present invention relates to a kind of implementation method of image search system, relate in particular to a kind of implementation method of the picture material search system based on image recognition.
Background technology:
Along with computing machine, Development of Multimedia Technology, increasing image information has appearred, and most mode with digital picture of these information is stored in the storage medium.Enter the later stage in last century, network technology has obtained huge development and popularization, and increasing view data transmits, stores and retrieve by network.Huge image file shop has appearred, at the line image share system, and image search engine.But the function of search of these systems at present is to analyze the related text that textual descriptions such as near the page epigraph of introducing this picture literal or image header, name are determined image mostly, and the user can't retrieving based on picture material.
Flame Image Process, Segmentation Research are more and more deep.After view data removed pre-service after noise, geometry and colour correction, the conversion, cut apart based on equidistant literal piece again, based on color, contrast, the technology such as cut apart of objects such as the object of texture similarity, portrait can effectively cut out object and subobjects such as literal, portrait and object from an image.
Utilize OCR technology such as (Optical Character Recognition optical character identification) that the literal piece that cuts out in the image is discerned, the technology of obtaining literal and layout information is also ripe and begin widespread use.
Since computer based AI (Artificial Intelligence artificial intelligence) technology began one's study from last century, the achievement of a lot of brilliances had appearred.With BP (Back Propagation error backpropagation algorithm) forward direction multitiered network neural network is that the neural network of representative is wherein a kind of, neural network is the mathematical operation model that a kind of structure that is similar to cerebral nerve cynapse connection is carried out information processing, by a large amount of node (or claim ' neuron ') and between be coupled to each other formation, each node is represented a kind of specific output function, is called excitation function (activation function).Per two internodal connections all represent one to be referred to as weight (weight) for the weighted value by this connection signal, and this is equivalent to the memory of artificial neural network.The output of network is then according to the connected mode of network, weighted value and excitation function different and different.And network self all is to the approaching of certain algorithm of nature or function usually, also may be the expression to a kind of logic strategy.SVM (SupportVector Machine support vector machine) is the sorter of another effective artificial intelligence, it is a kind of method of supervising formula study, it is widely used in statistical classification and the regretional analysis, SVM belongs to the vague generalization linear classifier, the characteristics of this family's sorter are that they can minimize how much marginariums of experience error and maximization simultaneously, so support vector machine is also referred to as maximal margin region class device.The subimage block that splits after the Flame Image Process classified, discerns by the intelligent algorithm of artificial neural network and SVM and so on can reach effect preferably.
The lasting lifting of the calculated performance of computing machine, the capacity of storage system, network exchange ability, the lasting research and the progress of technology such as artificial intelligence, Flame Image Process, image recognition make exploitation become possibility based on the picture material search system of image recognition.
Summary of the invention:
At the deficiency of conventional images search system, the purpose of this invention is to provide picture material search system based on image recognition.
The present invention is achieved in that a kind of implementation method of the image search system based on image recognition, comprises following main process and step:
Described image search system comprises image input module, image processing module, picture recognition module, image index module, image memory module, modules such as image search module.
When adding image, described image input module marks the image data transmission of input to image processing module;
Described image processing module carries out pre-service with view data, and cuts, and the result of cutting is sent into that picture recognition module is discerned, text marking, annotation results is gathered as replying then and beams back image input module;
After described image input module is received described annotation results, image, sub-image data are sent into the image memory module stores, then described subimage mark result is merged, processing such as participle, then described word segmentation result and described subimage index information are mail to the image index module and set up index.
When the user initiates searching request after image search module, described image search module is handled described searching request, obtain related text, described crucial text is carried out participle, request sends to described image index module acquisition index data to the lists of keywords that described participle is obtained as search index, described index data is merged, sorts and obtain the matching result collection, then described matching result collection is mail to image memory module inquiry subimage block and image related data, the related data of described subimage block and image is returned to the user.
Further, described image processing module comprises following function:
The pretreated process of image is to carry out format conversion, and various image data formats are converted into the internal data format of being convenient to Flame Image Process and identification, carries out convolution algorithm, binaryzation, image color, luminance transformation, rotation, convergent-divergent, noise reduction ... wait processing;
The character cutting of back gauges such as described image segmentation is based on, connected region based on similarities such as color, contrast, texture similarities is cut apart, and will cut apart the subimage block that obtains and carry out the recurrence multi-stage division, up to the subimage block size threshold value that reaches minimum, thereby be partitioned into the possible subimage block that all comprise single literal, portrait, object, object, subobject.
Further, the process discerned of described picture recognition module:
Described picture recognition module is at first carried out the contrast and the lines test of described subimage block, the potential alphabetic character of the conduct of the strong svelteness of contrast carries out literal identification, the comprehensive assessment of classifying of artificial neural network, SVM scheduling algorithm is adopted in the identification of described literal, the classification results assessed value preferably character as the label character of described subimage identification.
Described picture recognition module is carried out the profile identifying, carry out type and subtype loop test in the extensive sorter based on artificial neural network, SVM scheduling algorithm, with the type text of the sorter of all couplings label character as described subimage block identification.
The process of setting up of described extensive sorter further:
Adopt the way of automatic and semi-automatic combination, the passing through of sorter specifies a large amount of subimage blocks to carry out automatic, semi-automatic iteration training to each type word.The sample of training can obtain from 3 big approach at least: the existing normal pictures material database that marks the type text, carry out after the image segmentation by the online type text that marks out each height piece of user, the picture that picture name or heading message are enumerated in the website accurately (if any businessman to individual's e-commerce website).
Further, described image memory module needs the storage space of persistant data.
Further, described image index module exists, and is to be keyword with the word in the word segmentation result, the tabulation of in store subimage relative index information in the storage space of the persistant data of this module management.
Description of drawings:
Below in conjunction with accompanying drawing, the present invention is made detailed description.
Fig. 1 is an image labeling subfunction sequential chart
Fig. 2 is for adding the image sequential chart
Fig. 3 is based on text search image sequential chart
Fig. 4 is based on picture searching image sequential chart
Fig. 5 is sample picture (including notebook computer and pen and literal)
Fig. 6 is the main subimage block tabulation figure in the sample picture segmentation result
Embodiment:
The present invention is by to Flame Image Process, carries out the recurrence multi-level images then and cuts apart, and the identified sub-images piece also marks out that people such as the shape of describing object, position, type, subtype, texture, pattern, character can understand and the feature text of concern comparatively.And, carry out participle and set up index the Word message of these features.Query interface externally is provided, and the user can retrieve the image that occurs " red notebook computer " in the picture material by input " red notebook computer ".Also can submit picture to, by system picture is carried out Flame Image Process, carry out the multistage cutting of recurrence, intelligence mark then, being converted into text retrieves, as to submit a content to be the image of red notebook computer, system can retrieve the picture that contains red notebook computer in the picture material, and the picture that similarity is the highest is listed in the front.
The 4 main big functions of system of the present invention structure, the image labeling subfunction, the interpolation image based on the text search image, based on the picture searching image, below illustrates it respectively.
The process of " image labeling subfunction " is as shown in Figure 1:
Image is handled, cuts apart, discerned, return all literal that identify, object and mark text thereof.
The body step is: image processing module is received (shown in the P1 of Fig. 1) after the request of image labeling, and view data is carried out format conversion, and is unified to the internal image form, is convenient to subsequent treatment.Then the view data after the format conversion is carried out pre-service, remove noise.With pretreated view data based on the cutting of equidistant literal piece or based on color, the connected region isotype of similarities such as comparison, texture similarity is cut, the image block that is split reattempts in this piece to be cut apart once more, up to the critical size that reaches the minimum image piece, finish the recurrence multi-stage division.Possible literal, the object of sizes at different levels all are divided into subimage block, so that follow-up identification, Fig. 5 is sample picture (including notebook computer and pen and literal), and the result set of the main subimage block after result's cutting as shown in Figure 6.Each subimage block that splits is carried out the space mark, mark out position and border that this piece occurs, and obtain the custom space mark literal (as: centre ...) of the position of this piece in image.Each the height piece that splits is carried out the color mark, calculate the overall color of subimage block, adopt average weighted mode to obtain image value, be converted into the color mark literal of knowing (as: blueness, redness) according to Color Range.The image subblock that splits is sent to picture recognition module one by one to be discerned.Picture recognition module is received (shown in the I1 of Fig. 1) after the request of subimage block identification, the identification item of be correlated with trial.Picture recognition module is at first carried out character recognition.Can carry out the contrast test during identification earlier, whether contrast clearly demarcated, lines are tangible, there are not a large amount of acnodes, as satisfy character features, and carry out the character image feature extraction, view data is converted into the standard input data of sorter, by being discerned by artificial neural network, SVM algorithm, the character that identifies is as the mark text.Picture recognition module is then carried out Material Identification, subimage block is divided into the fritter of normal size, artificial neural network that each fritter utilization has trained or the sorter of SVM are discerned, the result of the fritter that each has been identified is weighted on average at last, obtain more excellent material texture recognition result relatively, if there is more excellent result, just obtain the pairing material texture mark of this result text (as: glass, timber ...).Picture recognition module is then carried out shape recognition, according to extracting the subimage block data, particularly contrasts the profile of the edge formation of strong part, carry out image characteristics extraction, the standard that view data is converted into sorter is imported data, by carrying out the correlation type loop test in a large amount of sorters that are made of artificial neural network, SVM, obtains type and the subtype and the text marking (as: people of all couplings, the woman, automobile, mobile phone, notebook computer, skirt, one-piece dress ...) tabulation.Identification, annotation results that picture recognition module is incited somebody to action gather, as a result of beam back image processing module (shown in the I2 of Fig. 1), at least the type and the subtype name (as: people that comprise the object that has identified literal, identified among this result, the woman, automobile, mobile phone, notebook computer, skirt, one-piece dress ...), material (as: glass, timber), texture and corresponding text.Image processing module merges, gathers the result of all images identification module.The image block that will belong to literal carries out the position relatedness computation, single literal is communicated with constitute continuous text.Beam back and reply (shown in the P2 of Fig. 1), reply the view data that comprises after the format conversion, the tabulation of annotation results of cutting apart, discerning and corresponding mark text.
" interpolation image " as shown in Figure 2:
Add image process and at first carried out " image labeling subfunction (so as Fig. 2 gray background part) ", set up index then according to annotating the result then.
Concrete steps are: after image input module is received the request of adding image, take out relevant informations such as view data, send to image processing module and carry out image labeling.After image processing module is received the image labeling request, carry out image labeling, the process that specifically marks is referring to " image labeling subfunction ".Image processing module is beamed back image input module with the result of image labeling.Image input module marks processing after receiving the result of image processing module, with each subimage block that identifies as an identifying object, for this object distributes ID one by one.The literal, object and the corresponding mark text list that identify are carried out processing such as participle.The view data after image input module will transform and the ID of identifying object and related data thereof send to the image memory module.The data storage that the image memory module is preserved the image input module request is in the persistent storage space.Image input module will mark text and carry out processing such as participle, set up index then, and the identified object (and subimage block information) of each word correspondence and the relative index information of the current input picture of handling are sent to the image index module.The image index module is added index data in the index of corresponding word after receiving that index adds request." based on the text search image " as shown in Figure 3:
Input text is carried out participle, obtain the relevant index of keyword, and carry out sequencing by merging, return coupling associated picture data.
Concrete steps are: image search module obtains the request based on the text search image.Image search module is carried out logical operator with the text and is detected, and carries out processing such as participle, merging, obtains searching for related keyword list.Image search module is sent the search index request of lists of keywords to the image index module.After the image index module is received the search index request of image search module, image search module is taken out and beamed back to the relative index of all lists of keywords in this request.Image search module is carried out the merging of index after receiving replying that the image index module beams back, specifically according to the character string of user's input with or relation is occured simultaneously or and set operation and then obtain the index amalgamation result, this result is carried out relevancy ranking.Image search module is sent the request of subimage block and the acquisition of associated picture data according to ranking results to the image memory module.After the image memory module obtained subimage block and view data acquisition request, the related data of taking out subimage block was beamed back image search module.
" based on the picture search image " as shown in Figure 4:
At first carry out " image labeling subfunction (so as Fig. 4 gray background part) ",, obtain the relevant index of keyword, and merge, sort, return and mate the associated picture data then with the mark text participle of annotation results correspondence.
Concrete steps are: after the request of image search module acquisition based on the picture search image, take out raw image data, send the image labeling request to image processing module.After image processing module is received the image labeling request, carry out image labeling, image processing module is beamed back image input module with the result of image labeling, and the process that specifically marks is referring to " image labeling subfunction ".Image search module is received the processing of the laggard rower notes of the annotation results of image processing module, takes out the mark text of all text block that mark out and all subimage block correspondences that identify, and carries out processing such as participle, merging, obtains related keyword list.Image search module is sent the search index request of this lists of keywords to the image index module.After the image index module is received the search index request of image search module, image search module is taken out and beamed back to the relative index of all lists of keywords in this request.Image search module is carried out the merging of index, and this result is carried out relevancy ranking after receiving replying that the image index module beams back.Image search module is sent the request of subimage block and the acquisition of associated picture data according to ranking results to the image memory module.After the image memory module obtained subimage block and view data acquisition request, the related data of taking out subimage block was beamed back image search module.
Claims (4)
1. the implementation method based on the image search system of image recognition is characterized in that, this method may further comprise the steps:
Described image search system comprises image input module, image processing module, picture recognition module, image index module, image memory module, modules such as image search module.
When adding image, described image input module marks the image data transmission of input to image processing module;
Described image processing module carries out pre-service with view data, and cuts, and the result of cutting is sent into that picture recognition module is discerned, text marking, annotation results is gathered as replying then and beams back image input module;
After described image input module is received described annotation results, image, sub-image data are sent into the image memory module stores, then described subimage mark result is merged, processing such as participle, then described word segmentation result and described subimage index information are mail to the image index module and set up index.
When the user initiates searching request after image search module, described image search module is handled described searching request, obtain related text, described crucial text is carried out participle, request sends to described image index module acquisition index data to the lists of keywords that described participle is obtained as search index, described index data is merged, sorts and obtain the matching result collection, then described matching result collection is mail to image memory module inquiry subimage block and image related data, the related data of described subimage block and image is returned to the user.
2. image processing module is characterised in that according to claim 1:
Described image processing module comprises following function:
The pretreated process of image is to carry out format conversion, and various image data formats are converted into the internal data format of being convenient to Flame Image Process and identification, carries out convolution algorithm, binaryzation, image color, luminance transformation, rotation, convergent-divergent, noise reduction ... wait processing;
The character cutting of back gauges such as described image segmentation is based on, connected region based on similarities such as color, contrast, texture similarities is cut apart, and will cut apart the subimage block that obtains and carry out the recurrence multi-stage division, up to the subimage block size threshold value that reaches minimum, thereby be partitioned into the possible subimage block that all comprise single literal, portrait, object, object, subobject.
3. the process discerned of picture recognition module is characterised in that according to claim 1:
Described picture recognition module is at first carried out the contrast and the lines test of described subimage block, the potential alphabetic character of the conduct of the strong svelteness of contrast carries out literal identification, the comprehensive assessment of classifying of artificial neural network, SVM scheduling algorithm is adopted in the identification of described literal, the classification results assessed value preferably character as the label character of described subimage identification.
Described picture recognition module is carried out the profile identifying, carry out type and subtype loop test in the extensive sorter based on artificial neural network, SVM scheduling algorithm, with the type text of the sorter of all couplings label character as described subimage block identification.
4. the foundation as extensive sorter as described in the claim 3 is characterized by:
Adopt the way of automatic and semi-automatic combination, sorter carries out automatic, semi-automatic iteration training by each type word being specified a large amount of subimage blocks.The sample of training can obtain from 3 big approach at least: the existing normal pictures material database that marks the type text, carry out after the image segmentation by the online type text that marks out each height piece of user, the picture that picture name or heading message are enumerated in the website accurately (if any businessman to individual's e-commerce website).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009101071569A CN101571875A (en) | 2009-05-05 | 2009-05-05 | Realization method of image searching system based on image recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2009101071569A CN101571875A (en) | 2009-05-05 | 2009-05-05 | Realization method of image searching system based on image recognition |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101571875A true CN101571875A (en) | 2009-11-04 |
Family
ID=41231233
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2009101071569A Pending CN101571875A (en) | 2009-05-05 | 2009-05-05 | Realization method of image searching system based on image recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101571875A (en) |
Cited By (63)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102156855A (en) * | 2011-03-30 | 2011-08-17 | 信雅达系统工程股份有限公司 | Bank voucher data collection method based on image segmentation |
| CN102231148A (en) * | 2010-11-29 | 2011-11-02 | 辜进荣 | Visual information searching method |
| CN102323926A (en) * | 2011-06-15 | 2012-01-18 | 百度在线网络技术（北京）有限公司 | Device and method for acquiring and requesting object information relevant to object |
| CN102509044A (en) * | 2011-10-17 | 2012-06-20 | 镇江金钛软件有限公司 | Mouse behavior characteristic-based password authentication method |
| CN102541999A (en) * | 2010-11-16 | 2012-07-04 | 微软公司 | Object-sensitive image search |
| CN102722709A (en) * | 2012-05-23 | 2012-10-10 | 杭州朗和科技有限公司 | Method and device for identifying garbage pictures |
| CN101763424B (en) * | 2009-12-14 | 2013-03-06 | 刘二中 | Method for determining characteristic words and searching according to file content |
| CN103020446A (en) * | 2012-12-10 | 2013-04-03 | 吴志宏 | Data processing system for information of chronicle spinal degenerative diseases and deformity |
| CN103425715A (en) * | 2012-05-25 | 2013-12-04 | 百度在线网络技术（北京）有限公司 | Method and system for confirming text annotations of pictures |
| CN103577414A (en) * | 2012-07-20 | 2014-02-12 | 富士通株式会社 | Data processing method and device |
| CN103597500A (en) * | 2011-06-29 | 2014-02-19 | 高通股份有限公司 | System and method for recognizing text information in object |
| CN103678305A (en) * | 2012-08-31 | 2014-03-26 | 北京网秦天下科技有限公司 | Method and system for displaying inquiring information based on image recognition |
| CN104025610A (en) * | 2011-12-28 | 2014-09-03 | 英特尔公司 | Systems, methods and apparatus for providing content based on a collection of images |
| CN104252475A (en) * | 2013-06-27 | 2014-12-31 | 腾讯科技（深圳）有限公司 | Method and device for positioning text messages in picture |
| CN105117706A (en) * | 2015-08-28 | 2015-12-02 | 小米科技有限责任公司 | Image processing method and apparatus and character recognition method and apparatus |
| CN105260727A (en) * | 2015-11-12 | 2016-01-20 | 武汉大学 | Academic-literature semantic restructuring method based on image processing and sequence labeling |
| CN105760390A (en) * | 2014-12-17 | 2016-07-13 | 富泰华工业（深圳）有限公司 | Image retrieval system and method |
| CN105900084A (en) * | 2013-12-20 | 2016-08-24 | 高通股份有限公司 | Systems, methods, and apparatus for image retrieval |
| CN106022024A (en) * | 2016-06-03 | 2016-10-12 | 深圳市金立通信设备有限公司 | System login method based on image recognition and terminal |
| CN106126755A (en) * | 2016-08-24 | 2016-11-16 | 广东华邦云计算股份有限公司 | A kind of purchase method based on image recognition |
| CN106155994A (en) * | 2016-06-30 | 2016-11-23 | 广东小天才科技有限公司 | The comparative approach of a kind of content of pages and device, terminal unit |
| CN106528739A (en) * | 2016-10-28 | 2017-03-22 | 杭州慕锐科技有限公司 | Method for establishing big data warehouse of digital printing and dyeing image materials |
| CN106650821A (en) * | 2016-12-30 | 2017-05-10 | 厦门幻世网络科技有限公司 | Method and device for transmitting information |
| CN106709488A (en) * | 2016-12-20 | 2017-05-24 | 深圳市深信服电子科技有限公司 | Business card identification method and device |
| WO2017092574A1 (en) * | 2015-12-01 | 2017-06-08 | 慧科讯业有限公司 | Mixed data type data based data mining method |
| CN106844614A (en) * | 2017-01-18 | 2017-06-13 | 天津中科智能识别产业技术研究院有限公司 | A kind of floor plan functional area system for rapidly identifying |
| CN106909548A (en) * | 2015-12-22 | 2017-06-30 | 北京奇虎科技有限公司 | Picture loading method and device based on server |
| CN106921804A (en) * | 2017-04-10 | 2017-07-04 | 青岛海信移动通信技术股份有限公司 | Method, device and the terminal device of schedule are created in the terminal |
| CN107016368A (en) * | 2017-04-07 | 2017-08-04 | 郑州悉知信息科技股份有限公司 | The information acquisition method and server of a kind of object |
| CN107054937A (en) * | 2017-03-23 | 2017-08-18 | 广东数相智能科技有限公司 | A kind of refuse classification suggestion device and system based on image recognition |
| CN107330750A (en) * | 2017-05-26 | 2017-11-07 | 北京三快在线科技有限公司 | A kind of recommended products figure method and device, electronic equipment |
| CN107644022A (en) * | 2016-07-21 | 2018-01-30 | 联想移动通信科技有限公司 | A kind of geographical position searching method, mobile terminal and server |
| CN107729547A (en) * | 2017-11-01 | 2018-02-23 | 上海掌门科技有限公司 | Retrieve the method and apparatus of circle of friends message |
| CN107920256A (en) * | 2017-11-30 | 2018-04-17 | 广州酷狗计算机科技有限公司 | Live data playback method, device and storage medium |
| CN108009251A (en) * | 2017-12-01 | 2018-05-08 | 珠海市魅族科技有限公司 | A kind of image file searching method and device |
| CN108170742A (en) * | 2017-12-19 | 2018-06-15 | 百度在线网络技术（北京）有限公司 | Picture public sentiment acquisition methods, device, computer equipment and storage medium |
| CN108256833A (en) * | 2018-04-09 | 2018-07-06 | 南京晓庄学院 | A kind of IN TIME schedule processing systems |
| CN108292311A (en) * | 2015-11-11 | 2018-07-17 | 三星电子株式会社 | Device and method for handling metadata |
| CN108287817A (en) * | 2017-05-08 | 2018-07-17 | 腾讯科技（深圳）有限公司 | A kind of information processing method and equipment |
| CN108681705A (en) * | 2018-05-15 | 2018-10-19 | 国网重庆市电力公司电力科学研究院 | A kind of measuring equipment consistency checking method and system based on figure identification |
| CN108763244A (en) * | 2013-08-14 | 2018-11-06 | 谷歌有限责任公司 | It searches for and annotates in image |
| CN108921185A (en) * | 2018-05-04 | 2018-11-30 | 广州图匠数据科技有限公司 | A kind of shelf sales promotion information recognition methods based on image recognition, device and system |
| CN109388649A (en) * | 2018-09-28 | 2019-02-26 | 湖南土流信息有限公司 | Intelligent land recommendation method and system |
| CN109635719A (en) * | 2018-12-10 | 2019-04-16 | 宽凳（北京）科技有限公司 | A kind of image-recognizing method, device and computer readable storage medium |
| CN109711289A (en) * | 2018-12-13 | 2019-05-03 | Oppo广东移动通信有限公司 | Based reminding method, device, electronic equipment and storage medium by bus |
| CN109791559A (en) * | 2016-06-28 | 2019-05-21 | 谷歌有限责任公司 | Promote use of the image as search inquiry |
| CN109858324A (en) * | 2018-12-10 | 2019-06-07 | 西南石油大学 | Language graphics context identifying system and its implementation |
| CN109874018A (en) * | 2018-12-29 | 2019-06-11 | 深兰科技（上海）有限公司 | Image encoding method, system, terminal and storage medium neural network based |
| CN109885842A (en) * | 2018-02-22 | 2019-06-14 | 谷歌有限责任公司 | Handle text neural network |
| CN109996107A (en) * | 2017-12-29 | 2019-07-09 | 百度在线网络技术（北京）有限公司 | Video generation method, device and system |
| CN110162639A (en) * | 2019-04-16 | 2019-08-23 | 深圳壹账通智能科技有限公司 | Knowledge figure knows the method, apparatus, equipment and storage medium of meaning |
| CN110377338A (en) * | 2019-07-23 | 2019-10-25 | 网易无尾熊（杭州）科技有限公司 | Information labeling method, medium, device and calculating equipment |
| CN110503087A (en) * | 2019-08-19 | 2019-11-26 | 广东小天才科技有限公司 | A kind of searching method, device, terminal and the storage medium of frame topic of taking pictures |
| CN110688504A (en) * | 2019-09-27 | 2020-01-14 | 中国工商银行股份有限公司 | Image data management method, apparatus, system, device and medium |
| CN111291738A (en) * | 2020-05-09 | 2020-06-16 | 支付宝(杭州)信息技术有限公司 | Element extraction method and device in front-end page image and electronic equipment |
| CN111353419A (en) * | 2020-02-26 | 2020-06-30 | 北京百度网讯科技有限公司 | Image comparison method and device, electronic equipment and storage medium |
| CN111460782A (en) * | 2020-04-01 | 2020-07-28 | 支付宝(杭州)信息技术有限公司 | Information processing method, device and equipment |
| CN111651674A (en) * | 2020-06-03 | 2020-09-11 | 北京妙医佳健康科技集团有限公司 | Bidirectional searching method and device and electronic equipment |
| CN111782850A (en) * | 2019-11-27 | 2020-10-16 | 北京沃东天骏信息技术有限公司 | Object searching method and device based on hand drawing |
| CN111783786A (en) * | 2020-07-06 | 2020-10-16 | 上海摩勤智能技术有限公司 | Picture identification method and system, electronic equipment and storage medium |
| CN112199545A (en) * | 2020-11-23 | 2021-01-08 | 湖南蚁坊软件股份有限公司 | Keyword display method and device based on picture character positioning and storage medium |
| CN112381038A (en) * | 2020-11-26 | 2021-02-19 | 中国船舶工业系统工程研究院 | Image-based text recognition method, system and medium |
| CN116959075A (en) * | 2023-08-01 | 2023-10-27 | 湖北省电子信息产品质量监督检验院 | Deep learning-based iterative optimization method for identity recognition robot |
-
2009
- 2009-05-05 CN CNA2009101071569A patent/CN101571875A/en active Pending
Cited By (94)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101763424B (en) * | 2009-12-14 | 2013-03-06 | 刘二中 | Method for determining characteristic words and searching according to file content |
| CN102541999A (en) * | 2010-11-16 | 2012-07-04 | 微软公司 | Object-sensitive image search |
| CN102541999B (en) * | 2010-11-16 | 2015-09-02 | 微软技术许可有限责任公司 | The picture search of object sensitivity |
| CN102231148A (en) * | 2010-11-29 | 2011-11-02 | 辜进荣 | Visual information searching method |
| CN102156855A (en) * | 2011-03-30 | 2011-08-17 | 信雅达系统工程股份有限公司 | Bank voucher data collection method based on image segmentation |
| CN102323926A (en) * | 2011-06-15 | 2012-01-18 | 百度在线网络技术（北京）有限公司 | Device and method for acquiring and requesting object information relevant to object |
| CN102323926B (en) * | 2011-06-15 | 2014-09-10 | 百度在线网络技术（北京）有限公司 | Device and method for acquiring and requesting object information relevant to object |
| US9418304B2 (en) | 2011-06-29 | 2016-08-16 | Qualcomm Incorporated | System and method for recognizing text information in object |
| CN103597500A (en) * | 2011-06-29 | 2014-02-19 | 高通股份有限公司 | System and method for recognizing text information in object |
| CN102509044A (en) * | 2011-10-17 | 2012-06-20 | 镇江金钛软件有限公司 | Mouse behavior characteristic-based password authentication method |
| CN104025610B (en) * | 2011-12-28 | 2017-11-17 | 英特尔公司 | For providing the system of content, method and apparatus based on a collection of image |
| CN104025610A (en) * | 2011-12-28 | 2014-09-03 | 英特尔公司 | Systems, methods and apparatus for providing content based on a collection of images |
| CN102722709B (en) * | 2012-05-23 | 2014-10-29 | 杭州朗和科技有限公司 | Method and device for identifying garbage pictures |
| CN102722709A (en) * | 2012-05-23 | 2012-10-10 | 杭州朗和科技有限公司 | Method and device for identifying garbage pictures |
| CN103425715A (en) * | 2012-05-25 | 2013-12-04 | 百度在线网络技术（北京）有限公司 | Method and system for confirming text annotations of pictures |
| CN103425715B (en) * | 2012-05-25 | 2016-06-29 | 百度在线网络技术（北京）有限公司 | A kind of method and system determining picture text marking |
| CN103577414B (en) * | 2012-07-20 | 2017-04-12 | 富士通株式会社 | Data processing method and device |
| CN103577414A (en) * | 2012-07-20 | 2014-02-12 | 富士通株式会社 | Data processing method and device |
| CN103678305A (en) * | 2012-08-31 | 2014-03-26 | 北京网秦天下科技有限公司 | Method and system for displaying inquiring information based on image recognition |
| CN103020446B (en) * | 2012-12-10 | 2016-02-10 | 吴志宏 | The data handling system of chronic vertebra degenerative disease and lopsided information |
| CN103020446A (en) * | 2012-12-10 | 2013-04-03 | 吴志宏 | Data processing system for information of chronicle spinal degenerative diseases and deformity |
| CN104252475B (en) * | 2013-06-27 | 2018-03-27 | 腾讯科技（深圳）有限公司 | Position the method and device of text in picture information |
| CN104252475A (en) * | 2013-06-27 | 2014-12-31 | 腾讯科技（深圳）有限公司 | Method and device for positioning text messages in picture |
| CN108763244A (en) * | 2013-08-14 | 2018-11-06 | 谷歌有限责任公司 | It searches for and annotates in image |
| CN108763244B (en) * | 2013-08-14 | 2022-02-01 | 谷歌有限责任公司 | Searching and annotating within images |
| CN105900084A (en) * | 2013-12-20 | 2016-08-24 | 高通股份有限公司 | Systems, methods, and apparatus for image retrieval |
| US10346465B2 (en) | 2013-12-20 | 2019-07-09 | Qualcomm Incorporated | Systems, methods, and apparatus for digital composition and/or retrieval |
| CN105900084B (en) * | 2013-12-20 | 2019-12-31 | 高通股份有限公司 | System, method and apparatus for image retrieval |
| CN105760390A (en) * | 2014-12-17 | 2016-07-13 | 富泰华工业（深圳）有限公司 | Image retrieval system and method |
| CN105760390B (en) * | 2014-12-17 | 2021-09-28 | 富泰华工业（深圳）有限公司 | Picture retrieval system and method |
| CN105117706B (en) * | 2015-08-28 | 2019-01-18 | 小米科技有限责任公司 | Image processing method and device, character identifying method and device |
| CN105117706A (en) * | 2015-08-28 | 2015-12-02 | 小米科技有限责任公司 | Image processing method and apparatus and character recognition method and apparatus |
| US11328186B2 (en) | 2015-11-11 | 2022-05-10 | Samsung Electronics Co., Ltd. | Device and method for processing metadata |
| CN108292311A (en) * | 2015-11-11 | 2018-07-17 | 三星电子株式会社 | Device and method for handling metadata |
| CN105260727B (en) * | 2015-11-12 | 2018-09-21 | 武汉大学 | Academic documents semanteme based on image procossing and sequence labelling structural method again |
| CN105260727A (en) * | 2015-11-12 | 2016-01-20 | 武汉大学 | Academic-literature semantic restructuring method based on image processing and sequence labeling |
| WO2017092574A1 (en) * | 2015-12-01 | 2017-06-08 | 慧科讯业有限公司 | Mixed data type data based data mining method |
| CN106909548A (en) * | 2015-12-22 | 2017-06-30 | 北京奇虎科技有限公司 | Picture loading method and device based on server |
| CN106022024B (en) * | 2016-06-03 | 2019-10-22 | 深圳市金立通信设备有限公司 | A kind of method and terminal for realizing system login based on image recognition |
| CN106022024A (en) * | 2016-06-03 | 2016-10-12 | 深圳市金立通信设备有限公司 | System login method based on image recognition and terminal |
| CN109791559B (en) * | 2016-06-28 | 2023-10-10 | 谷歌有限责任公司 | Promoting image processing apparatus as a means of use of search queries |
| CN109791559A (en) * | 2016-06-28 | 2019-05-21 | 谷歌有限责任公司 | Promote use of the image as search inquiry |
| CN106155994B (en) * | 2016-06-30 | 2019-04-26 | 广东小天才科技有限公司 | A kind of comparative approach and device, terminal device of content of pages |
| CN106155994A (en) * | 2016-06-30 | 2016-11-23 | 广东小天才科技有限公司 | The comparative approach of a kind of content of pages and device, terminal unit |
| CN107644022A (en) * | 2016-07-21 | 2018-01-30 | 联想移动通信科技有限公司 | A kind of geographical position searching method, mobile terminal and server |
| CN106126755A (en) * | 2016-08-24 | 2016-11-16 | 广东华邦云计算股份有限公司 | A kind of purchase method based on image recognition |
| CN106528739A (en) * | 2016-10-28 | 2017-03-22 | 杭州慕锐科技有限公司 | Method for establishing big data warehouse of digital printing and dyeing image materials |
| CN106528739B (en) * | 2016-10-28 | 2019-07-23 | 杭州米绘科技有限公司 | A kind of method for building up in Digital Dyeing picture material big data warehouse |
| CN106709488A (en) * | 2016-12-20 | 2017-05-24 | 深圳市深信服电子科技有限公司 | Business card identification method and device |
| CN106650821A (en) * | 2016-12-30 | 2017-05-10 | 厦门幻世网络科技有限公司 | Method and device for transmitting information |
| CN106844614A (en) * | 2017-01-18 | 2017-06-13 | 天津中科智能识别产业技术研究院有限公司 | A kind of floor plan functional area system for rapidly identifying |
| CN107054937A (en) * | 2017-03-23 | 2017-08-18 | 广东数相智能科技有限公司 | A kind of refuse classification suggestion device and system based on image recognition |
| CN107016368A (en) * | 2017-04-07 | 2017-08-04 | 郑州悉知信息科技股份有限公司 | The information acquisition method and server of a kind of object |
| CN106921804A (en) * | 2017-04-10 | 2017-07-04 | 青岛海信移动通信技术股份有限公司 | Method, device and the terminal device of schedule are created in the terminal |
| CN108287817A (en) * | 2017-05-08 | 2018-07-17 | 腾讯科技（深圳）有限公司 | A kind of information processing method and equipment |
| CN108287817B (en) * | 2017-05-08 | 2020-08-11 | 腾讯科技（深圳）有限公司 | Information processing method and device |
| CN107330750A (en) * | 2017-05-26 | 2017-11-07 | 北京三快在线科技有限公司 | A kind of recommended products figure method and device, electronic equipment |
| CN107729547A (en) * | 2017-11-01 | 2018-02-23 | 上海掌门科技有限公司 | Retrieve the method and apparatus of circle of friends message |
| CN107920256A (en) * | 2017-11-30 | 2018-04-17 | 广州酷狗计算机科技有限公司 | Live data playback method, device and storage medium |
| CN107920256B (en) * | 2017-11-30 | 2020-01-10 | 广州酷狗计算机科技有限公司 | Live broadcast data playing method and device and storage medium |
| CN108009251A (en) * | 2017-12-01 | 2018-05-08 | 珠海市魅族科技有限公司 | A kind of image file searching method and device |
| CN108170742A (en) * | 2017-12-19 | 2018-06-15 | 百度在线网络技术（北京）有限公司 | Picture public sentiment acquisition methods, device, computer equipment and storage medium |
| CN109996107A (en) * | 2017-12-29 | 2019-07-09 | 百度在线网络技术（北京）有限公司 | Video generation method, device and system |
| CN109885842B (en) * | 2018-02-22 | 2023-06-20 | 谷歌有限责任公司 | Processing text neural networks |
| CN109885842A (en) * | 2018-02-22 | 2019-06-14 | 谷歌有限责任公司 | Handle text neural network |
| CN108256833A (en) * | 2018-04-09 | 2018-07-06 | 南京晓庄学院 | A kind of IN TIME schedule processing systems |
| CN108921185A (en) * | 2018-05-04 | 2018-11-30 | 广州图匠数据科技有限公司 | A kind of shelf sales promotion information recognition methods based on image recognition, device and system |
| CN108681705A (en) * | 2018-05-15 | 2018-10-19 | 国网重庆市电力公司电力科学研究院 | A kind of measuring equipment consistency checking method and system based on figure identification |
| CN109388649A (en) * | 2018-09-28 | 2019-02-26 | 湖南土流信息有限公司 | Intelligent land recommendation method and system |
| CN109635719A (en) * | 2018-12-10 | 2019-04-16 | 宽凳（北京）科技有限公司 | A kind of image-recognizing method, device and computer readable storage medium |
| CN109635719B (en) * | 2018-12-10 | 2023-11-17 | 宽凳（北京）科技有限公司 | Image recognition method, device and computer readable storage medium |
| CN109858324B (en) * | 2018-12-10 | 2022-10-21 | 西南石油大学 | Language image-text recognition system and implementation method thereof |
| CN109858324A (en) * | 2018-12-10 | 2019-06-07 | 西南石油大学 | Language graphics context identifying system and its implementation |
| CN109711289A (en) * | 2018-12-13 | 2019-05-03 | Oppo广东移动通信有限公司 | Based reminding method, device, electronic equipment and storage medium by bus |
| CN109874018A (en) * | 2018-12-29 | 2019-06-11 | 深兰科技（上海）有限公司 | Image encoding method, system, terminal and storage medium neural network based |
| CN110162639A (en) * | 2019-04-16 | 2019-08-23 | 深圳壹账通智能科技有限公司 | Knowledge figure knows the method, apparatus, equipment and storage medium of meaning |
| CN110377338A (en) * | 2019-07-23 | 2019-10-25 | 网易无尾熊（杭州）科技有限公司 | Information labeling method, medium, device and calculating equipment |
| CN110503087A (en) * | 2019-08-19 | 2019-11-26 | 广东小天才科技有限公司 | A kind of searching method, device, terminal and the storage medium of frame topic of taking pictures |
| CN110688504A (en) * | 2019-09-27 | 2020-01-14 | 中国工商银行股份有限公司 | Image data management method, apparatus, system, device and medium |
| CN111782850A (en) * | 2019-11-27 | 2020-10-16 | 北京沃东天骏信息技术有限公司 | Object searching method and device based on hand drawing |
| CN111353419B (en) * | 2020-02-26 | 2023-08-11 | 北京百度网讯科技有限公司 | Image comparison method, device, electronic equipment and storage medium |
| CN111353419A (en) * | 2020-02-26 | 2020-06-30 | 北京百度网讯科技有限公司 | Image comparison method and device, electronic equipment and storage medium |
| CN111460782B (en) * | 2020-04-01 | 2023-08-22 | 支付宝(杭州)信息技术有限公司 | Information processing method, device and equipment |
| CN111460782A (en) * | 2020-04-01 | 2020-07-28 | 支付宝(杭州)信息技术有限公司 | Information processing method, device and equipment |
| CN111291738A (en) * | 2020-05-09 | 2020-06-16 | 支付宝(杭州)信息技术有限公司 | Element extraction method and device in front-end page image and electronic equipment |
| CN111651674B (en) * | 2020-06-03 | 2023-08-25 | 北京妙医佳健康科技集团有限公司 | Bidirectional searching method and device and electronic equipment |
| CN111651674A (en) * | 2020-06-03 | 2020-09-11 | 北京妙医佳健康科技集团有限公司 | Bidirectional searching method and device and electronic equipment |
| CN111783786A (en) * | 2020-07-06 | 2020-10-16 | 上海摩勤智能技术有限公司 | Picture identification method and system, electronic equipment and storage medium |
| CN112199545B (en) * | 2020-11-23 | 2021-09-07 | 湖南蚁坊软件股份有限公司 | Keyword display method and device based on picture character positioning and storage medium |
| CN112199545A (en) * | 2020-11-23 | 2021-01-08 | 湖南蚁坊软件股份有限公司 | Keyword display method and device based on picture character positioning and storage medium |
| CN112381038A (en) * | 2020-11-26 | 2021-02-19 | 中国船舶工业系统工程研究院 | Image-based text recognition method, system and medium |
| CN112381038B (en) * | 2020-11-26 | 2024-04-19 | 中国船舶工业系统工程研究院 | Text recognition method, system and medium based on image |
| CN116959075A (en) * | 2023-08-01 | 2023-10-27 | 湖北省电子信息产品质量监督检验院 | Deep learning-based iterative optimization method for identity recognition robot |
| CN116959075B (en) * | 2023-08-01 | 2024-02-06 | 湖北省电子信息产品质量监督检验院 | Deep learning-based iterative optimization method for identity recognition robot |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101571875A (en) | Realization method of image searching system based on image recognition | |
| Mi et al. | Hierarchical graph attention network for visual relationship detection | |
| CN101923653B (en) | Multilevel content description-based image classification method | |
| CN100414548C (en) | Search system and technique comprehensively using information of graphy and character | |
| CN107657008B (en) | Cross-media training and retrieval method based on deep discrimination ranking learning | |
| CN101739428B (en) | Method for establishing index for multimedia | |
| CN109165294A (en) | Short text classification method based on Bayesian classification | |
| CN110674407A (en) | Hybrid recommendation method based on graph convolution neural network | |
| US11195046B2 (en) | Method and system for image search and cropping | |
| CN104834693A (en) | Depth-search-based visual image searching method and system thereof | |
| CN101996191A (en) | Method and system for searching for two-dimensional cross-media element | |
| CN103425757A (en) | Cross-medial personage news searching method and system capable of fusing multi-mode information | |
| CN104715021A (en) | Multi-label learning design method based on hashing method | |
| CN109213845B (en) | Original news evaluation method and system based on article characteristics | |
| CN108388656A (en) | A kind of image searching method based on label correlation | |
| CN111931953A (en) | Multi-scale characteristic depth forest identification method for waste mobile phones | |
| CN109657063A (en) | A kind of processing method and storage medium of magnanimity environment-protection artificial reported event data | |
| CN112036511B (en) | Image retrieval method based on attention mechanism graph convolution neural network | |
| CN101515286B (en) | Image matching method based on image feature multi-level filtration | |
| CN103929499B (en) | A kind of Internet of Things isomery index identification method and system | |
| CN112148938B (en) | Cross-domain heterogeneous data retrieval system and retrieval method | |
| CN106529601B (en) | Image classification prediction technique based on sparse subspace multi-task learning | |
| US20120158784A1 (en) | Method and system for image search | |
| CN111159456B (en) | Multi-scale clothing retrieval method and system based on deep learning and traditional features | |
| CN103020630B (en) | The disposal route of characteristics of image and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| DD01 | Delivery of document by public notice |
Addressee: Cheng Zhiyong Document name: the First Notification of an Office Action |
|
| DD01 | Delivery of document by public notice |
Addressee: Cheng Zhiyong Document name: Notification that Application Deemed to be Withdrawn |
|
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20091104 |