CN117253244A - Book data processing method and equipment - Google Patents
Book data processing method and equipment Download PDFInfo
- Publication number
- CN117253244A CN117253244A CN202311074432.2A CN202311074432A CN117253244A CN 117253244 A CN117253244 A CN 117253244A CN 202311074432 A CN202311074432 A CN 202311074432A CN 117253244 A CN117253244 A CN 117253244A
- Authority
- CN
- China
- Prior art keywords
- book
- image
- back cover
- target
- front cover
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 48
- 239000011159 matrix material Substances 0.000 claims description 92
- 230000009466 transformation Effects 0.000 claims description 63
- 239000013598 vector Substances 0.000 claims description 45
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 38
- 238000001514 detection method Methods 0.000 claims description 34
- 230000006870 function Effects 0.000 claims description 24
- 238000012544 monitoring process Methods 0.000 claims description 23
- 238000003860 storage Methods 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 14
- 238000013519 translation Methods 0.000 claims description 8
- 238000010586 diagram Methods 0.000 description 16
- 230000008569 process Effects 0.000 description 15
- 238000012545 processing Methods 0.000 description 14
- 238000012937 correction Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 8
- 238000009826 distribution Methods 0.000 description 6
- 238000012015 optical character recognition Methods 0.000 description 5
- 230000036544 posture Effects 0.000 description 5
- 238000012549 training Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001627 detrimental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/40—Document-oriented image-based pattern recognition
- G06V30/41—Analysis of document content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
- G06K7/10—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation
- G06K7/10544—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation by scanning of the records by radiation in the optical part of the electromagnetic spectrum
- G06K7/10821—Methods or arrangements for sensing record carriers, e.g. for reading patterns by electromagnetic radiation, e.g. optical sensing; by corpuscular radiation by scanning of the records by radiation in the optical part of the electromagnetic spectrum further details of bar or optical code scanning devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/16—Image preprocessing
- G06V30/1607—Correcting image deformation, e.g. trapezoidal deformation caused by perspective
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/13—Type of disclosure document
- G06V2201/131—Book
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Electromagnetism (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Toxicology (AREA)
- Processing Or Creating Images (AREA)
Abstract
The embodiment of the application provides a book data processing method and device, wherein the method comprises the following steps: acquiring at least one frame of target image, wherein the target image comprises a front cover and a back cover of a book to be recorded; correcting the shape of the front cover and the shape of the back cover in the target image respectively, and determining the corrected front cover and back cover as a front cover image and a back cover image of the book to be recorded respectively; detecting a bar code corresponding to a book to be recorded in a target image, and acquiring book information of the book to be recorded according to the bar code; and storing the front cover image, the back cover image and the book information into a preset database. The embodiment of the application solves the technical problem that the related information of the book is difficult to enter into the database rapidly and accurately in the prior art.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a book data processing method and apparatus.
Background
With the rapid development of information technology, libraries are gradually changed to intelligent and digital directions. The fusion application of the intelligent library construction and the artificial intelligence technology is a necessary trend of library development and is also an important way for improving the service quality of libraries.
In intelligent library business, an important function is bookshelf digitalization, and bookshelf digitalization can push favorite books to readers according to big data, so that service experience is improved, meanwhile, an administrator can know book browsing frequency more accurately, and bookshelf layout is optimized.
The bookshelf is digitalized, related information of books is required to be input into a database in advance and used as a base for book retrieval and statistical analysis. However, how to quickly and accurately enter related information of books into a database is a technical problem to be solved at present.
Disclosure of Invention
Aspects of the application provide a book data processing method and device, which can solve the technical problem that related information of books is difficult to be quickly and accurately input into a database in the prior art.
In a first aspect, an embodiment of the present application provides a book data processing method, including:
acquiring at least one frame of target image, wherein the target image comprises a front cover and a back cover of a book to be recorded;
correcting the shape of the front cover and the shape of the back cover in the target image respectively, and determining the corrected front cover and back cover as a front cover image and a back cover image of the book to be recorded respectively; wherein the front cover image and the back cover image are regular polygons;
Detecting a bar code corresponding to the book to be recorded in the target image, and acquiring book information of the book to be recorded according to the bar code;
and storing the front cover image, the back cover image and the book information into a preset database.
In one possible implementation manner, the correcting the shape of the front cover and the shape of the back cover in the target image respectively includes:
detecting the vertex coordinates of the front cover and the vertex coordinates of the back cover in the target image respectively based on a preset arbitrary polygon detection model;
determining the target vertex coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the back cover;
determining an affine transformation matrix corresponding to the cover according to the vertex coordinates of the cover and the target vertex coordinates; determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates;
correcting the shape of the front cover in the target image by using an affine transformation matrix corresponding to the front cover, and correcting the shape of the back cover in the target image by using an affine transformation matrix corresponding to the back cover.
In one possible implementation manner, the determining the target vertex coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the back cover includes:
respectively determining the center point coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the vertex coordinates of the back cover;
determining a true vector and a predicted vector from the center point of the front cover to the top point according to the top point coordinates and the center point coordinates of the front cover, and determining a true vector and a predicted vector from the center point of the back cover to the top point according to the top point coordinates and the center point coordinates of the back cover;
determining target vertex coordinates of the cover based on a true vector and a predicted vector from a center point to a vertex of the cover and a loss function preset by the cover; and determining the target vertex coordinates of the back cover based on the true vector and the predicted vector from the central point to the vertex of the back cover and the preset loss function of the back cover.
In one possible implementation manner, the affine transformation matrix corresponding to the cover is determined according to the vertex coordinates of the cover and the target vertex coordinates; and determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates, wherein the affine transformation matrix comprises the following components:
Determining a transformation matrix corresponding to the front cover according to the vertex coordinates of the front cover and the target vertex coordinates, and determining a transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates; wherein the transformation matrix comprises at least one of the following matrices: a reflection matrix, a rotation matrix, a scaling matrix, a beveling matrix, and a translation matrix;
and determining an affine transformation matrix corresponding to the front cover according to the transformation matrix corresponding to the front cover, and determining an affine transformation matrix corresponding to the back cover according to the transformation matrix corresponding to the back cover.
In one possible implementation manner, the acquiring at least one frame of the target image includes:
receiving at least one frame of book image uploaded by terminal equipment, and taking the book image as the target image; the book image is an image of the book to be recorded, which is collected by the terminal equipment and placed in any posture.
In one possible implementation manner, the detecting the bar code corresponding to the book to be recorded in the target image, and obtaining the book information of the book to be recorded according to the bar code includes:
Detecting the position information of the bar code corresponding to the book to be recorded in the target image;
intercepting the bar code in the target image according to the position information;
reading the number corresponding to the bar code, and searching the book information of the book to be recorded in a preset book index library according to the number; the book information includes at least one of the following information: title, publisher, author, publication date, pricing.
In one possible embodiment, the method further comprises:
when a search instruction is received, acquiring a search keyword in the search instruction;
searching a matched book matched with the search keyword in the database;
and outputting the front cover image, the back cover image and the book information of the matched books stored in the database.
In one possible embodiment, the method further comprises:
based on a book ReID model, depth characteristic information of the front cover image and depth characteristic information of the back cover image are respectively obtained;
and storing the depth characteristic information of the front cover image and the depth characteristic information of the back cover image into the database.
In one possible embodiment, the method further comprises:
Acquiring a monitoring image and detecting whether a target book in a moving state exists in the monitoring image;
when a target book in a moving state exists in the monitoring image, acquiring a target book image corresponding to the target book in the monitoring image;
based on a book ReID model, acquiring depth characteristic information of the target book image;
searching target depth characteristic information matched with the depth characteristic information of the target book image in the stored depth characteristic information in the database;
and according to the target depth characteristic information, acquiring book information of the target book in the database, and adding the book information of the target book in a book reference record.
In one possible implementation manner, the obtaining depth feature information of the target book image based on the book ReID model includes:
correcting the shape of the target book image;
and acquiring depth characteristic information of the corrected target book image based on the book ReID model.
In one possible embodiment, the method further comprises:
counting the consulting times of each book in a preset time according to the book consulting record;
And generating attention degree information of each book according to the consulting times.
In a second aspect, an embodiment of the present application provides a book data processing apparatus, including:
the acquisition module is used for acquiring at least one frame of target image, wherein the target image comprises a front cover and a back cover of a book to be recorded;
the correction module is used for respectively correcting the shape of the front cover and the shape of the back cover in the target image, and respectively determining the corrected front cover and back cover as a front cover image and a back cover image of the book to be recorded; wherein the front cover image and the back cover image are regular polygons;
the bar code identification module is used for detecting a bar code corresponding to the book to be recorded in the target image and acquiring book information of the book to be recorded according to the bar code;
and the storage module is used for storing the front cover image, the back cover image and the book information into a preset database.
In one possible implementation, the correction module is configured to:
detecting the vertex coordinates of the front cover and the vertex coordinates of the back cover in the target image respectively based on a preset arbitrary polygon detection model;
Determining the target vertex coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the back cover;
determining an affine transformation matrix corresponding to the cover according to the vertex coordinates of the cover and the target vertex coordinates; determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates;
correcting the shape of the front cover in the target image by using an affine transformation matrix corresponding to the front cover, and correcting the shape of the back cover in the target image by using an affine transformation matrix corresponding to the back cover.
In one possible implementation, the correction module is configured to:
respectively determining the center point coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the vertex coordinates of the back cover;
determining a true vector and a predicted vector from the center point of the front cover to the top point according to the top point coordinates and the center point coordinates of the front cover, and determining a true vector and a predicted vector from the center point of the back cover to the top point according to the top point coordinates and the center point coordinates of the back cover;
Determining target vertex coordinates of the cover based on a true vector and a predicted vector from a center point to a vertex of the cover and a loss function preset by the cover; and determining the target vertex coordinates of the back cover based on the true vector and the predicted vector from the central point to the vertex of the back cover and the preset loss function of the back cover.
In one possible implementation, the correction module is configured to:
determining a transformation matrix corresponding to the front cover according to the vertex coordinates of the front cover and the target vertex coordinates, and determining a transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates; wherein the transformation matrix comprises at least one of the following matrices: a reflection matrix, a rotation matrix, a scaling matrix, a beveling matrix, and a translation matrix;
and determining an affine transformation matrix corresponding to the front cover according to the transformation matrix corresponding to the front cover, and determining an affine transformation matrix corresponding to the back cover according to the transformation matrix corresponding to the back cover.
In one possible implementation, the obtaining module is configured to:
receiving at least one frame of book image uploaded by terminal equipment, and taking the book image as the target image; the book image is an image of the book to be recorded, which is collected by the terminal equipment and placed in any posture.
In one possible implementation, the bar code identification module is used to:
detecting the position information of the bar code corresponding to the book to be recorded in the target image;
intercepting the bar code in the target image according to the position information;
reading the number corresponding to the bar code, and searching the book information of the book to be recorded in a preset book index library according to the number; the book information includes at least one of the following information: title, publisher, author, publication date, pricing.
In a possible implementation manner, the apparatus further includes a retrieval module, configured to:
when a search instruction is received, acquiring a search keyword in the search instruction;
searching a matched book matched with the search keyword in the database;
and outputting the front cover image, the back cover image and the book information of the matched books stored in the database.
In a possible implementation manner, the apparatus further includes a feature processing module, configured to:
based on a book ReID model, depth characteristic information of the front cover image and depth characteristic information of the back cover image are respectively obtained;
And storing the depth characteristic information of the front cover image and the depth characteristic information of the back cover image into the database.
In a possible implementation manner, the apparatus further includes a reference detection module, configured to:
acquiring a monitoring image and detecting whether a target book in a moving state exists in the monitoring image;
when a target book in a moving state exists in the monitoring image, acquiring a target book image corresponding to the target book in the monitoring image;
based on a book ReID model, acquiring depth characteristic information of the target book image;
searching target depth characteristic information matched with the depth characteristic information of the target book image in the stored depth characteristic information in the database;
and according to the target depth characteristic information, acquiring book information of the target book in the database, and adding the book information of the target book in a book reference record.
In one possible implementation, the reference detection module is configured to:
correcting the shape of the target book image;
and acquiring depth characteristic information of the corrected target book image based on the book ReID model.
In a possible implementation manner, the apparatus further includes a statistics module, configured to:
counting the consulting times of each book in a preset time according to the book consulting record;
and generating attention degree information of each book according to the consulting times.
In a third aspect, an embodiment of the present application provides an electronic device, including: a memory and a processor;
the memory stores computer-executable instructions;
the processor executes the computer-executable instructions stored in the memory, so that the processor executes the book data processing method according to the first aspect.
In a fourth aspect, embodiments of the present application provide a computer-readable storage medium, where computer-executable instructions are stored, where the computer-executable instructions, when executed by a computer, are configured to implement the book data processing method according to the first aspect.
In a fifth aspect, embodiments of the present application provide a computer program product, which includes a computer program, where the computer program is executed by a computer to implement the book data processing method according to the first aspect.
According to the book data processing method and device, after the target image is obtained, the shape of the front cover and the shape of the back cover in the target image are corrected, so that the front cover image and the back cover image of the book to be input can be corrected into regular polygons, and the standardized storage and the characteristic retrieval of the front cover of the subsequent book are facilitated; meanwhile, by detecting the bar code corresponding to the book to be recorded in the target image and acquiring the book information of the book to be recorded according to the bar code, the accuracy of the acquired book information can be ensured, and the technical problem that the related information of the book is difficult to be recorded in a database rapidly and accurately in the prior art is solved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:
fig. 1 is a schematic architecture diagram of an application system according to an exemplary embodiment of the present application;
FIG. 2 is a schematic flow chart of a book data processing method according to an exemplary embodiment of the present application;
FIG. 3 is a schematic diagram of a process for obtaining a front cover image or a back cover image from a book image according to an embodiment of the present application;
FIG. 4 is a schematic diagram of another process for obtaining a front cover image or a back cover image from a book image according to an embodiment of the present application;
FIG. 5 is another flow chart of a book data processing method according to an exemplary embodiment of the present application;
FIG. 6 is a schematic view of another process of the book data processing method according to the exemplary embodiment of the present application;
FIG. 7 is a schematic diagram of a book data processing apparatus according to an exemplary embodiment of the present application;
fig. 8 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application.
Detailed Description
For the purposes, technical solutions and advantages of the present application, the technical solutions of the present application will be clearly and completely described below with reference to specific embodiments of the present application and corresponding drawings. It will be apparent that the described embodiments are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
The term "module" as used in the embodiments of the present application refers to any known or later developed hardware, software, firmware, artificial intelligence, fuzzy logic, or combination of hardware or/and software code that is capable of performing the function associated with that element.
It should be noted that, the user information (including but not limited to user image information, user personal information, etc.) and the data (including but not limited to positioning data for analysis, stored review record data, etc.) related to the present application are information and data authorized by the user or fully authorized by each party, and the collection, use and processing of the related data need to comply with related laws and regulations and standards, and provide corresponding operation entries for the user to select authorization or rejection.
Some terms involved in the embodiments of the present application are explained below:
edge computation (Edge computation): the method is characterized in that an open platform integrating network, computing, storage and application core capabilities is adopted at one side close to an object or data source, so that nearest service is provided. The application program is initiated at the edge side, can generate faster network service response, and meets the basic requirements of the industry in the aspects of real-time service, application intelligence, security, privacy protection and the like. Edge computation may be a distributed computing architecture that handles the computation of applications, data materials, and services by hub nodes moving to edge nodes on the network logic.
YOLOv5 rectangular box detection model: in the target detection, the detected object may be marked with a rectangular frame. The position and size of the rectangular box can accurately represent the position and size of the detected object in the image.
Book ReID model: the method can also be used for a model called a cross-mirror tracking technology, and can be used for identifying and searching books under cross-camera and cross-scene conditions.
Task alignment learning (Task Alignment Learning, TAL): in the task of object detection, label assignment plays a key role, the purpose of which is to divide the sample into positive and negative samples, and then calculate loss with GT (object box), determining how the model learns and converges. TAL may be used as a label distribution policy in the target detection task.
In intelligent library business, bookshelf digitalization can push favorite books to readers according to big data, service experience is improved, meanwhile, an administrator can know book browsing frequency more accurately, and bookshelf layout is optimized. The bookshelf digitization needs to input related information of books, such as front cover images, back cover images, picture names, publishers, authors, publishing dates, pricing and the like of the books into a database in advance to serve as a base for book retrieval and statistical analysis.
However, in the actual operation process, the shooting angle of view of the camera may not be completely perpendicular to the shooting plane for entering the front cover image and the back cover image, and the placement of the book to be entered is difficult to be standardized, which may result in the front cover image and the back cover image being non-standard polygons. Taking a common book as an example, if a rectangular frame detection model is used for extracting a front cover image and a back cover image from a shot image, the extracted front cover image and back cover image can contain a certain background area, which is not beneficial to subsequent standardized storage and feature retrieval of the front cover. Meanwhile, in the prior art, optical character recognition (Optical Character Recognition, OCR) technology is generally used to obtain book information such as book names, publishers, authors, date of publication, pricing, etc. of books from photographed images. However, when the difference between the text typesetting of the front cover and the back cover of the book is too large, the OCR technology cannot ensure the accuracy of text extraction, which may lead to errors in the entered book information. Therefore, how to quickly and accurately enter the related information of the book into the database is a technical problem which needs to be solved at present.
In view of the above technical problems, the embodiments of the present application provide a book data processing method, which corrects a front cover image and a back cover image of a book to be recorded into regular polygons by correcting the shape of the front cover and the shape of the back cover in a target image, so as to facilitate standardized storage and feature retrieval of the front cover of a subsequent book; meanwhile, the accuracy of the acquired book information can be ensured by detecting the bar code corresponding to the book to be recorded in the target image and acquiring the book information of the book to be recorded according to the bar code.
The technical scheme shown in the application is described in detail through specific embodiments. It should be noted that the following embodiments may exist alone or in combination with each other, and for the same or similar content, the description will not be repeated in different embodiments.
Referring to fig. 1, fig. 1 is a schematic architecture diagram of an application system according to an exemplary embodiment of the present application. As shown in fig. 1, the application system includes a terminal device 101, an edge calculation server 102, and a digital management server 103.
The terminal device 101 includes a monitoring device, an electronic device (e.g., a mobile phone, a tablet computer, a digital camera, etc.) having a photographing function, and the like, which is not limited in the embodiment of the present application.
Edge application server 102 may be understood as a small server that sinks the computing power of a traditional server from a central store to a point of proximity to terminal device 101. Unlike a centralized network, where the edge application server 102 is located at the edge of the network, this change in location solves many of the common problems of a centralized network, can improve latency, reduce load time, and remove load from the source server.
The digital management server 103 can be used for storing related information of books, realizing lease management of books, counting book reading records and the like. Alternatively, the digital management server 103 may be a cloud server.
Alternatively, the book data processing method may be independently executed by the terminal device 101 or the edge computing server 102, or may be executed by the terminal device 101 and the edge computing server 102 together, which is not limited in the embodiment of the present application. For example, in some embodiments, the user sends the target image to the edge computing server 102 through the terminal device 101; after receiving the target image, the edge computing server 102 performs related processing operations to obtain a front cover image, a back cover image and book information of the book to be recorded, and stores the front cover image, the back cover image and the book information in a database preset in the digital management server 103.
Referring to fig. 2, fig. 2 is a flow chart of a book data processing method according to an exemplary embodiment of the present application. In some embodiments, the book data processing method may include:
s201, at least one frame of target image is acquired, wherein the target image comprises a front cover and a back cover of a book to be recorded.
Note that, the cover of a book generally includes the following 5 parts: front cover (seal one), back cover (seal two), back cover (seal three), back cover (seal four) and spine. Wherein the cover is the outermost page of the book, and is usually printed with a book name, a author, a publisher name, a book illustration, etc.; the seal refers to the inside of the cover; the inside of the bottom cover refers to the inside of the bottom cover; the back cover is the last page of the book, and is connected with the front cover, and is generally blank except for the uniform book number, pricing and bar codes.
In some embodiments of the present application, a terminal device with an image acquisition function may acquire a book image of a book to be recorded in a library, and then upload the acquired book image to an edge computing server; the acquired book image comprises a front cover and a back cover of a book to be recorded.
For example, a camera may be used to capture the front and back covers of the book to be entered, respectively, and then the captured images may be uploaded to an edge computing server.
Optionally, after the front cover and the back cover of the book to be recorded are respectively shot by adopting a camera, the shot two frames of book images can be spliced into one frame of image and then uploaded to the edge computing server; or, the two shot book images can be uploaded to the edge computing server at the same time.
And the edge computing server takes the received book image as the target image after receiving the book image uploaded by the terminal equipment.
S202, respectively correcting the shape of a front cover and the shape of a back cover in a target image, and respectively determining the corrected front cover and back cover as a front cover image and a back cover image of a book to be recorded; wherein, the front cover image and the back cover image are regular polygons.
It will be appreciated that in an actual operating scenario, the camera view may not be perfectly perpendicular to the shooting plane, and the placement of the book to be entered may not be sufficiently standardized, thereby resulting in irregular polygons for the front and/or back covers of the book.
Fig. 3 is a schematic diagram of a process of acquiring a front cover image or a back cover image from a book image according to an embodiment of the present application.
Referring to fig. 3, a book image 301 of a book to be entered may be acquired by a camera, the book image 301 including a front cover or back cover 302 of the book to be entered. Wherein the front or back cover 302 is a non-standard rectangle due to the camera view angle. Further, a rectangular frame detection model is adopted, an external rectangular frame 303 of the book to be recorded is detected, and a front cover image or a back cover image 304 of the book to be recorded is obtained according to the external rectangular frame 303.
As can be seen in fig. 3, when the front or back cover 302 is a non-standard rectangle, if a rectangular frame detection model is used, the extracted front or back cover image 304 will contain a portion of the background area (i.e., the gray area in fig. 3) in the book image 301, which is detrimental to subsequent standardized storage of the front cover and retrieval of features.
In the application, the edge computing server can respectively correct the shape of the front cover and the shape of the back cover in the target image to be regular polygons, and then respectively determine the corrected front cover and back cover as a front cover image and a back cover image of the book to be recorded.
Fig. 4 is a schematic diagram illustrating another process of acquiring a front cover image or a back cover image from a book image according to an embodiment of the present application.
Referring to fig. 4, a book image 401 of a book to be entered may be acquired by a camera, the book image 401 including a front cover or back cover 402 of the book to be entered. Wherein the front or back cover 402 is non-standard rectangular due to the camera view angle. Further, the shape of the front cover or back cover 402 in the book image 401 is corrected to be a standard rectangle, and then the corrected front cover or back cover 403 is determined as a front cover image or back cover image 404 of the book to be entered.
As can be seen from fig. 4, when the front cover or back cover 402 is a non-standard rectangle, by correcting the shape of the front cover or back cover 402 and then acquiring the front cover image or back cover image 404 from the corrected image, the acquired front cover image or back cover image 404 will not include the background area in the book image 401, thereby facilitating the subsequent standardized storage and feature retrieval of the front cover.
S203, detecting a bar code corresponding to the book to be recorded in the target image, and acquiring book information of the book to be recorded according to the bar code.
In the embodiment of the application, after the edge computing server acquires the target image, the bar code corresponding to the book to be recorded can be detected in the target image, and the book information of the book to be recorded can be obtained by identifying the bar code.
Alternatively, the book information may include at least one of the following information: the title, publisher, author, date of publication, and pricing of the title, embodiments of the application are not limited herein.
It should be understood that there is no sequence of execution between the step S202 and the step S203, that is, in some embodiments, the step S202 may be executed first and then the step S203 may be executed. Alternatively, step S203 may be performed first, and then step S202 may be performed. Alternatively, step S202 and step S203 may be performed simultaneously, which is not limited herein.
S204, storing the front cover image, the back cover image and the book information into a preset database.
In this embodiment of the present application, after acquiring the front cover image, the back cover image, and the book information, the edge computing server may store the acquired front cover image, back cover image, and book information in a preset database.
It can be understood that the edge computing server is adopted to process the target image, and the front cover image, the back cover image and the book information obtained by processing are uploaded to the database, so that the load of the digital management server corresponding to the library can be reduced.
According to the book data processing method, after the target image is obtained, the shape of the front cover and the shape of the back cover in the target image are corrected, so that the front cover image and the back cover image of the book to be input can be corrected into regular polygons, and the standardized storage and the characteristic retrieval of the front cover of the subsequent book are facilitated; meanwhile, by detecting the bar code corresponding to the book to be recorded in the target image and acquiring the book information of the book to be recorded according to the bar code, the accuracy of the acquired book information can be ensured, and the technical problem that the related information of the book is difficult to be recorded in a database rapidly and accurately in the prior art is solved.
Based on the descriptions in the foregoing embodiments, referring to fig. 5, fig. 5 is another schematic flow chart of a book data processing method according to an exemplary embodiment of the present application. In some embodiments, the book data processing method may include:
s501, collecting book images.
In some embodiments of the present application, a terminal device with an image acquisition function may acquire a book image of a book to be recorded in a library, and then upload the acquired book image to an edge computing server.
The edge computing server receives at least one frame of book image uploaded by the terminal equipment and takes the book image as a target image; the book image is an image of a book to be recorded, which is collected by the terminal equipment and placed in any posture.
Wherein, the book image comprises a front cover and a back cover of the book to be recorded.
S502, detecting and correcting the shapes of the front cover and the back cover.
In this embodiment of the present application, the vertex coordinates of the front cover and the vertex coordinates of the back cover in the book image may be detected respectively based on a preset arbitrary polygon detection model, and the target vertex coordinates of the front cover and the target vertex coordinates of the back cover are determined according to the vertex coordinates of the front cover and the vertex coordinates of the back cover.
In some embodiments, the three parts of the detection head, the loss function and the label distribution mode of the detection head can be improved on the basis of a YOLOv5 rectangular frame detection model, an arbitrary polygon detection model based on vertex regression is trained and generated, the vertex coordinates of the front cover and the vertex coordinates of the back cover in the book image are detected respectively according to the arbitrary polygon detection model based on vertex regression, and the target vertex coordinates of the front cover and the target vertex coordinates of the back cover are determined according to the vertex coordinates of the front cover and the vertex coordinates of the back cover.
For example, the detection head portion of any polygonal detection model based on vertex regression may be modified from a rectangular frame detection head to a vertex regression detection head. For example, the output of the original rectangular frame detection head is (x, y, w, h), wherein x, y are coordinates of a central point of the rectangular detection frame, w is a width of the rectangular detection frame, and h is a height of the rectangular detection frame; the output of the vertex return detection head is (x) 1 ,y 1 ,…,x n ,y n ) Wherein x is 1 ,y 1 ,…,x n ,y n The coordinates of each vertex of the front cover or the back cover in the book image.
In some embodiments of the present application, the coordinates of the center point of the front cover and the coordinates of the center point of the back cover may be determined according to the coordinates of the top point of the front cover and the coordinates of the top point of the back cover; determining a true vector and a predicted vector of a central point to a top point of the front cover according to the top point coordinates and the central point coordinates of the front cover, and determining a true vector and a predicted vector of a central point to a top point of the back cover according to the top point coordinates and the central point coordinates of the back cover; determining target vertex coordinates of the cover based on a true vector and a predicted vector from a center point to a vertex of the cover and a preset loss function of the cover; and determining the target vertex coordinates of the back cover based on the true vector and the predicted vector from the center point to the vertex of the back cover and a loss function preset by the back cover.
In some embodiments, the loss function portion described above may be modified to optimize the Cross-IOU function of vertex distance regression. Assume that the true vector from the center point to the vertex isThe predictive vector is +.>The Cross-IOU between the true vector and the predicted vector is defined as:
for an n-sided polygon, the total Cross-IOU loss function of its n regression vectors is defined as:
in some embodiments, for the label distribution portion, a TAL label distribution method based on task alignment learning may be used, so that the arbitrary polygon detection model based on vertex regression converges faster and better in the training process, and finally higher accuracy is obtained. For the predicted positive sample, the classification score is higher, and a high IOU value with the real object is guaranteed, so that the TAL label distribution method can take a high-order combination of the classification score and the IOU score as a measurement value of the training sample. Assuming s is the classification score, u is the IOU value, and the scaling factors for both, respectively, the metric values are defined as:
t=s α ×u β
in the training process, the TAL label distribution method is utilized to calculate the metric values corresponding to all samples, the first k samples with the largest metric value are selected as positive samples, the rest samples are selected as negative samples, and the consistency of the sample measurement mode in the training and prediction processes is ensured.
In some embodiments, after determining the target vertex coordinates of the front cover and the target vertex coordinates of the back cover, determining an affine transformation matrix corresponding to the front cover according to the vertex coordinates of the front cover and the target vertex coordinates; and determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates.
Optionally, affine transformation can be obtained through translation, rotation, scaling, miscut and translation calculation, so that a transformation matrix corresponding to the front cover can be determined according to the vertex coordinates of the front cover and the target vertex coordinates, and a transformation matrix corresponding to the back cover can be determined according to the vertex coordinates of the back cover and the target vertex coordinates; and then, determining an affine transformation matrix corresponding to the front cover according to the transformation matrix corresponding to the front cover, and determining an affine transformation matrix corresponding to the back cover according to the transformation matrix corresponding to the back cover.
Optionally, the transformation matrix includes at least one of the following matrices: reflection matrix, rotation matrix, scaling matrix, chamfer matrix, and translation matrix.
For example, affine transformation matrix = reflection matrix x rotation matrix x scaling matrix x bias matrix x translation matrix; the specific formula is as follows:
In some embodiments, the shape of the cover in the target image may be corrected by using an affine transformation matrix corresponding to the cover, so as to obtain the cover image; and correcting the shape of the back cover in the target image by using an affine transformation matrix corresponding to the back cover to obtain the back cover image.
S503, detecting and identifying the bar code.
In some embodiments of the present application, after the book image is obtained, detecting the position information of the bar code corresponding to the book to be recorded in the target image, and intercepting the bar code in the target image according to the position information; and then, reading the number corresponding to the bar code, and searching book information of the book to be recorded in a preset book index library according to the number.
S504, information is put in storage.
In some embodiments of the present application, after the front cover image, the back cover image and the book information are obtained, the obtained front cover image, back cover image and book information may be saved to a preset database, so as to complete the digital bookshelf database creation.
In some embodiments of the present application, after the completion of the digitalized book shelf library establishment, the method may be applied to the book retrieval of readers.
For example, when receiving a search instruction, the edge computing server acquires a search keyword in the search instruction, searches the database for a matching book matching the search keyword, and outputs a front cover image, a back cover image, and book information of the matching book stored in the database.
In other embodiments of the present application, after the digitalized library construction of the bookshelf is completed, the method may be also applied to feature retrieval of books in the reader reference stage, so as to implement the function of counting the reference times of all books.
After the front cover image and the back cover image of the book to be recorded are obtained, depth characteristic information of the front cover image and depth characteristic information of the back cover image can be obtained based on a book ReID model, and the depth characteristic information of the front cover image and the depth characteristic information of the back cover image are stored in a database.
For example, in some embodiments, the front cover image and the back cover image of the book to be recorded may be input into the book ReID model, the 256-dimensional depth feature information corresponding to the front cover image and the back cover image may be obtained, and then the front cover image, the back cover image, the depth feature information and the book information may be stored into the database, so as to complete the data recording of the single book, and further realize the digitalized information library establishment flow of the whole bookshelf book.
According to the book data processing method, based on any polygon detection model, when books are placed in different postures, the edges of the books can be accurately detected and corrected to be regular polygons, so that the defect that a traditional rectangular frame detection algorithm cannot accurately position is overcome; meanwhile, by adopting a method based on bar code detection and identification, detailed information such as book names, publishers and the like can be accurately obtained according to bar code numbers in a service networking query mode, the influence of character typesetting and style difference on an OCR-based book character identification scheme can be overcome, and book information accuracy is ensured, so that related information of books can be rapidly and accurately input into a database, and the digital construction of bookshelf books is realized.
Based on the descriptions in the foregoing embodiments, referring to fig. 6, fig. 6 is a schematic flow chart of a book data processing method according to an exemplary embodiment of the present application. In some embodiments, the book data processing method may include:
s601, acquiring a monitoring image.
In some embodiments, the edge computing server may acquire the monitoring image in real time, or may acquire the monitoring image within a certain period of time.
S602, acquiring a target book image.
In some embodiments, whether a target book in a moving state exists in the monitoring image is detected, and if yes, a target book image corresponding to the target book is obtained in the monitoring image.
It will be appreciated that when a reader refers to a book, there will be an action of taking the book out of the bookshelf and/or putting the book back into the bookshelf, and the book will be in a moving state during the process of taking the book out of the bookshelf and/or putting the book back into the bookshelf, so that the book in the moving state in the monitoring image can be regarded as the book that the reader is referring to.
Alternatively, the target book image may be a front cover image and/or a back cover image corresponding to the target book.
S603, obtaining depth characteristic information.
In some embodiments, the depth characteristic information of the target book image may be directly acquired based on the book ReID model.
In other embodiments, the shape of the target book image may be corrected first, and then the depth feature information of the corrected target book image may be obtained based on the book ReID model, so that the obtained depth feature information may have a higher similarity with the depth feature information stored in the database, thereby improving the retrieval accuracy.
S604, retrieving depth characteristic information.
And searching target depth characteristic information matched with the depth characteristic information of the target book image in the stored depth characteristic information in the database, and acquiring book information of the target book in the database according to the target depth characteristic information.
S605, recording the search result.
In some embodiments, the book information of the target book may be added to the book reference record.
In some embodiments, the number of times of referring to each book in the preset time period may be counted according to the book reference record, and the attention degree information of each book may be generated according to the number of times of referring.
In some embodiments, the book reading monitoring record of the reader and the video picture of the picking and placing book can be also referred to according to the book reading record. In addition, information such as monthly, annual book reading and ranking can be generated.
After the bookshelf is built digitally, the book data processing method can be applied to characteristic retrieval of books in the reading stage of readers, and statistics of the reading times of all books is achieved, so that favorite books can be pushed to readers according to big data, service experience is improved, meanwhile, an administrator can know book browsing frequency more accurately, and bookshelf layout is optimized.
Based on the descriptions in the above embodiments, some embodiments of the present application further provide a book data processing apparatus. Referring to fig. 7, fig. 7 is a schematic structural diagram of a book data processing apparatus according to an exemplary embodiment of the present application, and the book data processing apparatus 70 includes:
the obtaining module 701 is configured to obtain at least one frame of target image, where the target image includes a front cover and a back cover of a book to be recorded.
The correction module 702 is configured to correct the shape of the front cover and the shape of the back cover in the target image respectively, and determine the corrected front cover and back cover as a front cover image and a back cover image of the book to be recorded respectively; wherein the front cover image and the back cover image are regular polygons.
The bar code identification module 703 is configured to detect a bar code corresponding to the book to be recorded in the target image, and obtain book information of the book to be recorded according to the bar code.
And the storage module 704 is used for storing the front cover image, the back cover image and the book information into a preset database.
In one possible implementation, the correction module 702 is configured to:
detecting the vertex coordinates of the front cover and the vertex coordinates of the back cover in the target image respectively based on a preset arbitrary polygon detection model;
determining the target vertex coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the back cover;
determining an affine transformation matrix corresponding to the cover according to the vertex coordinates of the cover and the target vertex coordinates; determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates;
correcting the shape of the front cover in the target image by using an affine transformation matrix corresponding to the front cover, and correcting the shape of the back cover in the target image by using an affine transformation matrix corresponding to the back cover.
In one possible implementation, the correction module 702 is configured to:
respectively determining the center point coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the vertex coordinates of the back cover;
determining a true vector and a predicted vector from the center point of the front cover to the top point according to the top point coordinates and the center point coordinates of the front cover, and determining a true vector and a predicted vector from the center point of the back cover to the top point according to the top point coordinates and the center point coordinates of the back cover;
determining target vertex coordinates of the cover based on a true vector and a predicted vector from a center point to a vertex of the cover and a loss function preset by the cover; and determining the target vertex coordinates of the back cover based on the true vector and the predicted vector from the central point to the vertex of the back cover and the preset loss function of the back cover.
In one possible implementation, the correction module 702 is configured to:
determining a transformation matrix corresponding to the front cover according to the vertex coordinates of the front cover and the target vertex coordinates, and determining a transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates; wherein the transformation matrix comprises at least one of the following matrices: a reflection matrix, a rotation matrix, a scaling matrix, a beveling matrix, and a translation matrix;
And determining an affine transformation matrix corresponding to the front cover according to the transformation matrix corresponding to the front cover, and determining an affine transformation matrix corresponding to the back cover according to the transformation matrix corresponding to the back cover.
In one possible implementation, the obtaining module 701 is configured to:
receiving at least one frame of book image uploaded by terminal equipment, and taking the book image as the target image; the book image is an image of the book to be recorded, which is collected by the terminal equipment and placed in any posture.
In one possible implementation, the bar code identification module 703 is configured to:
detecting the position information of the bar code corresponding to the book to be recorded in the target image;
intercepting the bar code in the target image according to the position information;
reading the number corresponding to the bar code, and searching the book information of the book to be recorded in a preset book index library according to the number; the book information includes at least one of the following information: title, publisher, author, publication date, pricing.
In a possible implementation manner, the apparatus further includes a retrieval module, configured to:
When a search instruction is received, acquiring a search keyword in the search instruction;
searching a matched book matched with the search keyword in the database;
and outputting the front cover image, the back cover image and the book information of the matched books stored in the database.
In a possible implementation manner, the apparatus further includes a feature processing module, configured to:
based on a book ReID model, depth characteristic information of the front cover image and depth characteristic information of the back cover image are respectively obtained;
and storing the depth characteristic information of the front cover image and the depth characteristic information of the back cover image into the database.
In a possible implementation manner, the apparatus further includes a reference detection module, configured to:
acquiring a monitoring image and detecting whether a target book in a moving state exists in the monitoring image;
when a target book in a moving state exists in the monitoring image, acquiring a target book image corresponding to the target book in the monitoring image;
based on a book ReID model, acquiring depth characteristic information of the target book image;
searching target depth characteristic information matched with the depth characteristic information of the target book image in the stored depth characteristic information in the database;
And according to the target depth characteristic information, acquiring book information of the target book in the database, and adding the book information of the target book in a book reference record.
In one possible implementation manner, the above-mentioned reference detection module is used for:
correcting the shape of the target book image;
and acquiring depth characteristic information of the corrected target book image based on the book ReID model.
In a possible implementation manner, the apparatus further includes a statistics module, configured to:
counting the consulting times of each book in a preset time according to the book consulting record;
and generating attention degree information of each book according to the consulting times.
It should be noted that, the specific execution contents of the acquiring module 701, the correcting module 702, the barcode identifying module 703 and the storing module 704 may refer to the relevant contents in the above method embodiment, and are not described herein.
Based on the foregoing description of the embodiments, an electronic device is further provided in some embodiments of the present application, and referring to fig. 8, fig. 8 is a schematic structural diagram of an electronic device provided in an exemplary embodiment of the present application, where the electronic device 80 may include a processor 801 and a memory 802. The processor 801 and the memory 802 are illustratively interconnected by a bus 803.
Memory 802 stores computer-executable instructions;
the processor 801 executes computer-executable instructions stored in the memory 802, so that the processor 801 performs the book data processing method as shown in the above-described method embodiment.
Accordingly, the present application further provides a computer readable storage medium, in which computer executable instructions are stored, for implementing the book data processing method described in the above method embodiments when the computer executable instructions are executed by a computer.
Accordingly, the embodiments of the present application may also provide a computer program product, including a computer program, which when executed by a computer may implement the book data processing method shown in the foregoing method embodiments.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors, input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission medium which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and changes may be made to the present application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. which are within the spirit and principles of the present application are intended to be included within the scope of the claims of the present application.
Claims (14)
1. A book data processing method, comprising:
acquiring at least one frame of target image, wherein the target image comprises a front cover and a back cover of a book to be recorded;
correcting the shape of the front cover and the shape of the back cover in the target image respectively, and determining the corrected front cover and back cover as a front cover image and a back cover image of the book to be recorded respectively; wherein the front cover image and the back cover image are regular polygons;
Detecting a bar code corresponding to the book to be recorded in the target image, and acquiring book information of the book to be recorded according to the bar code;
and storing the front cover image, the back cover image and the book information into a preset database.
2. The method of claim 1, wherein the correcting the shape of the front cover and the shape of the back cover in the target image, respectively, comprises:
detecting the vertex coordinates of the front cover and the vertex coordinates of the back cover in the target image respectively based on a preset arbitrary polygon detection model;
determining the target vertex coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the back cover;
determining an affine transformation matrix corresponding to the cover according to the vertex coordinates of the cover and the target vertex coordinates; determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates;
correcting the shape of the front cover in the target image by using an affine transformation matrix corresponding to the front cover, and correcting the shape of the back cover in the target image by using an affine transformation matrix corresponding to the back cover.
3. The method of claim 2, wherein determining the target vertex coordinates of the front cover and the back cover based on the vertex coordinates of the front cover and the back cover comprises:
respectively determining the center point coordinates of the front cover and the back cover according to the vertex coordinates of the front cover and the vertex coordinates of the back cover;
determining a true vector and a predicted vector from the center point of the front cover to the top point according to the top point coordinates and the center point coordinates of the front cover, and determining a true vector and a predicted vector from the center point of the back cover to the top point according to the top point coordinates and the center point coordinates of the back cover;
determining target vertex coordinates of the cover based on a true vector and a predicted vector from a center point to a vertex of the cover and a loss function preset by the cover; and determining the target vertex coordinates of the back cover based on the true vector and the predicted vector from the central point to the vertex of the back cover and the preset loss function of the back cover.
4. The method according to claim 2, wherein the affine transformation matrix corresponding to the cover is determined according to the vertex coordinates of the cover and the target vertex coordinates; and determining an affine transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates, wherein the affine transformation matrix comprises the following components:
Determining a transformation matrix corresponding to the front cover according to the vertex coordinates of the front cover and the target vertex coordinates, and determining a transformation matrix corresponding to the back cover according to the vertex coordinates of the back cover and the target vertex coordinates; wherein the transformation matrix comprises at least one of the following matrices: a reflection matrix, a rotation matrix, a scaling matrix, a beveling matrix, and a translation matrix;
and determining an affine transformation matrix corresponding to the front cover according to the transformation matrix corresponding to the front cover, and determining an affine transformation matrix corresponding to the back cover according to the transformation matrix corresponding to the back cover.
5. The method of claim 1, wherein the acquiring at least one frame of the target image comprises:
receiving at least one frame of book image uploaded by terminal equipment, and taking the book image as the target image; the book image is an image of the book to be recorded, which is collected by the terminal equipment and placed in any posture.
6. The method according to claim 1, wherein detecting a bar code corresponding to the book to be recorded in the target image and acquiring book information of the book to be recorded according to the bar code includes:
Detecting the position information of the bar code corresponding to the book to be recorded in the target image;
intercepting the bar code in the target image according to the position information;
reading the number corresponding to the bar code, and searching the book information of the book to be recorded in a preset book index library according to the number; the book information includes at least one of the following information: title, publisher, author, publication date, pricing.
7. The method as recited in claim 1, further comprising:
when a search instruction is received, acquiring a search keyword in the search instruction;
searching a matched book matched with the search keyword in the database;
and outputting the front cover image, the back cover image and the book information of the matched books stored in the database.
8. The method as recited in claim 1, further comprising:
based on a book ReID model, depth characteristic information of the front cover image and depth characteristic information of the back cover image are respectively obtained;
and storing the depth characteristic information of the front cover image and the depth characteristic information of the back cover image into the database.
9. The method as recited in claim 8, further comprising:
acquiring a monitoring image and detecting whether a target book in a moving state exists in the monitoring image;
when a target book in a moving state exists in the monitoring image, acquiring a target book image corresponding to the target book in the monitoring image;
based on a book ReID model, acquiring depth characteristic information of the target book image;
searching target depth characteristic information matched with the depth characteristic information of the target book image in the stored depth characteristic information in the database;
and according to the target depth characteristic information, acquiring book information of the target book in the database, and adding the book information of the target book in a book reference record.
10. The method of claim 9, wherein the obtaining depth characteristic information of the target book image based on a book ReID model comprises:
correcting the shape of the target book image;
and acquiring depth characteristic information of the corrected target book image based on the book ReID model.
11. The method as recited in claim 9, further comprising:
Counting the consulting times of each book in a preset time according to the book consulting record;
and generating attention degree information of each book according to the consulting times.
12. An electronic device, comprising: a memory and a processor;
the memory stores computer-executable instructions;
the processor executes computer-executable instructions stored in the memory, causing the processor to perform the book data processing method of any one of claims 1 to 11.
13. A computer-readable storage medium, wherein computer-executable instructions are stored in the computer-readable storage medium, which when executed by a computer, implement the book data processing method according to any one of claims 1 to 11.
14. A computer program product comprising a computer program which, when executed by a computer, implements the book data processing method of any one of claims 1 to 11.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311074432.2A CN117253244A (en) | 2023-08-23 | 2023-08-23 | Book data processing method and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311074432.2A CN117253244A (en) | 2023-08-23 | 2023-08-23 | Book data processing method and equipment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117253244A true CN117253244A (en) | 2023-12-19 |
Family
ID=89132156
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311074432.2A Pending CN117253244A (en) | 2023-08-23 | 2023-08-23 | Book data processing method and equipment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117253244A (en) |
-
2023
- 2023-08-23 CN CN202311074432.2A patent/CN117253244A/en active Pending
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Yousif et al. | Animal Scanner: Software for classifying humans, animals, and empty frames in camera trap images | |
EP2565804B1 (en) | Text-based searching of image data | |
US20230376527A1 (en) | Generating congruous metadata for multimedia | |
RU2668717C1 (en) | Generation of marking of document images for training sample | |
CN109241374B (en) | Book information base updating method and library book positioning method | |
US9076069B2 (en) | Registering metadata apparatus | |
CN101133429B (en) | Document/image searching method, and document/image recording and searching device | |
CN100430938C (en) | Image retrieving apparatus and image retrieving method | |
US20120093421A1 (en) | Detection of duplicate document content using two-dimensional visual fingerprinting | |
US9171203B2 (en) | Scanbox | |
Clinchant et al. | Comparing machine learning approaches for table recognition in historical register books | |
US9665773B2 (en) | Searching for events by attendants | |
CN110209863B (en) | Method and equipment for searching similar pictures | |
CN111428656A (en) | Mobile terminal identity card identification method based on deep learning and mobile device | |
Tsai et al. | Using cell phone pictures of sheet music to retrieve MIDI passages | |
CN111104542A (en) | Part identification management method and device | |
Iwanowski et al. | Comparing images for document plagiarism detection | |
CN111860122B (en) | Method and system for identifying reading comprehensive behaviors in real scene | |
CN112925939A (en) | Picture searching method, description information generating method, device and storage medium | |
Chumwatana et al. | Using OCR framework and information extraction for Thai documents digitization | |
CN117253244A (en) | Book data processing method and equipment | |
Ghazal et al. | Mobile-based archival and retrieval of missing objects using image matching | |
US20220164571A1 (en) | Model training method and apparatus, electronic device and readable storage medium | |
CN112364790B (en) | Airport work order information identification method and system based on convolutional neural network | |
CN114491134A (en) | Trademark registration success rate analysis method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |