CN109344275B

CN109344275B - Image identification-based resource acquisition device and method

Info

Publication number: CN109344275B
Application number: CN201810956056.2A
Authority: CN
Inventors: 米玖松; 赵芬花
Original assignee: Shenzhen Zhiyuan Youxue Education Technology Co ltd
Current assignee: Shenzhen Zhiyuan Youxue Education Technology Co ltd
Priority date: 2018-08-21
Filing date: 2018-08-21
Publication date: 2021-10-26
Anticipated expiration: 2038-08-21
Also published as: CN109344275A

Abstract

The invention discloses a resource acquisition device and a method based on picture identification, wherein the device comprises a client picture acquisition device for acquiring an external picture; the picture identification device is used for carrying out picture identification and outputting an identification result; the matching device is used for matching data according to the identification result and outputting a matching result; the output device outputs corresponding resource contents in the corresponding education resource packet according to the matching result; the picture identification device comprises a picture preprocessing module and a picture processing module, wherein the picture preprocessing module is used for cutting an external picture into a plurality of cut pictures; the picture characteristic matching module is used for carrying out characteristic matching on the cut pictures and outputting pictures which are successfully matched; and the result screening module receives the successfully matched picture and outputs an identification result. In the resource acquisition device and method based on image recognition, a user scans paper textbooks/test papers and the like into external images through handheld equipment such as a tablet personal computer and the like, and learning contents can be automatically searched by a server and directly enter a learning interface, so that the user operation is simplified, and the learning efficiency is improved.

Description

Image identification-based resource acquisition device and method

Technical Field

The invention relates to the field of information-based education, in particular to a resource acquisition device and method based on picture recognition.

Background

The traditional electronic courseware learning resource positioning mode mainly has three types: the first is to locate the courseware information through the courseware resource directory; the second is to input the name of the textbook and the page number information to position the courseware information; the third is to locate learning resources such as courseware, knowledge points, test questions and the like through a search engine. The first mode of the courseware resource directory is to download courseware resource directories corresponding to the appointed textbook or complete multimedia courseware with the courseware resource directories, and then open the courseware resource directories to select the resources needing to be learned step by step; the second mode of inputting textbook and page number information requires the user to manually input the page number information of the book, and learning resources corresponding to the page number of the book are obtained; and thirdly, after the keywords are input, a search engine is started to search and position learning resources such as courseware, knowledge points, test questions and the like.

However, the existing methods have certain defects. The first method is through courseware resource directory can not be through the page location. The second method of inputting the page number of the book requires the user to manually input the information of the book and the page number of the book, requires the user to repeatedly perform input operation positioning, and cannot ensure the accuracy of the input information at one time. And thirdly, the search engine can only take input characters as input keywords, cannot input information such as graphic pictures and formulas as the input keywords, and has limited search results, thereby causing low search efficiency. The three modes have complicated operation steps and cannot be simply and quickly realized.

Disclosure of Invention

The invention aims to provide an improved resource acquisition device and method based on picture recognition.

The technical scheme adopted by the invention for solving the technical problems is as follows: the resource acquisition device based on picture identification comprises

The client picture acquisition device is used for acquiring an external picture;

the picture identification device is connected with the client picture acquisition device and used for carrying out picture identification according to the acquired external picture and outputting an identification result;

the matching device is connected with the picture recognition device and used for selectively matching data according to the recognition result and outputting a matching result;

the output device is connected with the matching device and used for outputting corresponding resource contents in the corresponding education resource packages according to the matching result;

wherein the picture recognition device comprises

The picture preprocessing module is used for cutting the external picture into a plurality of cut pictures and setting the priority and weight information of each cut picture;

the picture feature matching module is used for performing feature matching on each cut picture and outputting a picture which is successfully matched;

and the result screening module is used for receiving the successfully matched picture and outputting the identification result.

Preferably, the picture preprocessing module comprises

The first cropping unit is used for subtracting a first non-effective area from the left side and the right side of the external picture along an X axis and generating a first cropping picture, wherein the ratio of the width sum of the first non-effective areas on the left side and the right side to the width of the external picture is 5-45%, and the ratio of the width of the first cropping picture to the width of the external picture is 55-95%;

and the second clipping unit is used for subtracting a second non-effective area from the upper side and the lower side of the external picture along the Y axis and generating a second clipping picture, wherein the height ratio of the sum of the heights of the second non-effective areas on the upper side and the lower side to the external picture is 5-45%, and the height ratio of the second clipping picture to the external picture is 55-95%.

Preferably, the picture preprocessing module further comprises

A third cropping unit, configured to subtract a third non-effective region from a far end of the external picture along a Y axis and generate a third cropped picture, where a height ratio of the third non-effective region to the external picture is 25-45%, and a height ratio of the third cropped picture to the external picture is 55-75%;

and the fourth cropping unit is used for subtracting a fourth non-effective area from the far end of the external picture along the Y axis and generating a fourth cropped picture, wherein the height ratio of the fourth non-effective area to the external picture is 25-45%, and the height ratio of the fourth cropped picture to the external picture is 55-75%.

Preferably, the picture preprocessing module further includes a fifth cropping unit, configured to subtract a fifth non-effective region from a far end of the external picture along the Y axis and generate a fifth cropped picture, where a height ratio of the fifth non-effective region to the external picture is 46% to 60%, and a height ratio of the fifth cropped picture to the external picture is 40% to 54%.

Preferably, the result screening module comprises

The receiving judgment unit judges whether the matching success picture is received or not, and if yes, a receiving result is output; if not, outputting matching failure;

the maximum weight judging unit judges whether the successfully matched picture is the maximum weight or not according to the receiving result, and if so, the cut picture with the maximum weight is output as the identification result; if not, outputting a judgment result;

the weight comparison unit is used for comparing whether the successfully matched picture is the picture with the maximum weight in the received pictures according to the judgment result, and if so, outputting a comparison result; if not, returning to the receiving judgment unit;

a receiving end judging unit, which judges whether all the successfully matched pictures are received according to the comparison result, if so, the comparison result is output as the identification result; if not, returning to the receiving judgment unit.

Also provides a resource acquisition method based on picture identification, which comprises

Acquiring an external picture;

performing picture identification according to the acquired external picture, and outputting an identification result;

selectively performing data matching according to the identification result, and outputting a matching result;

outputting corresponding resource contents in the corresponding education resource packages according to the matching results;

wherein, the image recognition according to the acquired external image and outputting a recognition result comprises:

cutting the external picture into a plurality of cut pictures, and setting the priority and weight information of each cut picture;

performing feature matching on each cut picture, and outputting a picture successfully matched;

and receiving the successfully matched picture and outputting the identification result.

Preferably, the cropping the external picture into a plurality of cropped pictures includes

Subtracting a first non-effective area from the left side and the right side of the external picture along an X axis to generate a first cut picture, wherein the ratio of the sum of the widths of the first non-effective areas on the left side and the right side to the width of the external picture is 5-45%, and the ratio of the width of the first cut picture to the width of the external picture is 55-95%;

and subtracting a second non-effective area from the upper side and the lower side of the external picture along the Y axis to generate a second cut picture, wherein the height ratio of the sum of the heights of the second non-effective areas on the upper side and the lower side to the external picture is 5-45%, and the height ratio of the second cut picture to the external picture is 55-95%.

Preferably, the cropping the external picture into a plurality of cropped pictures further comprises

Subtracting a third non-effective area from the far end of the external picture along the Y axis, and generating a third cut picture, wherein the height ratio of the third non-effective area to the external picture is 25-45%, and the height ratio of the third cut picture to the external picture is 55-75%;

and subtracting a fourth non-effective area from the far end of the external picture along the Y axis to generate a fourth cut picture, wherein the height ratio of the fourth non-effective area to the external picture is 25-45%, and the height ratio of the fourth cut picture to the external picture is 55-75%.

Preferably, the cropping the external picture into a plurality of cropped pictures further comprises: and subtracting a fifth non-effective area from the far end of the external picture along the Y axis to generate a fifth cut picture, wherein the height ratio of the fifth non-effective area to the external picture is 46% -60%, and the height ratio of the fifth cut picture to the external picture is 40% -54%.

Preferably, the receiving the successfully matched picture and outputting the recognition result includes:

judging whether the matching success picture is received or not, if so, outputting a receiving result; if not, outputting matching failure;

judging whether the successfully matched picture is the maximum weight or not according to the receiving result, and if so, outputting the cut picture with the maximum weight as the identification result; if not, outputting a judgment result;

comparing whether the successfully matched picture is the picture with the maximum weight in the received pictures according to the judgment result, and if so, outputting a comparison result; if not, returning to the judgment whether the matching success picture is received or not;

judging whether all the successfully matched pictures are received or not according to the comparison result, if so, outputting the comparison result as the identification result; if not, returning to the judgment whether the matching success picture is received or not.

The beneficial effects of the implementation of the invention are as follows: according to the resource acquisition device and method based on picture recognition, the external picture generated by scanning the printed matters such as the paper textbook and the like is used as the input key information, and a user only needs to scan the printed matters such as the paper textbook/the test paper and the like through handheld equipment such as a tablet personal computer and the like to generate the external picture, so that the server can automatically search related learning contents and directly enter a learning interface, user operation is simplified, and the learning efficiency of the user is improved.

In addition, the resource acquisition device and method based on image identification also have the following three technical advantages:

1. the user does not need to have the computer character input capability;

2. the correctness of the input book information is ensured;

3. the operation steps are simplified.

Drawings

The invention will be further described with reference to the accompanying drawings and examples, in which:

FIG. 1 is a schematic diagram of the overall configuration of an educational resource positioning system in which a picture recognition based resource acquisition device is located, according to some embodiments of the present invention;

FIG. 2 is a schematic diagram of the connection between the resource acquisition device and other modules of the image recognition-based educational resource positioning system of FIG. 1 in some embodiments of the present invention;

FIG. 3 is a schematic diagram illustrating the connection relationship between the client image capturing device, the image recognizing device and the matching device in FIG. 2;

FIG. 4 is a diagram illustrating clipping of an external picture by a first clipping unit in a resource acquisition method based on picture recognition according to some embodiments of the invention;

FIG. 5 is a diagram illustrating clipping an external picture by a third clipping unit in a resource obtaining method based on picture recognition according to some embodiments of the invention;

FIG. 6 is a diagram illustrating clipping an external picture by a fifth clipping unit in a resource obtaining method based on picture recognition according to some embodiments of the invention;

FIG. 7 is a diagram illustrating clipping of an external picture by a second clipping unit in a resource acquisition method based on picture recognition according to some embodiments of the invention;

FIG. 8 is a diagram illustrating clipping an external picture by a fourth clipping unit in a resource obtaining method based on picture recognition according to some embodiments of the invention;

FIG. 9 is a schematic diagram of the construction of the educational asset library module 10 of FIG. 1;

FIG. 10 is a diagram illustrating the composition of the original picture library module 30 of the picture server in FIG. 1;

FIG. 11 is a schematic diagram of the composition of the original picture and index library module 20 of FIG. 1;

fig. 12 is a schematic diagram showing the structure of the chapter associating module 40 in fig. 1;

FIG. 13 is a block diagram of the original picture and index compression packet and alias association module 50 of FIG. 1;

FIG. 14 is an overall flow chart of a resource acquisition method based on picture recognition in some embodiments of the present invention;

FIG. 15 is a detailed flow chart of a resource acquisition method based on picture recognition in some embodiments of the present invention;

FIG. 16 is a schematic diagram of the picture recognition apparatus shown in FIG. 3;

fig. 17 is a schematic flow chart illustrating result screening in the resource obtaining method based on picture recognition according to some embodiments of the present invention.

Detailed Description

For a more clear understanding of the technical features, objects and effects of the present invention, embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

Fig. 1 shows the general configuration of an educational resource positioning system based on picture recognition in which a resource acquisition device is located in some embodiments of the present invention, and fig. 2 is a schematic diagram of the configuration of the resource acquisition device in fig. 1. The educational resource positioning system 700 comprises a picture server original picture library module 30, an educational resource library module 10, an original picture and index library module 20, a chapter association module 40, an original picture compression package and alias association module 50 and a resource acquisition device 800, wherein the original picture library module 30 is disposed on a picture server 500, and the educational resource library module 10, the original picture and index library module 20, the chapter association module 40, the original picture compression package and alias association module 50 are disposed on a search engine server 600. The resource acquiring device 800 comprises a picture acquiring device 100, a picture identifying device 200, a matching device 300, an output device 400, and a client receiving and displaying device 900.

The invention provides a resource acquisition method and a resource acquisition device which can be used for searching and positioning educational resources by scanning printed matters such as paper textbooks/test papers and generating external pictures as input key information and returning search results to a user, namely, the user directly scans the paper textbooks/test papers through handheld equipment such as a tablet personal computer and generates the external pictures, then the external pictures are used as the input key information and are uploaded to a picture server and are matched with picture server original pictures in a picture server original picture library module 30 stored in the picture server, and then the file names of the picture server original pictures which are successfully matched are used as the input key information to search matched learning contents in an educational resource library module 10 stored in a search engine server, and then the successfully searched learning contents are returned to the user. The method and the device have the advantages that the user can immediately search related learning contents and directly enter a learning interface by scanning printed matters such as the paper textbook and the like through education electronic learning equipment such as a tablet personal computer and the like, and the process that the user manually inputs keywords is omitted, so that the input of high-difficulty key information such as graphs, pictures, formulas and equations is avoided, the user operation is simplified, the input of the key information is enriched, the operation time is saved, the searching efficiency and the learning efficiency are improved, and the user is helped to construct a good knowledge system.

Referring to fig. 2, a picture generated after a scanning program on a handheld device such as a tablet computer directly scans an inner page or a non-inner page of a paper textbook/test paper is called an external picture, the external picture is sent to a picture recognition device 200 by a client picture acquisition device 100, and is matched with an original picture of a picture server in a picture server 500, and then a file name of the successfully matched original picture of the picture server is sequentially matched with an original picture, an index compression packet and alias association table, an original picture index table and a chapter association table through a matching device 300, and an identification of an education resource packet, an original picture file name and corresponding chapter information which are successfully matched are output, and the output device 400 outputs corresponding resource contents in the corresponding education resource packet according to a result of successful matching of the matching device 300; the client receiving and displaying device 900 receives the corresponding resource content in the educational resource package output by the output device 400 and displays the resource content to the user.

The terms appearing in some embodiments of the invention are explained and defined below in conjunction with fig. 1-17.

A paper textbook: the paper textbook refers to a primary school textbook, a middle school textbook, a primary school teaching aid, a middle and primary school end test paper, a middle and high school entrance examination question and other types of learning type paper presswork. Each paper textbook corresponds to an original picture and index compression packet and an education resource packet. The original image and index compression package comprises an original image folder and an original image index table, and the original image folder comprises series scanned images in JPG (joint photographic experts group) or PNG (public network group) and other formats generated after the paper textbook is scanned according to pages. The educational resource package contains various learning resources of the paper textbook. Each paper textbook and the original picture and index compression packet correspond to a unique alias.

Alias: each original picture and index compression packet is defined with a unique alias, which is also called a prefix of the original picture and index compression packet. The alias naming method comprises the following steps: the method comprises the steps of using a resource type, a subject name, an original picture and an index compression packet ID number, namely using the resource type of an education resource packet corresponding to each paper textbook as a first part, using the subject to which the paper textbook belongs as a second part, using the original picture and the index compression packet ID number corresponding to the paper textbook as a third part, and combining the three parts according to the sequence to form the alias of the original picture and the index compression packet. If PS in the alias "PSCHINESEG 110245" indicates that the resource type is a textbook reading resource, CHINESE is a subject, and G110245 is the original picture and the index compression packet ID number corresponding to the paper textbook. It should be understood that the alias is also an alias for each paper textbook.

An educational resource package: the education resource package comprises various learning resources displayed to users, such as analysis of text of a textbook, explanation of example questions, explanation of related knowledge points, synchronous training, teaching and assisting questions or examination papers and the like, and the display form of the education resource package comprises one or more types of figures, pictures, characters, voice, videos, exercises, Flash, interesting games and the like. The set of learning resources corresponding to each paper textbook forms an education resource package of the paper textbook, namely, each paper textbook corresponds to one education resource package.

Identification of the educational resource package: the file name of each educational resource package serves as the "educational resource package identifier" of the educational resource package, which is unique in the educational resource library module 10.

Original picture: the original pictures refer to series inner page and non-inner page scanning pictures in a format of JPG or PNG and the like generated after all contents of a paper textbook are scanned from a front cover to a back cover according to pages, the non-inner page scanning pictures refer to scanning pictures corresponding to the non-front contents of the front cover, the back cover, a preamble, a catalogue, a postscript and the like, and the inner page scanning pictures refer to a plurality of scanning pictures corresponding to the front contents.

Original picture clip: all scanned pictures of each paper textbook are stored under the same picture folder, and the file name of the picture folder can be named as 'picture'. In this patent document, "original image folder" is simply referred to as "original image folder", and the scanned image in the original image folder is updated along with the update of the content of the paper textbook.

Original picture file name: all the inside page scanned pictures and non-inside page scanned pictures in the original picture folder have corresponding file names, called original picture file names, which are unique in the corresponding original picture folder. The file name of the inner page scanning picture is consistent with the page number of the corresponding text content page in the paper textbook corresponding to the inner page scanning picture; the name of the non-inner page scanning picture file is the same as the content of the non-inner page scanning picture file, such as a front cover, a catalogue, a postscript, a back cover and the like.

External picture: the picture generated after the scanning program on the handheld equipment such as the tablet personal computer directly scans the inner page or the non-inner page of the paper textbook is called as an external picture.

Specific embodiments are described below in conjunction with fig. 1-17 and the above explanations and definitions.

The educational resource library module 10 is used for storing a plurality of educational resource packages. The education resource library module 10 is stored in the search engine server 600, and as shown in fig. 1 to 3, the education resource library module 10 is formed by assembling a plurality of education resource packages corresponding to paper textbooks. Each paper textbook corresponds to an education resource package which contains various learning resources displayed to users, such as the analysis of textbook texts, the explanation of example questions, the explanation of related knowledge points, synchronous training, teaching and assisting questions or examination papers and the like, and the display form of the textbook comprises one or more types of figures, pictures, characters, sounds, videos, exercises, flashes, interesting games and the like.

Each educational resource package has a unique file name as its identifier in the educational resource library module 10, thus ensuring that all educational resource packages within the educational resource library module 10 are not renamed. The file name is called 'education resource package identification' of the education resource package, and the naming method is as follows: the "resource type, teaching material version, belonging school section, belonging subject, belonging grade, school period, teaching material edition, teaching material printing number, and other information" of the paper textbook corresponding to the education resource package are combined in sequence, for example, the name "nine synchronization-human education edition-junior-mathematics-seven grade-lower-12 edition" is the file name corresponding to the paper textbook of the 12 th edition of the school period under seven-year grade of nine synchronization human education edition junior-middle mathematics (i.e., "education resource package identification" of the education resource package corresponding to the paper textbook), and the identification is unique in the education resource library module 10.

It should be understood that the educational resource package contains various types of learning resources that are uploaded and stored in the educational resource library module 10 of the search engine server 600 by the manufacturer after being manufactured on the local device. The process of making the educational resource package is not described herein in detail.

Referring to fig. 1 again, the picture server original picture library module 30 is used for storing a plurality of picture server original pictures. Optionally, the system further includes a picture server 500, and the picture server original picture library module 30 is stored in the picture server 500.

All original pictures in the original picture folders corresponding to all paper textbooks on the local equipment are uploaded and integrated under the same folder of the picture server 500 to form a picture server original picture library module 30, and each picture in the library is called as a picture server original picture.

Each picture server original picture is set with an integer type unique serial number as a "picture server original picture identifier" which is unique in the "picture server original picture library module 30".

The file name of each picture server original picture is called as the 'picture server original picture file name' of the picture, and is called as the 'picture server original picture name' for short. Each picture server raw picture file name is unique in the picture server raw picture library module 30. The naming method of the original picture file name of the picture server comprises the following steps: the original picture to which the original picture corresponding to the original picture of the picture server belongs, the alias of the index compression packet and the file name of the original picture corresponding to the original picture of the picture server are as follows: the first part of the file name of the original picture of the picture server is the original picture to which the local original picture corresponding to the original picture of the picture server belongs and the alias of the index compression packet; the second part is the original picture file name corresponding to the original picture of the picture server. If "PSCHINESEG 110245 cover" is the file name of the original picture of the picture server, where "PSCHINESEG 110245" part is an alias, "cover" is the file name of the original picture corresponding to the original picture of the picture server, i.e. the file name of the original picture is "cover", i.e. the original picture is a scanned picture of the cover corresponding to a paper textbook; also, for example, in "PSCHINESEGG 11024510 1102451", the part "PSCHINESEG 110245" is an alias, and 10 is the original picture file name, i.e., the 10 th page of the corresponding paper textbook page. The alias naming method is described above and will not be described herein.

The "picture server original picture identifier" and the "picture server original picture file name" are identification information of the picture server original picture in the picture server original picture library module 30.

It should be understood that the "picture server original pictures" in the picture server original picture library module 30 are obtained by the manufacturer by first uploading the original pictures in the original picture folder made for the paper textbook by the local device to the picture server 500 for storage, wherein the picture server original pictures can be added, deleted and updated.

Referring to fig. 1, the original image and index library module 20 is configured to store a plurality of original images and index packages, each of which includes an original image folder and an original image index table, the original image folder includes original images generated by scanning corresponding paper textbooks according to pages, and each index in the original image index table corresponds to a corresponding original image in the original image folder; each paper textbook corresponds to an original picture, an index packet and an education resource packet.

The original picture and the index packet are formed by uploading the original picture and the index compression packet which are made by the manufacturer on the local equipment aiming at the paper textbook in advance to the search engine server 600 and decompressing, and the specific contents and file names of the original picture and the index compression packet are the same with those of the original picture and the index compression packet, so that the invention mainly describes the related information of the original picture and the index compression packet.

The file name of the original picture and the index compression packet is called as an original picture and index compression packet identifier, and as mentioned above, each original picture and index compression packet identifier is unique, and the naming method is as follows: the resource type + teaching material version + school period + subject + grade + school period + others, namely the resource type, teaching material version, school period, subject, grade, school period and other information of the paper textbook are combined in sequence, the resource type comprises textbook point reading resources, test question shooting resources and other types, and the subject refers to subjects of primary school and middle school such as Chinese and mathematics. If the original picture and the index compression package are marked as 'teaching and assistant subjects-synchronous test paper-human teaching five and four plate making-primary school-mathematics-six grades-up-teaching materials full explanation reference full-17 years autumn-shooting search', the original picture and the file name of the index compression package are corresponding to a paper textbook 'teaching materials full explanation reference full-2017 autumn synchronous teaching and assistant test paper of five and four plate making primary school mathematics six grades of people education publishers'.

The following describes a resource acquisition system based on picture recognition according to an embodiment of the present invention with reference to fig. 2 to 13.

As shown in fig. 2, the resource acquisition apparatus 800 includes: the client picture acquiring device 100, the picture identifying device 200, the matching device 300, the output device 400 and the receiving and displaying device 900.

The client-side picture acquiring apparatus 100 is configured to acquire an external picture.

The picture recognition device 200 is connected to the client picture acquiring device 100, and is configured to perform picture recognition according to the acquired external picture, and output a recognition result. It mainly includes three modules: a picture preprocessing module 2001, a picture feature matching module 2002, and a result screening module 2003.

The picture preprocessing module 2001 is connected to the client-side picture capturing device 100, and is configured to crop an external picture into multiple cropped pictures, and set priority and weight information of each cropped picture. Preferably, the picture preprocessing module 2001 is configured to perform corresponding processing on an external picture to generate a plurality of cut pictures, so as to improve the recognition rate of the external picture, specifically: and cutting the external picture into a plurality of cut pictures, and setting the priority and the weight of each cut picture.

The picture feature matching module 2002 is connected to the picture preprocessing module 2001, and configured to perform feature matching on each cut picture, and output a picture that is successfully matched. Preferably, the picture feature matching module 2002 is configured to extract specific features of the cropped picture, perform feature matching on the cropped picture and a picture server original picture in the picture server original picture library module 30, and further match the cropped picture to a specific picture server original picture.

The result screening module 2003 is connected to the picture preprocessing module 2002, and is configured to receive the successfully matched picture and output the recognition result. Preferably, the result filtering module 2003 is used for filtering out the picture server original picture closest to the external picture, and sending the file name of the picture server original picture to the matching device 300.

The matching device 300 is connected to the image recognition device 200, and is configured to selectively perform data matching according to the recognition result and output the matching result. Specifically, the matching device 300 is connected to the result filtering module 2003 and configured to perform resource matching according to the file name of the original picture of the picture server sent by the result filtering module 2003 and output an education resource package, a chapter name and a chapter number that are successfully matched;

and the output device 400 is connected with the matching device 300 and is used for outputting the corresponding resource content in the corresponding education resource package according to the matching result.

The client receiving and displaying device 900 is connected to the output device 400, and is configured to receive the corresponding resource content in the educational resource package output by the output device 400 and display the resource content to the user.

The following briefly describes a resource acquisition method based on picture recognition in an embodiment of the present invention with reference to fig. 1 to 17.

First, an external picture is acquired by the client-side picture acquisition apparatus 100.

Here, the respective modules in the present embodiment are described: a picture server original picture library module 30, configured to store a plurality of picture server original pictures; an education resource library module 10 for storing a plurality of education resource packages; the original picture and index library module 20 is used for storing a plurality of original pictures and index packages, each original picture and index package comprises an original picture folder and a corresponding original picture index table, the original picture folder comprises original pictures generated after corresponding paper textbooks are scanned according to pages, and each index in the original picture index table corresponds to a corresponding original picture in the original picture folder; each paper textbook corresponds to an original picture, an index packet and an education resource packet; the chapter association module 40 is used for storing a plurality of chapter association tables, wherein the chapter association tables comprise the corresponding relation between each chapter in a paper textbook and an original picture; and an original picture and index compressed packet and alias association module 50, configured to store a plurality of original picture and index compressed packet and alias association tables, where the original picture and index compressed packet and alias association tables include a corresponding relationship between an identifier of the original picture and index compressed packet and an alias.

Next, the picture recognition device 200 performs picture recognition based on the acquired external picture, and outputs a recognition result. Specifically, the external picture is compared with the original picture in the picture server original picture library module 30, and the picture server original picture name of the corresponding picture in the picture server original picture library module 30 is obtained according to the comparison result, where the picture server original picture name includes an alias and an original picture file name. As mentioned above, all the intra-page scanned pictures and non-intra-page scanned pictures in the "original picture folder" have corresponding file names, which are called "original picture file names", and the names are unique in the corresponding "original picture folder". The file name of the inner page scanning picture is consistent with the page number of the corresponding text content page in the paper textbook corresponding to the inner page scanning picture; the name of the non-inner page scanning picture file is the same as the content of the non-inner page scanning picture file, such as non-text content of a front cover, a catalogue, a postscript, a back cover and the like.

Then, the matching device 300 selectively performs data matching according to the recognition result, and outputs the matching result. Specifically, according to the alias in the original picture name of the picture server, a query and a comparison are performed in the "original picture and index compression packet and alias association table" in the original picture and index compression packet and alias association module 50 to obtain a corresponding original picture and index compression packet identifier; because the original picture and index compression packet identification is the same as the corresponding original picture index table, the corresponding original picture index table is positioned in the original picture and index library module 20 according to the original picture and index compression packet identification; positioning a corresponding index record on an original picture index table according to the original picture file name contained in the original picture name of the picture server; and because the original picture and index compression packet identification are identical to the corresponding original picture index table and chapter association table, the corresponding chapter association table is located on the chapter association module 40 according to the original picture and index compression packet identification, and the specific index record in the chapter association table according to the original picture and index compression packet identification and the original picture page number or original picture file serial number in the original picture index table, and the corresponding education resource packet and chapter name and chapter serial number in the education resource library module 10 according to the specific index record in the chapter association table, and the corresponding chapter content and chapter content of the corresponding education resource packet in the education resource library module according to the education resource packet and name and serial number, and the corresponding page resource learning content includes the analysis of the textbook, example questions, related knowledge points, synchronous training, teaching and assisting questions, test paper, exercise, video links, Flash links, interesting game links and the like.

The output device 400 outputs the corresponding educational resource package and the corresponding resource content in the chapter according to the matching result.

Finally, the client receiving and displaying device 900 receives the corresponding educational resource packages and corresponding resource contents in chapters and sections output by the output device 400 and displays the corresponding educational resource packages and sections to the user.

The following describes in detail a resource acquisition method based on picture recognition in an embodiment of the present invention with reference to fig. 14 and 15.

In step S100, the client image obtaining device 100 obtains an external image, and sends the generated external image file to the image recognition device 200, which specifically includes:

s1001, starting a client scanning program;

s1002, the scanning program automatically focuses and shoots and scans the scanned object in the scanning area to generate an external picture file;

s1003, sends the generated external picture file to the picture preprocessing module 2001.

In step S200, after the picture recognition apparatus 200 acquires the external picture from the client-side picture acquisition apparatus 100, the obtained external picture is subjected to picture recognition, and the picture server original picture file name most similar to the external picture is output. The method specifically comprises the following steps:

s2001, the picture preprocessing module 2001 receives the external picture and cuts the external picture to generate a plurality of cut pictures, and sets a priority and a weight of each cut picture.

S2002, the picture feature matching module 2002 extracts specific features of the cut picture, and performs feature matching between the plurality of cut pictures and the original picture of the picture server, respectively.

S2003, the result filtering module 2003 receives the successfully matched clipped pictures, filters a best matched clipped picture from the successfully matched clipped pictures according to the priority and weight of the clipped pictures, and sends the file name of the original picture of the picture server corresponding to the clipped picture to the matching device 300.

The following is a detailed explanation of the above S2001-S2003:

s2001, the picture preprocessing module 2001 receives the external picture and cuts the external picture to generate a plurality of cut pictures, and sets a priority and a weight of each cut picture:

the picture preprocessing module 2001 is connected to the client-side picture obtaining device 100, and is configured to crop an external picture to generate multiple cropped pictures, so as to improve an identification rate of the external picture, specifically: and cutting the external picture into a plurality of pictures, and setting the priority and the distribution weight of each cut picture.

(1) Cutting the picture: the client-side image obtaining device 100 captures an external image, where there are some fuzzy non-effective areas at the edge of the obtained external image, for the external image captured by a paper textbook (such textbook is a mainstream textbook) whose vertical height (i.e., height in Y direction) is greater than horizontal width (i.e., width in X direction), the non-effective areas on the left and right sides of the image are relatively large, for the external image captured by a paper textbook whose vertical height is less than horizontal width, the non-effective areas on the left and right sides of the image are relatively small, and the non-effective areas cause certain interference to image recognition, so that the image needs to be cut off, and a cut image including a central area is left. The picture cutting follows the principle that the effective area is as large as possible and the non-effective area is as small as possible, and the same external picture is cut for many times by different methods to generate a plurality of cut pictures with different effective areas. In order to improve the overall recognition rate, in this example, one external picture is cropped for multiple times to generate 5 cropped pictures.

In some embodiments, the picture pre-processing module includes a first cropping unit 2001A and a second cropping unit 2001B. The first cropping unit 2001A is configured to subtract a first non-valid region from each of left and right sides of the external picture along the X axis and generate a first cropped picture, where a ratio of a sum of widths of the first non-valid regions on the left and right sides to a width of the external picture is 5-45%, and a ratio of the width of the first cropped picture to the width of the external picture is 55-95%. The second cropping unit 2001B is configured to subtract a second non-effective region from each of the upper and lower sides of the external picture along the Y axis and generate a second cropped picture, where a height ratio of a sum of heights of the second non-effective regions at the upper and lower sides to the external picture is 5 to 45%, and a height ratio of the second cropped picture to the external picture is 55 to 95%.

Preferably, the picture preprocessing module further includes a third clipping unit 2001C and a fourth clipping unit 2001D. The third cropping unit 2001C is configured to subtract a third non-effective region from the external picture along the Y-axis and generate a third cropped picture, where a height ratio of the third non-effective region to the external picture is 25-45%, and a height ratio of the third cropped picture to the external picture is 55-75%. A fourth cropping unit 2001D is configured to subtract a fourth non-effective region from the outer picture along the Y-axis and generate a fourth cropped picture, where a height ratio of the fourth non-effective region to the outer picture is 25-45%, and a height ratio of the fourth cropped picture to the outer picture is 55-75%.

In some preferred embodiments, the picture preprocessing module further includes a fifth cropping unit 2001E, configured to subtract a fifth non-effective region from the external picture along the Y-axis, and generate a fifth cropped picture, where a height ratio of the fifth non-effective region to the external picture is 46% to 60%, and a height ratio of the fifth cropped picture to the external picture is 40% to 54%.

(2) In the best embodiment of the invention, the picture cutting method comprises the following steps: in this example, one external picture is cut into 5 pictures, and 5 different cutting methods are as follows:

1) for external pictures shot by paper textbooks (such textbooks are mainstream textbooks) with the upper and lower heights larger than the left and right widths, the non-effective areas on the left and right sides are relatively large, and three cut pictures are generally taken:

first, as shown in fig. 4, picture cropping is performed by the first cropping unit 2001A. In order to reduce interference areas on the left side and the right side of an external picture, namely in the X direction, the left upper corner of the picture is taken as an original point, non-effective areas to which width values of M pixels belong are removed from the left side and the right side of the external picture along the X direction by taking a Y axis as a dotted line, a remaining middle part is a first cut picture 101, the first cut picture 101 is called a first cut picture 101, the width value of the first cut picture 101 is W1, and the value is obtained by subtracting 2M pixel values from the width value of the external picture; the height of the first cropped picture 101 is equal to the height of the external picture, and is H pixel values. As shown in fig. 4, the region composed of O1O2O4O3 is the region to which the external picture belongs, where O1 is the upper left corner of the external picture and is also the origin of the intersection of the X axis and the Y axis, O1O2 is the width of the external picture in the X direction, and O1O3 is the height of the external picture in the Y direction. The region composed of A1A2A4A3 is the region to which the first cropping picture 101 belongs, where A1 is the upper left corner of the first cropping picture 101, A1A2 is the width of the first cropping picture 101, and its width value is W1; a1a3 is the height of the first cropped picture 101, and its height value is H; the O1 A3O3 area and the A2O 4a4 are the first blurred non-effective areas clipped on the left and right sides, respectively, where O1A1 and A2O2 are the widths of the non-effective areas clipped on the left and right sides, respectively, the widths thereof are M, respectively, and the height thereof is H as in the first clipped picture 101. In some embodiments, the ratio of the sum of the widths of the first non-active areas on the left and right sides to the width of the external picture is 5-45%, and the ratio of the width of the first cropped picture to the width of the external picture is 55-95%.

②, as shown in fig. 5, picture clipping is performed by the third clipping unit 2001C. In order to reduce the interference of a fuzzy area in the Y direction of an external picture, taking the upper left corner of the picture as an origin, taking an image of a clearer part area at a near end in the Y direction, generally taking an X axis as a starting point line, removing a far end area of the external picture along the Y direction, and leaving the near end area, wherein the ratio of the height of the far end area to the height of the external picture is generally 25-45%; the ratio of the height of the closer region to the height of the outer picture is typically 55-75%. The second cropped picture containing the near end region is thus generated and referred to as the third cropped picture 102. As shown in fig. 5, the region composed of O1O2O4O3 is the region to which the external picture belongs, where O1 is the upper left corner of the external picture and is also the origin of the intersection of the X axis and the Y axis, O1O2 is the width of the external picture in the X direction, and O1O3 is the height of the external picture in the Y direction. The region composed of O1B 2O2 is the region to which the third cropped picture 102 belongs, where O1 is the upper left corner of the third cropped picture 102, and O1O2 is the width of the third cropped picture 102, and the width value is the same as the width value of the external picture; O1B1 is the height of the third cropped picture 102, its height value is H2, and the ratio of the height value H2 to the height value H of the external picture is 55% -75%; the B1O3O4B2 area is a third non-effective area with the blur cut out at the far end, wherein B1B2 is the width of the non-effective area, the width is the same as the width of the external picture, B1O3 is the height of the non-effective area, the height is the height H of the external picture minus the height H2 of the third cut picture 102, and the ratio of the height to the height of the external picture is 25% -45%. In some embodiments, a height ratio of the third non-active area to the external picture is 25-45%, and a height ratio of the third cropped picture to the external picture is 55-75%.

Third, as shown in fig. 6, the fifth clipping unit 2001E clips a picture. Some third cropped pictures 102 may also be blurred and inconvenient to recognize, and therefore, the interference of a blurred region in the Y direction needs to be further reduced, an image of a more proximal clearer partial region is taken in the Y direction with the upper left corner of the picture as an origin, generally with the X axis as a dotted line, a more distal region of the external picture is removed in the Y direction, and a more proximal region is left, wherein the ratio of the height of the more distal region to the height of the external picture is generally 46% -60%; the ratio of the height of the more proximal region to the height of the outer picture is typically 40% to 54%. The third cropped picture containing the more proximal region is thus generated and is referred to as fifth cropped picture 103. As shown in fig. 6, the region composed of O1O2O4O3 is the region to which the external picture belongs, where O1 is the upper left corner of the external picture and is also the origin of the intersection of the X axis and the Y axis, O1O2 is the width of the external picture in the X direction, and O1O3 is the height of the external picture in the Y direction. A region composed of O1C 2O2 is a region to which the fifth cropped picture 103 belongs, where O1 is the upper left corner of the fifth cropped picture, and O1O2 is the width of the fifth cropped picture 103, and the width value is the same as the width value of the external picture; O1C1 is the height of the fifth cropped picture 103, with a height value of H3, the ratio of the height value H3 to the height value H of the external picture is 40% to 54%; the C1O3O4C2 area is a fifth non-effective area with blurred image clipped far away, where C1C2 is the width of the fifth non-effective area, the width is the same as the width of the external picture, C1O3 is the height of the fifth non-effective area, the height is the height H of the external picture minus the height H3 of the fifth clipped picture 103, and the ratio of the height to the height of the external picture is 46% -60%. In some embodiments, a height ratio of the fifth non-active area to the external picture is 46% -60%, and a height ratio of the fifth cropped picture to the external picture is 40% -54%.

2) For external pictures shot by a paper teaching material textbook with the height smaller than the width, the effective areas on the left side and the right side are more, and a small part of blank non-effective areas above and below cause interference, and two cut pictures are generally taken:

first, as shown in fig. 7, picture cropping is performed by the second cropping unit 2001B. In order to reduce the interference areas in the upper and lower sides of the external picture, the upper left corner of the external picture is taken as an origin, the non-effective areas to which the height values of N pixels belong are removed from the upper and lower sides of the external picture along the Y direction by taking the X axis as a dotted line, and then the remaining middle part is a fourth cut picture called a second cut picture 201, the width value of the second cut picture 201 is W and is the same as the width value of the external picture, and the height value of the cut picture 201 is the height value of the external picture minus 2N pixels and is H4 pixels. As shown in fig. 7, the region composed of O1O2O4O3 is the region to which the external picture belongs, where O1 is the upper left corner of the external picture and is also the origin of the intersection of the X axis and the Y axis, O1O2 is the width of the external picture in the X direction, and O1O3 is the height of the external picture in the Y direction. The region composed of D1D2D4D3 is the region to which the second cropped picture 201 belongs, where D1 is the upper left corner of the second cropped picture 201, and D1D2 is the width of the second cropped picture 201, and its width value is W; D1D3 is the height of the second cropped picture 201, with a height value of H4; the O1D 2O2 area and the D3O 4D4 are respectively the blurred second non-effective areas clipped at the upper and lower sides, wherein the O1D1 and the D3O3 are respectively the heights of the clipped second non-effective areas at the upper and lower sides, the height values are respectively N, and the width value is W as same as the external picture. In some embodiments, the ratio of the sum of the heights of the second non-effective areas on the upper and lower sides to the height of the external picture is 5-45%, and the ratio of the height of the second cropped picture to the height of the external picture is 55-95%.

② as shown in fig. 8, picture clipping is performed by the fourth clipping unit 2001D. In order to reduce the interference of a far-end fuzzy area in the Y direction of an external picture, taking the upper left corner of the picture as an origin, taking an image of a clearer part area at a near end in the Y direction, generally taking an X axis as a starting point line, removing a far-end area of the external picture along the Y direction, and leaving the near-end area, wherein the ratio of the height of the far-end area to the height of the external picture is generally 25-45%; the ratio of the height of the closer region to the height of the outer picture is typically 55-75%. The fifth cropped picture containing the near end region is thus generated and referred to as the fourth cropped picture 202. As shown in fig. 8, a region composed of O1O2O4O3 is a region to which the external picture belongs, where O1 is the upper left corner of the external picture and is also the origin of the intersection of the X axis and the Y axis, O1O2 is the width of the external picture in the X direction, and O1O3 is the height of the external picture in the Y direction. The region composed of O1E 2O2 is the region to which the fourth cropped picture 202 belongs, where O1 is the upper left corner of the fourth cropped picture 202, and O1O2 is the width of the fourth cropped picture 202, and the width value is the same as the width value of the external picture; O1E1 is the height of the fourth cropped picture 202, its height value is H5, and the ratio of the height value H5 to the height value H of the external picture is 55% to 75%; the E1O3O4E2 area is a fourth non-effective area with blurred images cropped off at the far end, where E1E2 is the width of the fourth non-effective area, the width is the same as the width of the external picture, E1O3 is the height of the fourth non-effective area, the height is the height H of the external picture minus the height H5 of the fourth cropped picture 202, and the ratio of the height to the height of the external picture is 25% -45%. In some embodiments, a height ratio of the fourth non-active area to the external picture is 25-45%, and a height ratio of the fourth cropped picture to the external picture is 55-75%.

It should be noted that, since the image preprocessing module 201 cannot determine whether the acquired external image is larger than the width or smaller than the width, the image preprocessing module 201 generally cuts the external image into the above 5 cut images. And because the height of the mainstream paper textbook is generally larger than the width, the weight of 5 cut pictures is sequentially ordered from large to small according to the principle that the effective area of the picture is as large as possible and the ineffective area is as small as possible: 101>201>102>202> 103.

It should be understood that, preferably, all external pictures are cropped into 5 cropped pictures, and 5 cropped pictures corresponding to each external picture are cropped and assigned with serial numbers according to the above method, that is, the serial numbers of the 5 cropped pictures corresponding to all external pictures are 101, 102, 103, 201 and 202.

It is worth noting that: the cropped pictures generally adopt imagemap function, and only four parameters (X, Y, W, H) are provided to the function to obtain corresponding cropped pictures 101,201,102,202, and 103, wherein (X, Y) is an origin coordinate of an upper left corner of the cropped picture, wherein X is a pixel position of the upper left corner of the cropped picture in the X-axis direction in the outer picture, it can be understood that X is the pixel position of the original point position of the upper left corner of the external picture plus M pixel points along the X-axis direction, Y is the pixel position of the original point of the upper left corner of the clipping picture plus N pixel points along the Y-axis direction, in this embodiment, the value M is generally 0 to 200, the value N is generally 0 to 100, W is the width of the cropped picture formed by extending W pixel points in the X-axis direction with the upper left-corner origin position as the starting point, and H is the height of the cropped picture formed by extending H pixel points in the Y-axis direction with the upper left-corner origin position as the starting point.

Alternatively, in some embodiments, only the first clipping unit 2001A and the second clipping unit 2001B may be provided, which has an advantage that both a picture with a large top-bottom height can be clipped by the first clipping unit 2001A and a picture with a large left-right width can be clipped by the second clipping unit 2001B. In some preferred embodiments, a third clipping unit 2001C and a fourth clipping unit 2001D may be further provided on the basis of the first clipping unit 2001A and the second clipping unit 2001B, and a picture with a large top-bottom height and a picture with a large left-right width may be further optimally clipped. Further, as in the five cropping modes included in the foregoing embodiment, a first cropping unit 2001A, a second cropping unit 2001B, a third cropping unit 2001C, a fourth cropping unit 2001D, and a fifth cropping unit 2001E may be included, and such a cropping effect is better.

It should be noted that the cutting method is not limited to the above five cutting methods, and various cutting methods may be added as long as the related functions can be realized.

The concrete description is as follows:

the picture feature matching module 2002 receives the cropped pictures 101,201,102,202, and 103 sent from the picture preprocessing module 2001, starts multithreading, matches the cropped pictures 101,201,102,202, and 103 with the stored picture server original pictures in the picture server original picture library module 30 one by one, and returns matching results. And respectively carrying out the following processing aiming at different matching results:

(1) if matching between the cut picture and the picture server original picture stored in the picture server original picture library module 30 fails, a corresponding failure code is returned, and the failure code is sent to the result screening module 2003;

(2) if the matching between the cut picture and the stored picture server original picture in the picture server original picture library module 30 is successful, the picture server original picture name corresponding to the picture server original picture which is successfully matched is returned, and the picture server original picture name is sent to the result filtering module 2003.

It should be understood that the format of the failure code returned by the picture feature matching module 2002 is: result _ fail _ clip picture number, i.e.,

picture numbers

101, 102, 103, 201, and 202 corresponding to the clip pictures 101, 102, 103, 201, and 202,

if result _ fail _101 indicates that the first cropped picture 101 fails to match.

The format of the matching success returned by the picture feature matching module 2002 is result _ success _ cropped picture serial number, which is the picture

serial numbers

101, 102, 103, 201, and 202 corresponding to the cropped

pictures

101, 102, 103, 201, and 202,

for example, result _ success _102 indicates that the second cropped picture 102 is successfully matched.

It should be noted that the picture feature matching module 2002 is a multi-thread asynchronous process, so that each time a matching result is returned, the matching result is sent to the result filtering module 2003 for processing. Since there are 5 cropped pictures in the preferred embodiment, the picture feature matching module 2002 returns a total of 5 matching results to the result filtering module 2003.

Each time a picture matching result sent by the picture feature matching module 2002 is received, the picture matching result is recorded in the memory and is correspondingly processed, and a specific processing method for the matching result is as follows:

(1) if all of the cropped

pictures

101, 102, 103, 201, and 202 fail to match the picture server original pictures in the picture server original picture library module, the result filtering module 2003 receives the failure code returned by the picture feature matching module 2002, which indicates that the external picture fails to match the corresponding picture server original picture in the picture server original picture library module.

(2) If the clipped

pictures

101, 102, 103, 201, and 202 match successfully with the original pictures of the picture server in the original picture library module of the picture server, the result filtering module 2003 filters out the clipped picture with the largest relative weight according to the weight corresponding to the clipped picture, and sends the file name of the original picture of the picture server in the picture server 500 corresponding to the clipped picture to the matching device 300. Specifically, when the result filtering module 2003 receives a result of successful matching of a certain clipped picture sent from the picture feature matching module 2002, it checks whether all other clipped pictures with a larger weight than the clipped picture have returned matching results:

1) if all other cut pictures with weights larger than the weight of the current cut picture return matching results and the results are matching failures, the result of the current matching success is the final matching result, and the file name of the original picture of the picture server in the picture server 500 corresponding to the current cut picture is sent to the matching device 300 without waiting for the matching results of other cut pictures with weights lower than the weight of the current cut picture;

2) if the matching result returned by other cut pictures with the weight larger than that of the current cut picture contains information of successful matching, the result of successful matching of the cut picture with the maximum weight is taken as a final matching result, and the file name of the original picture of the picture server in the picture server 500 corresponding to the cut picture with the maximum weight is sent to the matching device 300;

3) if the matching result of other clipped pictures with the weight larger than that of the current clipped picture is not returned, waiting for the matching result returned by other clipped pictures with the weight larger than that of the current clipped picture until the matching result of the clipped picture with the maximum weight is found, and sending the file name of the original picture of the picture server in the picture server 500 corresponding to the clipped picture with the maximum weight to the matching device 300.

Preferably, as shown in fig. 16 and 17 in conjunction, the result filtering means includes a reception judging unit 2003A, a maximum weight judging unit 2003B, a weight comparing unit 2003C, and a reception end judging unit 2003D. The receiving and judging unit 2003A is configured to judge whether a matching-successful picture is received, and if yes, output a receiving result; if not, the output matching fails. The maximum weight judging unit 2003B is configured to judge whether the successfully matched picture is the maximum weight according to the receiving result, and if so, output the clipped picture with the maximum weight as the recognition result; if not, outputting a judgment result. The weight comparison unit 2003C is configured to compare whether the successfully matched picture is the picture with the largest weight in the received pictures according to the determination result, and if so, output a comparison result; if not, the flow returns to the reception determination unit 2003A. The receiving end judging unit 2003D is configured to judge whether all the successfully matched pictures have been received according to the comparison result, and if yes, output the comparison result as an identification result; if not, the flow returns to the reception determination unit 2003A.

Specific embodiments of the results screening module 2003 are as follows:

if the matching results returned by the cut picture 101,201,102,202,103 sent by the picture feature matching module 2002 are result _ success _101, result _ fail _201, result _ success _102, result _ fail _202, and result _ fail _103, respectively, the result is arranged as result _ fail _103, result _ success _102, result _ fail _202, result _ success _101, and result _ fail _201 in sequence according to the sequence of the return (note: success in matching is indicated in the result, failure in matching is indicated in the result). Firstly, receive result _ fail _103, which is the result of failed match, record it to the memory directly, then receive result _ success _102, the result is a successful matching result, at this time, the result is read from the memory, and it is found that the clipped picture 101 and the clipped picture 201 with the weight larger than that of the clipped picture 102 have not returned a matching result, so that the result _ success _102 is also recorded into the memory, and the matching result between the clipped picture 101 and the clipped picture 201 is continuously waited for, then, it receives the result of failure matching of the cropped picture 202 to the memory, then, a result "result _ success _ 101" of successful matching of the cropped picture 101 is received, and the weight of the cut picture 101 is the largest, so the final matching result is result _ success _101, and the file name of the original picture of the picture server in the picture server 500 corresponding to the cut picture 101 is directly sent to the matching device 300 without waiting for the return of the matching result of the cut picture 201.

In step S300, the matching device 300 is connected to the picture recognition device 200, specifically to the result filtering module 2003, and is configured to sequentially match the file name of the original picture of the picture server sent by the result filtering module 2003 with the corresponding original picture, the compressed index packet and alias associated table, the original picture index table, and the chapter associated table, and output the successfully matched educational resource packet identifier, chapter name, and chapter number.

S300, the matching device 300 receives the file name of the original picture of the picture server sent by the picture recognition device 200, matches the file name with the corresponding original picture, the index compression packet and alias association table, the original picture index table and the chapter association table in sequence, and outputs the successfully matched education resource packet identification, chapter name and chapter number.

The concrete description is as follows:

s3001, receiving the original picture file name of the picture server from the result filtering module 2003, and querying and comparing the alias name in the original picture file name of the picture server with the index record in the corresponding "original picture and index compression packet and alias association table" to obtain the corresponding original picture and index compression packet identifier.

Specifically, the method comprises the following steps:

the file name of each picture server original picture is called as the 'picture server original picture file name' of the picture, and is called as the 'picture server original picture name' for short. Each picture server raw picture file name is unique in the picture server raw picture library module 30. The naming method of the original picture file name of the picture server comprises the following steps: the original picture and the alias of the index compressed packet corresponding to the original picture of the picture server + the original picture file name corresponding to the original picture of the picture server are obtained when the matching device 300 receives the original picture file name of the picture server sent by the result filtering module 2003, and then the original picture and the alias of the index compressed packet corresponding to the original picture of the picture server are found in the corresponding table 3 in the original picture and index compressed packet and alias association module 50, that is, the table "association table of the original picture and index compressed packet and alias" corresponding to the alias.

S3002, according to the original picture and the index compression packet identification, locating to the corresponding original picture index table, and then locating to the original picture page number or original picture file serial number in the corresponding index record on the original picture index table according to the original picture file name contained in the original picture name of the picture server.

Specifically, the method comprises the following steps:

because the original picture and index compression packet identification is the same as the corresponding original picture index table, namely table 1, the corresponding original picture index table is positioned in the original picture and index database module 20 according to the original picture and index compression packet identification; then, according to the original picture file name contained in the original picture name of the picture server, positioning the original picture page number or the original picture file serial number in the corresponding index record on an original picture index table;

s3003, according to the original picture and index compression packet identification, locating to the corresponding chapter association table, according to the original picture page number or original picture file serial number in the original picture index table, locating to the specific index record in the chapter association table, and then according to the specific index record in the chapter association table, locating to the corresponding education resource packet, chapter name and chapter serial number in the education resource library module 10.

Specifically, the method comprises the following steps:

because the original picture and index compression packet identification are identical to the corresponding original picture index table and chapter association table, the corresponding chapter association table, namely table 2, is positioned on the chapter association module 40 according to the original picture and index compression packet identification, the original picture page number or original picture file serial number in the original picture index table is positioned to a specific index record in the chapter association table according to the original picture and index compression packet identification, the original picture page number or the original picture file serial number in the original picture index table, the corresponding education resource packet, chapter name and chapter serial number in the education resource library module 10 are positioned according to the specific index record in the chapter association table, and the corresponding chapter content in the education resource packet and the resource learning content of the corresponding chapter page number in the education resource library module are positioned according to the education resource packet, chapter name and serial number.

It should be understood that the original picture index table, the chapter association table, the original picture and index compression packet and alias association table, the original picture and index packet, the original picture and index compression packet and the education resource packet corresponding to the same paper textbook have the same file name, and the same file name does not conflict because the storage paths are different.

The following description will be made with respect to the structures of "original picture and index compression package and alias association table", "original picture index table", and "chapter association table", respectively:

each paper textbook corresponds to an original picture index table, namely table 1, and is used for recording the index of each picture in an original picture clip associated with the textbook. The original picture index table may be an Excel-type file. The file name of the original picture index table is the same as the file names of the corresponding original picture and index compression packet and the original picture and index packet, so that when the file name of the original picture of the picture server is found, the alias in the file name of the original picture of the picture server is obtained, the corresponding original picture and index compression packet identification is obtained, and then the corresponding original picture index table is quickly found through the original picture and index compression packet identification. The naming method of the original picture index table comprises the following steps: resource type + textbook version + school lesson + subject + grade + school time + others.

Table 1 is a common format of the original picture index table, and table 1' is the original picture index table corresponding to the original picture and index compression package identified as "teaching and assistance questions-synchronous examination paper-human teaching five and four plate making-primary school-mathematics-six grades-upper-teaching material full explanation reference big whole-17 years of autumn-shooting and searching". As shown in table 1, the "original picture index table" contains information of "original picture index table file name, original picture file serial number, original picture and index compression packet identification, original picture page number, original picture file name, original picture URL, original picture MD5, picture server original picture file name, status, creator, creation date, modifier, and modification time".

TABLE 1

Table 1'

In table 1, the file name of the original image index table is the identifier of the original image index table corresponding to each paper textbook, and the type of the identifier field is big 16. The file name of the original picture index table corresponding to each paper textbook is unique in the original picture and index compression package. As mentioned above, the identifier is the same as the file name of the "original picture and index compressed packet" corresponding to the identifier, that is: resource type + textbook version + school lesson + subject + grade + school time + others.

The original picture file serial number refers to the serial number of each original picture in the original picture folder, the serial number is also regarded as the index serial number recorded in the original picture index table, and the field type of the serial number is VARCHAR (100). The serial number of each picture is unique within the original picture folder.

The original picture and index compression packet identifier can be understood as the file name of the original picture and index compression packet corresponding to each paper textbook, and the identifier field type is VARCHAR (400). The identifier is the same as the identifier of the original picture index table, and the naming method thereof is mentioned previously, and is not described herein again.

The original picture page number field type is INT (10), and comprises an inner page scanning picture page number and a non-inner page scanning picture page number, wherein the inner page scanning picture page number is consistent with the page number of a corresponding text content page in a corresponding paper textbook, and the top is not closed from 1; the page numbers of the directory scanning pictures are sequentially increased by 1 from a negative number until the page number of the last directory page is 0; the page numbers of other non-inner page scanning pictures such as front cover, back cover and the like are uniformly default to be-1000'.

The original picture file name field type is VARCHAR (200), which is mentioned above, and comprises an inner page scanning picture file name and a non-inner page scanning picture file name, wherein the inner page scanning picture file name is consistent with the page number of the corresponding text content page in the corresponding paper textbook; the non-inner page scanning picture file name is the same as the subject content, such as front cover, catalogue, postscript, back cover and the like.

The original picture URL refers to a link address of the original picture in the search engine server 600, and its field type is VARCHAR (300).

The original picture MD5, i.e. Message-Digest 5 information-Digest Algorithm 5, has a field type of VARCHAR (100), and is used to ensure the complete consistency of the information after the original picture is uploaded to the search engine server 600.

The file name of the original picture of the picture server, that is, the file name of the original picture of the picture server in the original picture library module 30 of the picture server, is formed by uploading an original picture corresponding to a paper textbook previously made by a manufacturer on a local device to the picture server 500, and the file name of the original picture of the picture server is unique in the original picture library module 30 of the picture server, and the field type of the file name is VARCHAR (100). The naming method of the original picture file name of the picture server comprises the following steps: the original picture to which the local original picture corresponding to the original picture of the picture server belongs, the alias of the compressed packet of the index library and the file name of the local original picture corresponding to the original picture of the picture server are as follows: the first part of the file name of the original picture of the picture server is the original picture and the alias of the index library compression packet to which the corresponding local original picture belongs, and the second part of the file name of the local original picture corresponding to the original picture of the picture server. If "PSCHINESEG 110245 cover" is the file name of the original picture of the picture server, where "PSCHINESEG 110245" is the first part, i.e. the alias part, and "cover" is the second part, i.e. the file name of the local original picture corresponding to the original picture of the picture server, i.e. the cover of the corresponding paper textbook; also, for example, in "PSCHINESEGG 11024510 1102451", the part "PSCHINESEG 110245" is an alias, and 10 is the original picture file name, i.e., page 10 of the inner page of the corresponding paper textbook.

As mentioned above, each original picture and index compression packet has a unique alias, which can also be used as a prefix of the file name of the original picture of the picture server. The naming rule of the alias is as follows: the method comprises the steps of using a resource type, a subject name, an original picture and an index compression packet ID number, namely using the resource type of an education resource packet corresponding to each paper textbook as a first part, using the subject to which the paper textbook belongs as a second part, using the original picture and the index compression packet ID number corresponding to the paper textbook as a third part, and combining the three parts according to the sequence to form the alias of the original picture and the index compression packet. If PS in the alias "PSCHINESEG 110245" indicates that the resource type is a textbook reading resource, CHINESE is a subject, and G110245 is the original picture and the index compression packet ID number corresponding to the paper textbook. It should be understood that the alias is also the alias for each paper textbook and its corresponding original pictures and index packets.

The states include valid and invalid states, which means whether the state of the index record in the original picture index table is valid, 1 is valid, and 0 is invalid. The field type is CHAR (2).

The creator, creation date, modifier and modification time refer to the related creation and modification information of the index, which is convenient for verification, statistics and updating, and are not described herein.

It should be noted that in the original picture index table, that is, the name of the original picture file of the picture server in table 1, all the local original pictures corresponding to the paper textbook are uploaded to the picture server 500 to generate the corresponding original picture of the picture server, and then the name of the original picture file of the picture server is written back into the original picture index table. The write back process is prior art and will not be described herein.

It should be understood that the original pictures and the index packets in the original picture and index library module 20 are first produced by the manufacturer in the local device, and then compressed into original pictures and index compression packets, and then uploaded to the search engine server 600, and then decompressed into original pictures and index packets, and stored in the search engine server 600, wherein the original pictures and index packets can be added, deleted and updated.

As shown in fig. 12, the chapter association module 40 is configured to store a plurality of chapter association tables, namely table 2, where the chapter association tables include a corresponding relationship between each chapter in the paper textbook and the original picture. The chapter association module 40 is stored in the search engine server 600 and is formed by a collection of a plurality of chapter association tables, i.e., table 2. The original picture folder corresponding to each paper textbook corresponds to a chapter association table, the chapter association table establishes association relationship between the chapters and the inner page scanned pictures in the original picture folder through the original picture file names, namely page numbers, of the inner page scanned pictures, and the chapter association table is used for describing all original pictures contained in each chapter.

The chapter association table can be an Excel-type file, as shown in fig. 1, the storage paths of the original picture index table and the chapter association table are different, so the chapter association table and the original picture index table are the same name, that is, the chapter association table and the corresponding original picture and index compression package are the same name, and since the chapter association module 40 comprises a plurality of chapter association tables, the corresponding chapter association table can be quickly located through the names of the original picture and index compression package or the names of the original picture index table, so that the retrieval efficiency is improved.

Table 2 is a common format of the chapter association table, and table 2' is the chapter association table corresponding to a paper textbook labeled "teaching and assistant questions, synchronous examination paper, five and four plate making by human education, primary school, mathematics, six grades, upper school, full explanation reference of teaching materials, full-17 years of autumn-shooting and searching". As shown in table 2, the chapter association table includes: the method comprises the steps of original picture and index compression packet identification, page numbers corresponding to chapter start pages, page numbers corresponding to chapter end pages, index serial numbers corresponding to original pictures of chapter start pages, index serial numbers corresponding to original pictures of chapter end pages, education resource packet identification, chapter names, chapter serial numbers, states, creators, creation dates, modifiers and modification time information.

TABLE 2

TABLE 2'

In table 2:

the original picture and index compression packet is identified as the file name of the original picture and index compression packet corresponding to each paper textbook, and the identification field type is VARCHAR (400). It should be understood that the original picture and index compression packet id in table 2 is the original picture and index compression packet id in table 1, so that the same paper textbook content can associate table 1 with table 2 and table 3 through the original picture and index compression packet id.

The page number corresponding to the chapter starting page refers to the page number of the paper textbook corresponding to the chapter starting page, and the field type is BIGINT (16).

The page number corresponding to the chapter ending page refers to the page number of the paper textbook to which the chapter ending page belongs, and the field type is BIGINT (16).

The index serial number of the original picture corresponding to the chapter start page refers to the index serial number of the corresponding original picture in the paper textbook to which the chapter start page belongs in the corresponding original picture index table, namely the original picture file serial number in table 1. Therefore, the same paper textbook content can establish the association relationship between table 1 and table 2 by the index number of the original picture corresponding to the starting page of the chapter.

The index serial number of the original picture corresponding to the chapter ending page refers to the index serial number of the corresponding original picture in the paper textbook to which the chapter ending page belongs in the corresponding original picture index table, namely the original picture file serial number in table 1. Therefore, the same paper textbook content can establish the association relationship between table 1 and table 2 through the index sequence number of the original picture corresponding to the chapter ending page.

The education resource package identification refers to the identification of the education resource package corresponding to the paper textbook, and the identification field type is VARCHAR (400). The naming method of the education resource package identifier is described above and will not be described herein.

The chapter name refers to the name of the chapter in a paper textbook corresponding to the chapter, and the field type of the name is VARCHAR (200).

The chapter number refers to the serial number of the chapter in the paper textbook corresponding to the chapter, and the serial number type is INT (16).

The state includes two states of valid and invalid, and means whether the state of the record in the chapter association table is valid or not. Its field type is CHAR (2), 1 is valid and 0 is invalid.

The creator, the creation date, the modifier, and the modification time refer to related creation and modification information of the content associated with the chapter, which is convenient for verification, statistics, and updating, and are not described herein again.

It should be understood that the table is used to establish the association relationship between the inner page of the paper textbook, i.e. the front content page, and the chapters, so as to locate the chapters in the educational resource package corresponding to the scanned picture of the user, regardless of the non-inner page front cover, back cover, catalog, etc.

It should be understood that the chapter association module 40 is formed by a plurality of chapter association tables previously created by the manufacturer on the local device and uploaded to the search engine server 600.

Referring to fig. 13, the original picture and index compressed packet and alias association module 50 is configured to store a plurality of original picture and index compressed packet and alias association tables, where the original picture and index compressed packet and alias association tables include the corresponding relationship between the identifier of the original picture and index compressed packet and the alias. The original picture and index compression packet and alias association module 50 are stored on the search engine server 600. The original picture and index compression packet and alias association table is a table formed by establishing a one-to-one correspondence relationship between the original picture and index compression packet identifier and the alias of the original picture and index compression packet.

The original picture and the index compression package and alias association table can be an EXCEL type table file, and can also be automatically generated by a program. As can be seen from fig. 1, the storage paths of the original picture and index compression packet and alias association table, the original picture index table, and the chapter association table are different, so that the original picture and index compression packet and alias association table is the same as the chapter association table and the original picture index table, that is, the original picture and index compression packet is the same as the original picture and index compression packet, so that the original picture and index compression packet and alias association table can be quickly located by the name of the original picture and index compression packet, the name of the original picture index table, or the name of the chapter association table, and the retrieval efficiency is improved.

Table 3 is a common format of the original picture and index compression package and alias association table, and table 3' is a partial example content of the original picture and index compression package and alias association table. As shown in table 3, the "original picture and index compressed packet and alias association table" includes: the picture server original picture name, the picture server original picture identification, the original picture and index compression packet identification, the education resource type, the alias, the state, the creator, the creation date, the modifier and the modification time information.

TABLE 3

Table 3'

In table 3:

the name of the original picture of the picture server, i.e. the name of the file of the original picture of the picture server, is unique in the original picture library module 30 of the picture server, and as mentioned above, the naming method of the name is as follows: the original picture to which the original picture corresponding to the original picture of the picture server belongs, the alias of the index compression packet and the file name of the original picture corresponding to the original picture of the picture server. It should be understood that the file names of the original pictures in the corresponding original pictures and the index packets are divided into the file names of the intra-page scanned pictures and the file names of the non-intra-page scanned pictures, the file names of the intra-page scanned pictures are consistent with the page numbers of the corresponding paper textbook contents, the file names of the non-intra-page scanned pictures are consistent with the page subject contents, and when the original pictures are uploaded to the picture server original picture library module 30, the alias of the original picture and the index compressed packet to which the original pictures belong is added in front of the file names of the original pictures as a prefix to form the corresponding picture server original picture file names. As shown in "PSCHINESEG 110245" in "PSCHINESEG 110245_ 1" in table 3', the alias of the corresponding original picture and the index compression packet, that is, the prefix of the corresponding original picture file name, and "1" is the original picture file name. This field type is BIGINT (16). The original picture name of the picture server in table 3 is the original picture file name of the picture server in table 1, so that the table 1 and table 3 can be associated with each other through the original picture name of the picture server in the same paper textbook content.

The picture server original picture identification is an integer type of serial number that is unique in the "picture server original picture library module 30".

The original picture and index compression packet is identified as the file name of the original picture and index compression packet corresponding to each paper textbook, and the identification field type is VARCHAR (400). It should be understood that the original picture and index compression packet id in table 3 is the same as the original picture and index compression packet id in table 2 and table 1, so that the same paper textbook content can associate table 3 with table 1 and table 2 through the original picture and index compression packet id.

Education resource types including textbook point reading resources, test question shooting resources and the like can be represented by different serial numbers, and the field is used for counting the number of the resources of various types. This field type is BIGINT (16).

The alias is the unique alias of each original picture and the index compression packet, and can also be regarded as the alias of the paper textbook, the original picture and the index packet, and the alias is used as the prefix of the file name of the original picture of the picture server. The naming method of the alias is as follows: the method comprises the steps of using a resource type, a subject name, an original picture and an index compression packet ID number, namely using the resource type of an education resource packet corresponding to each paper textbook as a first part, using the subject to which the paper textbook belongs as a second part, using the original picture and the index compression packet ID number corresponding to the paper textbook as a third part, and combining the three parts according to the sequence to form the alias of the original picture and the index compression packet. If the first part "PS" in the alias "PSCHINESEG 110245" indicates that the resource type is a textbook point-reading resource, the second part "CHINESE" is a language subject, and the third part G110245 is the original picture and the index compression package ID number corresponding to the paper textbook. The field type is VARCHAR (100). It should be understood that the original picture file name of the picture server in table 3 is obtained by removing the prefix, i.e. the alias, from the original picture name of the picture server in table 3, i.e. the original picture file name in table 1, so that the table 3 and table 1 can be associated by removing the alias from the original picture name of the picture server in table 3 to obtain the original picture file name in table 1.

The states include a valid state and an invalid state, which refer to whether the state of the record in the original picture and index compression packet and alias association table is valid, 1 is valid, 0 is invalid, and the field type is CHAR (2).

The creator, creation date, modifier and modification time refer to the related creation and modification information of the record in the original picture and index compression packet and alias association table, which is convenient for verification, statistics and updating, and are not described herein again.

It should be understood that the original picture and index compression package and alias association module 50 may be formed by the manufacturer by first making the "original picture and index compression package and alias association table" on the local device and uploading the table to the search engine server 600.

In step S400, the output device 400 locates the corresponding chapter content and the resource learning content corresponding to the page number in the corresponding education resource package in the education resource library module according to the education resource package, the chapter name and the chapter number, and outputs the corresponding chapter content and the resource learning content corresponding to the page number to the client receiving and displaying device 900.

Specifically, the corresponding chapter content and the resource learning content of the corresponding page number in the corresponding education resource packet in the education resource library module are located according to the education resource packet, the chapter name and the chapter number, and the corresponding page resource learning content comprises one or more types of analysis of text of a textbook, explanation of example questions, explanation of related knowledge points, synchronous training, teaching and assisting questions, test paper, exercise, video link, Flash link, interesting game link and the like.

In step S500, the client receiving and displaying device 900 receives and displays the chapter content and the resource learning content corresponding to the page number output by the output device 400 to the user.

It is worth mentioning that for the situation that the content of the chapters of the paper textbook is less than one page, a chapter association table is added for positioning the chapters, so that the positioning accuracy is improved. When the education resource package identification and the specific chapter information corresponding to the chapter association table are obtained, the chapter content of the corresponding education resource package and the resource learning content of the corresponding page can be located by combining the original picture page number information in the original picture index table.

It should be noted that if the original picture page number in the original picture index table is negative or 0, the first index record in the chapter association table with the same name as the original picture index table is directly located, and the first page learning content of the corresponding education resource package is located according to the first index record. It should be understood that the original picture page number being a negative number or 0 herein means that the original picture is a non-intra scanned picture, i.e., non-text content such as front cover, catalog, postscript, back cover, etc.

It is worth mentioning that the original picture index table, the chapter association table, the original picture and index compression packet and alias association table, the original picture and index packet, the original picture and index compression packet and the education resource packet corresponding to the same paper textbook have the same file name.

It is noted that, alternatively, in some embodiments, the system further includes a general server as a collection of the picture server 500 and the search engine server 600, in which the picture server original picture library module 30, the educational resource library module 10, the original picture and index library module 20, the chapter association module 40, the original picture compression package and alias association module 50 are stored.

In the method in this embodiment, other parts correspond to the systems in the foregoing embodiments, and are not described herein again.

The above description is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments, and all technical solutions belonging to the idea of the present invention belong to the protection scope of the present invention. It should be noted that several modifications and adaptations to those skilled in the art without departing from the principles of the present invention should also be considered within the scope of the present invention.

Claims

1. A resource acquisition device based on picture identification is characterized by comprising

A client picture acquisition device (100) for acquiring an external picture;

the picture identification device (200) is connected with the client picture acquisition device (100) and is used for carrying out picture identification according to the acquired external picture and outputting an identification result;

the matching device (300) is connected with the picture recognition device (200) and used for selectively matching data according to the recognition result and outputting a matching result;

the output device (400) is connected with the matching device (300) and is used for outputting the corresponding resource content in the corresponding education resource package according to the matching result;

wherein the picture recognition device (200) comprises

The picture preprocessing module (2001) is used for cutting the external picture into a plurality of cut pictures and setting the priority and the weight information of each cut picture, and the picture preprocessing module (2001) is connected with the client picture acquisition device (100);

the picture characteristic matching module (2002) is used for carrying out characteristic matching on each cut picture and outputting a picture which is successfully matched;

a result screening module (2003) for receiving the successfully matched picture and outputting the identification result;

wherein the picture pre-processing module (2001) comprises

A first cropping unit (2001A) for subtracting a first non-effective region from each of left and right sides of the external picture along an X-axis and generating a first cropped picture, wherein the ratio of the sum of the widths of the first non-effective regions on the left and right sides to the width of the external picture is 5-45%, and the ratio of the width of the first cropped picture to the width of the external picture is 55-95%;

a second cropping unit (2001B) for subtracting a second non-effective region from both upper and lower sides of the external picture along the Y axis and generating a second cropped picture, wherein the ratio of the sum of the heights of the second non-effective regions on the upper and lower sides to the height of the external picture is 5-45%, and the ratio of the height of the second cropped picture to the height of the external picture is 55-95%;

the matching device (300) is connected with the picture recognition device (200) and used for selectively matching data according to the recognition result, and comprises:

according to the alias in the original picture name of the picture server, inquiring and comparing in an original picture and index compression packet and alias association table in an original picture and index compression packet and alias association module (50) to obtain a corresponding original picture and index compression packet identifier;

according to the original picture and index compression packet identification, positioning a corresponding original picture index table in an original picture and index library module (20); positioning corresponding index records on the original picture index table according to original picture file names contained in the original picture names of the picture server;

positioning a corresponding chapter association table on a chapter association module (40) according to the original picture and index compression packet identification, and positioning a specific index record in the chapter association table according to the original picture and index compression packet identification and an original picture page number or an original picture file serial number in the original picture index table;

positioning the corresponding education resource packages, chapter names and chapter sequence numbers in an education resource library module (10) according to the specific index records in the chapter association table;

positioning corresponding chapter contents and corresponding page numbers of the corresponding chapter contents in the education resource library module (10) according to the education resource packages, the chapter names and the chapter sequence numbers, wherein the corresponding page numbers of the resource learning contents comprise one or more types of analysis of text texts, explanation of example questions, explanation of related knowledge points, synchronous training, teaching and assisting questions, test papers, exercises, video links, Flash links and interesting game links;

the original picture index table, the chapter association table, the original picture and index compression packet and alias association table, the original picture and index packet, the original picture and index compression packet and the education resource packet corresponding to the same paper textbook have the same file name.

2. The apparatus of claim 1, wherein the picture pre-processing module (2001) further comprises

And the third cropping unit (2001C) is used for subtracting a third non-effective area from the far end of the external picture along the Y axis and generating a third cropping picture, wherein the height ratio of the third non-effective area to the external picture is 25-45%, and the height ratio of the third cropping picture to the external picture is 55-75%.

3. The apparatus according to claim 2, wherein the picture pre-processing module (2001) further comprises a fifth cropping unit (2001E) for subtracting a fifth non-active area from the external picture along the Y-axis and generating a fifth cropped picture, wherein the height ratio of the fifth non-active area to the external picture is 46% -60%, and the height ratio of the fifth cropped picture to the external picture is 40% -54%.

4. The apparatus according to any of claims 1-3, wherein the result screening module (2003) comprises

A receiving judgment unit (2003A) for judging whether the matching success picture is received or not, if yes, outputting a receiving result; if not, outputting matching failure;

a maximum weight judging unit (2003B) for judging whether the matching success picture is the maximum weight according to the receiving result, if so, outputting the clipping picture with the maximum weight as the identification result; if not, outputting a judgment result;

a weight comparison unit (2003C) for comparing whether the matching success picture is the picture with the maximum weight in the received pictures according to the judgment result, if so, outputting the comparison result; if not, returning to the receiving judgment unit (2003A);

a receiving end judging unit (2003D) which judges whether all the matching success pictures are received according to the comparison result, if so, the comparison result is output as the identification result; if not, the processing returns to the receiving judgment unit (2003A).

5. A resource acquisition method based on picture identification is characterized by comprising

Acquiring an external picture;

receiving the successfully matched picture and outputting the identification result;

wherein the cropping the external picture into a plurality of cropped pictures comprises

subtracting a second non-effective area from the upper side and the lower side of the external picture along the Y axis to generate a second cut picture, wherein the height ratio of the sum of the heights of the second non-effective areas on the upper side and the lower side to the external picture is 5-45%, and the height ratio of the second cut picture to the external picture is 55-95%;

wherein the selectively performing data matching according to the recognition result and outputting the matching result includes:

6. The method according to claim 5, wherein said cropping the external picture into a plurality of cropped pictures further comprises

And subtracting a third non-effective area from the far end of the external picture along the Y axis to generate a third cut picture, wherein the height ratio of the third non-effective area to the external picture is 25-45%, and the height ratio of the third cut picture to the external picture is 55-75%.

7. The method of claim 6, wherein the cropping the external picture into a plurality of cropped pictures further comprises: and subtracting a fifth non-effective area from the far end of the external picture along the Y axis to generate a fifth cut picture, wherein the height ratio of the fifth non-effective area to the external picture is 46% -60%, and the height ratio of the fifth cut picture to the external picture is 40% -54%.

8. The method according to any one of claims 5-7, wherein the receiving the matching success picture and outputting the recognition result comprises: