CN110599665A - Paper pattern recognition method and device, computer equipment and storage medium - Google Patents
Paper pattern recognition method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN110599665A CN110599665A CN201810609999.8A CN201810609999A CN110599665A CN 110599665 A CN110599665 A CN 110599665A CN 201810609999 A CN201810609999 A CN 201810609999A CN 110599665 A CN110599665 A CN 110599665A
- Authority
- CN
- China
- Prior art keywords
- characteristic
- image
- feature
- points
- paper pattern
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/80—Recognising image objects characterised by unique random patterns
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/181—Testing mechanical properties or condition, e.g. wear or tear
- G07D7/187—Detecting defacement or contamination, e.g. dirt
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/2016—Testing patterns thereon using feature extraction, e.g. segmentation, edge detection or Hough-transformation
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07D—HANDLING OF COINS OR VALUABLE PAPERS, e.g. TESTING, SORTING BY DENOMINATIONS, COUNTING, DISPENSING, CHANGING OR DEPOSITING
- G07D7/00—Testing specially adapted to determine the identity or genuineness of valuable papers or for segregating those which are unacceptable, e.g. banknotes that are alien to a currency
- G07D7/20—Testing patterns thereon
- G07D7/202—Testing patterns thereon using pattern matching
- G07D7/2033—Matching unique patterns, i.e. patterns that are unique to each individual paper
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The application relates to a paper pattern recognition method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring a decontamination characteristic area image of the paper pattern image to be verified; extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points; performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image; and carrying out characteristic matching on the characteristic points of the pre-stored paper pattern image and the characteristic points of the decontamination characteristic area image to obtain a paper pattern recognition result. The paper pattern recognition method only needs to obtain the decontamination characteristic image of the fingerprint image to be verified, and does not need to shoot a large number of images, so that a large amount of time can be saved.
Description
Technical Field
The present application relates to the field of computer technologies, and in particular, to a method and an apparatus for identifying a bill, a computer device, and a storage medium.
Background
Paper veins, which refer to the natural fiber texture distribution characteristics within the paper, may be referred to as "paper veins". The paper pattern in the traditional sense means that paper pulp is uniformly distributed on a moving felt when a paper machine makes paper, so that fibers in the paper pulp form randomly distributed patterns, and the concept of the paper pattern only stays on the areas of brightness, glossiness, paper pattern depth, paper pattern width, paper pattern direction and the like. However, in fact, paper veins are a unique feature that nature imparts to each sheet of paper.
The bills are generally made of paper, and criminals often forge or alter the bills by altering, describing, copying, printing and other methods. Therefore, the bill needs to be identified and checked, and the bill can be identified by a paper-pattern identification method at present. I.e. comparing paper patterns. The method is characterized in that paper patterns of paper are extracted and identified, and when the bill is released (chartered and accepted), the paper patterns extracted when the bill is issued are compared to determine whether the bill is the original bill or not.
However, in the existing paper pattern identification method, a large number of images need to be shot in the paper pattern detection process, so that the paper pattern detection process is time-consuming.
Disclosure of Invention
In view of the above, it is necessary to provide a paper print identification method, apparatus, computer device and storage medium for solving the above technical problems.
A method of paper print identification, the method comprising:
acquiring a decontamination characteristic area image of the paper pattern image to be verified;
extracting characteristic points in the image of the decontamination characteristic area, and generating characteristic vectors according to the characteristic points;
generating a retrieval feature vector according to the feature vector;
performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image;
and carrying out feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
In one embodiment, the obtaining the image of the decontamination feature area of the paper print image to be verified further comprises:
obtaining a decontamination characteristic area image of a prestored paper pattern image;
extracting characteristic points in the decontamination characteristic region image, and acquiring retrieval characteristic vectors according to the characteristic points;
and storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database.
In one embodiment, the obtaining the image of the decontamination feature area of the paper print image to be verified comprises:
acquiring a characteristic area image of a paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, and removing the region where the pollution is located in the characteristic area image to obtain a decontamination characteristic area image.
In one embodiment, the extracting feature points in the image of the contamination removal feature area, and the obtaining of the retrieval feature vector according to the feature points specifically includes:
extracting characteristic points in the image of the decontamination characteristic area, and generating characteristic vectors according to the characteristic points;
and generating a retrieval feature vector according to the feature vector.
In one embodiment, the extracting feature points in the image of the contamination removal feature area, and the generating feature vectors according to the feature points includes:
constructing a scale space according to the decontamination feature area image;
acquiring an extreme point of the scale space;
removing the extreme points with asymmetric curvatures in the extreme points to obtain the characteristic points of the decontamination characteristic region image;
determining direction parameters according to the characteristic points of the decontamination characteristic region image;
and generating a characteristic vector according to the direction parameter.
In one embodiment, the generating a retrieval feature vector according to the feature vector includes:
and inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector.
In one embodiment, the performing feature search in a preset paper print database according to the search feature vector further includes, after obtaining a pre-stored paper print image:
acquiring pixel coordinates of feature points of the paper pattern image of the issued bill and the feature points of the decontamination feature area image which are successfully matched;
and calculating the shortest path of pixel coordinates between the successfully matched feature points.
A paper print recognition device, the device comprising:
the decontamination image acquisition module is used for acquiring a decontamination characteristic area image of the paper pattern image to be verified;
the retrieval feature vector extraction module is used for extracting feature points in the decontamination feature area image and generating retrieval feature vectors according to the feature points;
the characteristic retrieval module is used for carrying out characteristic retrieval in a preset paper pattern database according to the retrieval characteristic vector to obtain a prestored paper pattern image;
and the paper pattern recognition module is used for performing feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring a decontamination characteristic area image of the paper pattern image to be verified;
extracting characteristic points in the decontamination characteristic region image, and acquiring retrieval characteristic vectors according to the characteristic points;
performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image;
and carrying out feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring a decontamination characteristic area image of the paper pattern image to be verified;
extracting characteristic points in the decontamination characteristic region image, and acquiring retrieval characteristic vectors according to the characteristic points;
performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image;
and carrying out feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
According to the bill identification method, the bill identification device, the computer equipment and the storage medium, the decontamination characteristic region image of the paper pattern image to be verified is obtained, then the characteristic points of the decontamination characteristic region image are extracted, the characteristic vectors are obtained according to the characteristic points, the retrieval characteristic vectors are further obtained, so that the pre-stored paper pattern image matched with the paper pattern image to be verified is searched in the preset paper pattern database, the obtained characteristic points of the pre-stored paper pattern image are compared with the characteristic points of the paper pattern image to be verified, and the paper pattern identification result is obtained. The paper pattern recognition method only needs to obtain the decontamination characteristic image of the fingerprint image to be verified, and does not need to shoot a large number of images, so that a large amount of time can be saved.
Drawings
FIG. 1 is a schematic flow chart illustrating a method for identifying paper prints according to an embodiment;
FIG. 2 is a schematic flow chart illustrating a paper print recognition method according to another embodiment;
FIG. 3 is a schematic flow chart illustrating a paper print recognition method according to another embodiment;
FIG. 4 is a network architecture diagram of the YOLO-V2 model in one embodiment;
FIG. 5 is a flowchart illustrating step S400 according to an embodiment;
FIG. 6 is a flowchart illustrating step S410 according to an embodiment;
FIG. 7 is a flowchart illustrating a paper print recognition method according to another embodiment;
FIG. 8 is a block diagram showing the structure of a paper pattern recognition apparatus according to an embodiment;
FIG. 9 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
The paper pattern recognition method can be applied to the terminal. The captured image can be processed on the terminal. The terminal may be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, and portable wearable devices.
In an embodiment, as shown in fig. 1, a paper pattern recognition method is provided, which is described by taking the method as an example for being applied to the terminal, and includes the following steps:
and S200, obtaining a decontamination characteristic area image of the paper pattern image to be verified.
The paper pattern image to be verified refers to an image of a paper pattern to be recognized. In one embodiment, the paper print image to be verified can be an issued ticket whose paper print information is stored in a preset paper print database at the time of issuance. The paper pattern image to be verified contains the characteristic area image. The stain removal feature area image refers to a feature area image from which stains have been removed. In one embodiment, the characteristic region image of the recognized paper print may be acquired by infrared shooting. The pollution specifically refers to characteristic interference information such as pollutants or printed characters on the image. In one embodiment, pollution in the paper print image to be verified can be identified through a Yolo-v2 deep learning model, and the pollution is removed.
Firstly, a decontamination characteristic area image of a paper pattern image to be verified is obtained, and the interference of pollution in the paper pattern image to be verified on paper pattern recognition is prevented.
And S400, extracting characteristic points in the decontamination characteristic area image, and acquiring retrieval characteristic vectors according to the characteristic points.
The characteristic points are special points used for representing the decontamination characteristic area image and further representing the paper pattern image to be verified, and for one decontamination characteristic area image, a plurality of characteristic points corresponding to the decontamination characteristic area image can be arranged. The retrieval feature vector is a special vector used for quick retrieval in a preset paper pattern database.
Firstly, extracting characteristic points in the image of the decontamination characteristic area, and then acquiring retrieval characteristic vectors according to the characteristic points. In one embodiment, feature points in the image of the decontamination area can be extracted by a Scale-invariant feature transform (SIFT) algorithm and processed to obtain feature vectors. And then inputting the feature vector into a preset DBN (Deep Belief networks) model to obtain a retrieval feature vector.
And S600, performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image.
The preset paper pattern database is a database in which retrieval characteristic vectors of prestored paper pattern images are stored. The pre-stored paper pattern image refers to a paper pattern image which can be matched with the paper pattern image to be verified one by one.
And performing characteristic retrieval in a preset paper pattern database storing pre-stored paper pattern images according to the obtained retrieval characteristic vector, and when the pre-stored retrieval characteristic vector which is the same as the input retrieval characteristic vector is retrieved in the preset database, obtaining the pre-stored paper pattern image corresponding to the pre-stored retrieval characteristic vector.
And S800, carrying out feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
And then carrying out feature matching on the obtained feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image, and judging that the image corresponding to the paper pattern is the pre-stored paper pattern image in the preset database when the matching is successful. In one embodiment, the feature points of the obtained pre-stored paper print image and the feature points of the stain removal feature area image can be matched by a knmatch (K-nearest neighbor matching) algorithm. In one embodiment, the other stain removal feature area image is from a ticket, and the ticket is determined to be true when the matching with the pre-stored paper print image in the pre-set database is successful.
According to the bill identification method, the decontamination characteristic region image of the paper pattern image to be verified is obtained, the characteristic points of the decontamination characteristic region image are extracted, the characteristic vectors are obtained according to the characteristic points, the retrieval characteristic vectors are further obtained, the pre-stored paper pattern image matched with the paper pattern image to be verified is searched in the preset paper pattern database, the obtained characteristic points of the pre-stored paper pattern image are compared with the characteristic points of the paper pattern image to be verified, and the paper pattern identification result is obtained. The structure of the paper pattern recognition instrument can be simplified, and the portable paper pattern device is formed. The paper pattern recognition method only needs to obtain the decontamination characteristic image of the fingerprint image to be verified, and does not need to shoot a large number of images, so that a large amount of time can be saved. In addition, when the method is used for identifying the bill paper lines, the bill paper line characteristic data must be associated with the bill number, and the paper line identification method can quickly retrieve the issuing data according to the bill paper line characteristic data under the condition that the bill number information is not identified, so that the problem of paper line matching failure caused by bill number identification errors is reduced; the direct note paper grain characteristic retrieval comparison can realize the note acceptance without human-computer interaction and basically realize the unattended requirement of a paper grain instrument.
As shown in fig. 2, in one embodiment, before obtaining the stain removal feature area image of the paper print image to be verified, step S200 further includes:
and S121, obtaining a stain removal characteristic area image of the pre-stored paper pattern image.
And S123, extracting characteristic points in the decontamination characteristic area image, and acquiring retrieval characteristic vectors according to the characteristic points.
And S125, storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database.
Before the paper pattern image is identified, the retrieval characteristic vector of the pre-stored paper pattern image needs to be stored in the preset paper pattern database, and the processing process of the pre-stored paper pattern image is similar to the steps S200 to S400. In one embodiment, the pre-stored paper print image is an issued ticket, and the issued ticket may be subjected to the above-mentioned processing when issued for subsequent verification of the ticket.
As shown in fig. 3, in one embodiment, the obtaining of the stain removal characteristic area image of the paper print image to be verified in step S200 includes:
s140, obtaining a characteristic area image of the paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, removing the region where the pollution is located in the characteristic area image, and obtaining a decontamination characteristic area image.
Before the paper pattern image to be verified is identified, the characteristic region image of the paper pattern image to be verified can be detected through a Yolo-v2 algorithm. Specifically, as shown in fig. 4, a specific flow of the Yolo-v2 network model is to obtain a feature region image of an image to be verified, then divide the feature region image of the image to be verified into SxS grids, and if the center of an object in the feature region image falls in a certain grid, the grid is responsible for predicting the object. And each grid needs to predict a plurality of target windows, each target window needs to additionally predict a confidence value besides the position of the target window, and the confidence value represents the confidence of the object contained in the predicted window and how much quasi-dual information is predicted by the window. Each window also predicts a category information. At the time of testing, the class information of each grid prediction is multiplied by the confidence information of the target window prediction, and then class-specific confidence score of each target window is obtained. After obtaining the class-specific confidence score of each window, setting a threshold value, filtering out the target window with low score, and performing NMS (network management system) processing on the reserved target window to obtain a final detection result. The pollution in the characteristic area image of the image to be verified and the face printing content are identified through the Yolo-v2, and the interference of the pollution on the paper print identification is prevented.
As shown in fig. 5, in one embodiment, in step S400, extracting feature points in the image of the decontamination feature area, and acquiring and retrieving feature vectors according to the feature points specifically include:
and S410, extracting characteristic points in the image of the decontamination characteristic area, and generating characteristic vectors according to the characteristic points.
And S430, generating a retrieval feature vector according to the feature vector.
Feature points of the feature region image may be extracted first, and feature vectors may be generated from the obtained feature points. And then obtaining a retrieval feature vector according to the feature vector. The retrieval is carried out only by retrieving the characteristic vectors instead of the characteristic points, so that a large amount of retrieval time can be shortened, and the effect of identifying the paper pattern image to be verified in real time is achieved.
As shown in fig. 6, in one embodiment, the step S410 of extracting feature points in the image of the decontamination feature area and generating feature vectors according to the feature points includes:
and S411, constructing a scale space according to the image of the decontamination characteristic area.
And S413, acquiring an extreme point of the scale space.
And S415, removing the extreme points with asymmetric curvatures from the extreme points to obtain the characteristic points of the decontamination characteristic region image.
And S417, determining direction parameters according to the characteristic points of the decontamination characteristic area image.
And S419, generating a feature vector according to the direction parameter.
The feature vector of the paper pattern image to be verified can be extracted through the SIFT algorithm. The SIFT algorithm is a computer vision algorithm used for detecting and describing local features in an image, searching extreme points in a spatial scale and extracting position, scale and rotation invariants of the extreme points. Firstly, a scale space is constructed according to an image to be verified and is used for simulating the multi-time characteristics of image data. And then, detecting extreme points in the scale space, removing the extreme points with asymmetric curvatures in the extreme points, and taking other extreme points as characteristic points of the decontamination characteristic region image. The edge threshold of the extreme point may be specifically used as a criterion for determining whether the measurement is very asymmetric. And then calculating a direction for each feature point, further calculating according to the direction, and assigning a direction parameter for each feature point by using the gradient direction distribution characteristics of the pixels in the neighborhood of the feature point so as to enable the feature point to have rotation invariance. Specifically, the formula can be used:
the modulus of the gradient at point (x, y) is calculated. Wherein L is the scale at which the feature point is located.
And by the formula:
the direction at point (x, y) is calculated.
Until the feature points of the image are detected, each feature point has three pieces of information: position, scale, direction, from which a SIFT feature region can be determined. In actual calculation, sampling is carried out in a neighborhood window with the feature point as the center, and the gradient direction of a neighborhood pixel is counted by using a histogram. The gradient histogram ranges from 0 to 360 degrees with one column every 45 degrees for a total of 8 columns, or one column every 10 degrees for a total of 36 columns. The peak of the histogram represents the main direction of the neighborhood gradient at the feature point, i.e., the direction of the feature point.
And then the coordinate axes are rotated to be the direction of the characteristic points so as to ensure the rotation invariance. A 16 x 16 window is taken centered around the feature point, the gradient of each pixel in the 16 x 16 window around the keypoint is calculated, and the weights away from the center are reduced using a gaussian reduction function. Thus, a descriptor with 4 x 8-128 dimensions is formed for each feature point, the descriptor is the acquired feature vector matched with the feature point, and each dimension of the descriptor can represent the gradient and the direction of one of 4 x 4 grids.
In one embodiment, generating the search feature vector from the feature vector comprises:
and inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector.
The DBN network model can be used for unsupervised learning and is similar to a self-coding machine; and also can be used for supervised learning and used as a classifier. From unsupervised learning, the goal is to preserve the features of the original features as much as possible while reducing the dimensionality of the features. From supervised learning, the aim is to make the classification error rate as small as possible. Here we use its function for unsupervised learning with the goal of reducing the dimensionality of the feature vector while preserving metadata features. And inputting the feature vector into a pre-trained DBN network model to obtain a retrieval feature vector for retrieval.
As shown in fig. 7, in one embodiment, S600, after performing feature search in the preset paper print database according to the searched feature vector, and obtaining the pre-stored paper print image, further includes:
s710, acquiring pixel coordinates of feature points successfully matched with the feature points of the paper print image and the feature points of the decontamination feature area image of the issued bill.
And S730, calculating the shortest path of the pixel coordinates between the successfully matched feature points.
The pixel coordinates are quantities that describe the location of the pixel point in the image. And in the characteristic point matching process, calculating the shortest path of pixel coordinates between the successfully matched characteristic points. Random feature points are formed by discrete highlight points (namely, polluted points), a large polluted point can form a plurality of cross feature matching points, and the condition of paper-line-free feature mismatching is prevented after the threshold value of the matching points is exceeded.
In one embodiment, S800, performing feature matching on the feature points of the pre-stored paper print image and the feature points of the stain removal feature area image, and after obtaining the paper print recognition result, further includes:
and when the feature matching is successful, outputting the pixel coordinate data of all the feature matching point pairs at the same time.
The coordinate data can be restored to real image data, so that the data is convenient to keep and can be visually verified.
In one embodiment, the paper print identification method of the present application includes the following steps:
and acquiring a stain removal characteristic area image of the pre-stored paper pattern image. And extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points. And storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database. And acquiring a characteristic area image of the paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, and removing the region where the pollution is located in the characteristic area image to obtain a decontamination characteristic area image. And acquiring a decontamination characteristic area image of the paper pattern image to be verified. And constructing a scale space according to the decontamination characteristic region image. And acquiring an extreme point of the scale space. And removing the extreme points with asymmetric curvatures in the extreme points to obtain the characteristic points of the image of the decontamination characteristic region. And determining the direction parameters according to the characteristic points of the image of the decontamination characteristic area. And generating a characteristic vector according to the direction parameter. And inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector. And performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image. And acquiring pixel coordinates of the feature points of the paper print image of the issued bill and the feature points successfully matched in the feature points of the decontamination feature area image. And calculating the shortest path of pixel coordinates between the successfully matched feature points. And S800, carrying out feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result. And when the feature matching is successful, outputting the pixel coordinate data of all the feature matching point pairs at the same time.
It should be understood that although the various steps in the flow charts of fig. 1-7 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-7 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternating with other steps or at least some of the sub-steps or stages of other steps.
As shown in fig. 8, the present application also provides a paper pattern recognition apparatus, including:
a stain removal image acquisition module 200, configured to obtain a stain removal characteristic area image of a paper print image to be verified;
a retrieval feature vector extraction module 400, configured to extract feature points in the image of the decontamination feature area, and generate a retrieval feature vector according to the feature points;
the characteristic retrieval module 600 is configured to perform characteristic retrieval in a preset paper print database according to the retrieved characteristic vector to obtain a pre-stored paper print image;
and the paper pattern recognition module 800 is configured to perform feature matching on the feature points of the pre-stored paper pattern image and the feature points of the stain removal feature area image to obtain a paper pattern recognition result.
In one embodiment, the system further comprises a paper pattern image pre-storing module, configured to:
obtaining a decontamination characteristic area image of a prestored paper pattern image;
extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points;
and storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database.
In one embodiment, the paper pattern decontamination module is further included for:
and acquiring a characteristic area image of the paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, and removing the region where the pollution is located in the characteristic area image to obtain a decontamination characteristic area image.
In one embodiment, the retrieval feature vector extraction module 400 specifically includes:
the characteristic vector extraction unit is used for extracting characteristic points in the decontamination characteristic region image and generating characteristic vectors according to the characteristic points;
and the retrieval feature vector extraction unit is used for generating a retrieval feature vector according to the feature vector.
In one embodiment, the feature vector extraction unit is configured to:
constructing a scale space according to the decontamination feature area image;
acquiring an extreme point of a scale space;
removing the extreme points with asymmetric curvatures from the extreme points to obtain the characteristic points of the decontamination characteristic region image;
determining direction parameters according to the characteristic points of the image of the decontamination characteristic area;
and generating a characteristic vector according to the direction parameter.
In one embodiment, the retrieval feature vector extraction unit is configured to:
and inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector.
In one embodiment, the system further includes a matching feature point shortest path calculation unit, configured to:
acquiring pixel coordinates of feature points successfully matched with the feature points of the paper pattern image of the issued bill and the feature points of the decontamination feature area image;
and calculating the shortest path of pixel coordinates between the successfully matched feature points.
For the specific definition of the fingerprint identification device, reference may be made to the above definition of the paper print identification method, which is not described herein again. The modules in the paper pattern recognition device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a terminal, and its internal structure diagram may be as shown in fig. 9. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement a paper print recognition method. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on the shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
Those skilled in the art will appreciate that the architecture shown in fig. 9 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring a decontamination characteristic area image of the paper pattern image to be verified;
extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points;
performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image;
and carrying out characteristic matching on the characteristic points of the pre-stored paper pattern image and the characteristic points of the decontamination characteristic area image to obtain a paper pattern recognition result.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
obtaining a decontamination characteristic area image of a prestored paper pattern image;
extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points;
and storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and acquiring a characteristic area image of the paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, and removing the region where the pollution is located in the characteristic area image to obtain a decontamination characteristic area image.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
extracting characteristic points in the image of the decontamination characteristic area, and generating characteristic vectors according to the characteristic points;
and generating a retrieval feature vector according to the feature vector.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
constructing a scale space according to the decontamination feature area image;
acquiring an extreme point of a scale space;
removing the extreme points with asymmetric curvatures from the extreme points to obtain the characteristic points of the decontamination characteristic region image;
determining direction parameters according to the characteristic points of the image of the decontamination characteristic area;
and generating a characteristic vector according to the direction parameter.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
and inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
acquiring pixel coordinates of feature points successfully matched with the feature points of the paper pattern image of the issued bill and the feature points of the decontamination feature area image;
and calculating the shortest path of pixel coordinates between the successfully matched feature points.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring a decontamination characteristic area image of the paper pattern image to be verified;
extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points;
performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image;
and carrying out characteristic matching on the characteristic points of the pre-stored paper pattern image and the characteristic points of the decontamination characteristic area image to obtain a paper pattern recognition result.
In one embodiment, the computer program when executed by the processor further performs the steps of:
obtaining a decontamination characteristic area image of a prestored paper pattern image;
extracting characteristic points in the image of the decontamination characteristic area, and acquiring retrieval characteristic vectors according to the characteristic points;
and storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database.
In one embodiment, the computer program when executed by the processor further performs the steps of: and acquiring a characteristic area image of the paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, and removing the region where the pollution is located in the characteristic area image to obtain a decontamination characteristic area image.
In one embodiment, the computer program when executed by the processor further performs the steps of:
extracting characteristic points in the image of the decontamination characteristic area, and generating characteristic vectors according to the characteristic points;
and generating a retrieval feature vector according to the feature vector.
In one embodiment, the computer program when executed by the processor further performs the steps of:
constructing a scale space according to the decontamination feature area image;
acquiring an extreme point of a scale space;
removing the extreme points with asymmetric curvatures from the extreme points to obtain the characteristic points of the decontamination characteristic region image;
determining direction parameters according to the characteristic points of the image of the decontamination characteristic area;
and generating a characteristic vector according to the direction parameter.
In one embodiment, the computer program when executed by the processor further performs the steps of:
and inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector.
In one embodiment, the computer program when executed by the processor further performs the steps of:
acquiring pixel coordinates of feature points successfully matched with the feature points of the paper pattern image of the issued bill and the feature points of the decontamination feature area image;
and calculating the shortest path of pixel coordinates between the successfully matched feature points.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. A method of paper print identification, the method comprising:
acquiring a decontamination characteristic area image of the paper pattern image to be verified;
extracting characteristic points in the decontamination characteristic region image, and acquiring retrieval characteristic vectors according to the characteristic points;
performing feature retrieval in a preset paper pattern database according to the retrieval feature vector to obtain a prestored paper pattern image;
and carrying out feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
2. The method of claim 1, wherein obtaining the image of the desmear feature area of the paper print image to be verified further comprises:
obtaining a decontamination characteristic area image of a prestored paper pattern image;
extracting characteristic points in the decontamination characteristic region image, and acquiring retrieval characteristic vectors according to the characteristic points;
and storing the retrieval characteristic vector of the pre-stored paper pattern image into a pre-stored paper pattern database.
3. The method of claim 1, wherein obtaining the image of the desmear feature area of the paper print image to be verified is preceded by:
acquiring a characteristic area image of a paper pattern image to be verified, detecting a region where pollution is located in the characteristic area image through a Yolo-v2 model, and removing the region where the pollution is located in the characteristic area image to obtain a decontamination characteristic area image.
4. The method according to claim 1, wherein the extracting of the feature points in the image of the contamination removal feature area and the obtaining of the retrieval feature vector from the feature points specifically comprise:
extracting characteristic points in the image of the decontamination characteristic area, and generating characteristic vectors according to the characteristic points;
and generating a retrieval feature vector according to the feature vector.
5. The method of claim 4, wherein the extracting feature points within the de-contamination feature region image, and the generating feature vectors from the feature points comprises:
constructing a scale space according to the decontamination feature area image;
acquiring an extreme point of the scale space;
removing the extreme points with asymmetric curvatures in the extreme points to obtain the characteristic points of the decontamination characteristic region image;
determining direction parameters according to the characteristic points of the decontamination characteristic region image;
and generating a characteristic vector according to the direction parameter.
6. The method of claim 4, wherein generating the search feature vector from the feature vector comprises:
and inputting the characteristic vector into a preset DBN network model to obtain a retrieval characteristic vector.
7. The method according to claim 1, wherein the performing a feature search in a preset paper print database according to the search feature vector further comprises, after obtaining a pre-stored paper print image:
acquiring pixel coordinates of feature points of the paper pattern image of the issued bill and the feature points of the decontamination feature area image which are successfully matched;
and calculating the shortest path of pixel coordinates between the successfully matched feature points.
8. A paper print recognition apparatus, the method comprising:
the decontamination image acquisition module is used for acquiring a decontamination characteristic area image of the paper pattern image to be verified;
the retrieval feature vector extraction module is used for extracting feature points in the decontamination feature area image and generating retrieval feature vectors according to the feature points;
the characteristic retrieval module is used for carrying out characteristic retrieval in a preset paper pattern database according to the retrieval characteristic vector to obtain a prestored paper pattern image;
and the paper pattern recognition module is used for performing feature matching on the feature points of the pre-stored paper pattern image and the feature points of the decontamination feature area image to obtain a paper pattern recognition result.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810609999.8A CN110599665B (en) | 2018-06-13 | 2018-06-13 | Paper pattern recognition method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810609999.8A CN110599665B (en) | 2018-06-13 | 2018-06-13 | Paper pattern recognition method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110599665A true CN110599665A (en) | 2019-12-20 |
| CN110599665B CN110599665B (en) | 2021-08-24 |
Family
ID=68849554
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810609999.8A Active CN110599665B (en) | 2018-06-13 | 2018-06-13 | Paper pattern recognition method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110599665B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111680549A (en) * | 2020-04-28 | 2020-09-18 | 肯维捷斯（武汉）科技有限公司 | Paper pattern recognition method |
| CN115257198A (en) * | 2022-07-26 | 2022-11-01 | 上海商米科技集团股份有限公司 | Thermal paper printing identification method and device and thermal printer |
Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1510629A (en) * | 2002-12-25 | 2004-07-07 | 鲍东山 | Compound pattern processing banknote identifying system |
| CN1916975A (en) * | 2006-09-13 | 2007-02-21 | 顾泽苍 | Processing method for guiding against false of paper currency |
| CN101425194A (en) * | 2007-11-02 | 2009-05-06 | 广州广电运通金融电子股份有限公司 | Method for verifying banknote origin, relevant device and system |
| CN102063758A (en) * | 2009-11-16 | 2011-05-18 | 中科院成都信息技术有限公司 | Banknote paper texture detection method and device |
| CN102214305A (en) * | 2011-04-08 | 2011-10-12 | 大连理工大学 | Method for taking evidence for source of printing paper sheet by using grain characteristic |
| CN103035032A (en) * | 2011-10-09 | 2013-04-10 | 天津市阿波罗信息技术有限公司 | Method for generating admission tickets of public places |
| CN103208004A (en) * | 2013-03-15 | 2013-07-17 | 北京英迈杰科技有限公司 | Automatic recognition and extraction method and device for bill information area |
| US20130287263A1 (en) * | 2010-12-24 | 2013-10-31 | Sinosun Technology Co., Ltd. | Anti-counterfeiting test method, device and system based on textures |
| CN103761799A (en) * | 2014-01-13 | 2014-04-30 | 王波 | Bill anti-counterfeit method and device based on texture image characteristics |
| CN104217221A (en) * | 2014-08-27 | 2014-12-17 | 重庆大学 | Method for detecting calligraphy and paintings based on textural features |
| CN105046191A (en) * | 2015-05-13 | 2015-11-11 | 信阳师范学院 | Texture image identifying method |
| CN106127930A (en) * | 2016-06-17 | 2016-11-16 | 深圳怡化电脑股份有限公司 | A kind of method and system detecting banknote security line integrity |
| CN106447908A (en) * | 2016-09-20 | 2017-02-22 | 深圳怡化电脑股份有限公司 | Paper money authentic identification method and device |
| CN106447361A (en) * | 2016-10-28 | 2017-02-22 | 王友炎 | Anti-counterfeiting authentication and recording tracing system and method for paper-medium-based artworks |
| US20170124795A1 (en) * | 2015-10-30 | 2017-05-04 | Ncr Corporation | Document geometric deformation watermarking and tracking |
| CN106780962A (en) * | 2016-11-10 | 2017-05-31 | 深圳怡化电脑股份有限公司 | A kind of method and device of paper money discrimination |
| CN107533631A (en) * | 2015-02-13 | 2018-01-02 | 帕佩尔Dna股份公司 | Use the verification method of surface paper texture |
| CN107680245A (en) * | 2017-09-27 | 2018-02-09 | 深圳怡化电脑股份有限公司 | A kind of banknote image processing method, banknote image processing unit and electronic equipment |
-
2018
- 2018-06-13 CN CN201810609999.8A patent/CN110599665B/en active Active
Patent Citations (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1510629A (en) * | 2002-12-25 | 2004-07-07 | 鲍东山 | Compound pattern processing banknote identifying system |
| CN1916975A (en) * | 2006-09-13 | 2007-02-21 | 顾泽苍 | Processing method for guiding against false of paper currency |
| CN101425194A (en) * | 2007-11-02 | 2009-05-06 | 广州广电运通金融电子股份有限公司 | Method for verifying banknote origin, relevant device and system |
| CN102063758A (en) * | 2009-11-16 | 2011-05-18 | 中科院成都信息技术有限公司 | Banknote paper texture detection method and device |
| US20130287263A1 (en) * | 2010-12-24 | 2013-10-31 | Sinosun Technology Co., Ltd. | Anti-counterfeiting test method, device and system based on textures |
| CN102214305A (en) * | 2011-04-08 | 2011-10-12 | 大连理工大学 | Method for taking evidence for source of printing paper sheet by using grain characteristic |
| CN103035032A (en) * | 2011-10-09 | 2013-04-10 | 天津市阿波罗信息技术有限公司 | Method for generating admission tickets of public places |
| CN103208004A (en) * | 2013-03-15 | 2013-07-17 | 北京英迈杰科技有限公司 | Automatic recognition and extraction method and device for bill information area |
| CN103761799A (en) * | 2014-01-13 | 2014-04-30 | 王波 | Bill anti-counterfeit method and device based on texture image characteristics |
| CN104217221A (en) * | 2014-08-27 | 2014-12-17 | 重庆大学 | Method for detecting calligraphy and paintings based on textural features |
| CN107533631A (en) * | 2015-02-13 | 2018-01-02 | 帕佩尔Dna股份公司 | Use the verification method of surface paper texture |
| CN105046191A (en) * | 2015-05-13 | 2015-11-11 | 信阳师范学院 | Texture image identifying method |
| US20170124795A1 (en) * | 2015-10-30 | 2017-05-04 | Ncr Corporation | Document geometric deformation watermarking and tracking |
| CN106127930A (en) * | 2016-06-17 | 2016-11-16 | 深圳怡化电脑股份有限公司 | A kind of method and system detecting banknote security line integrity |
| CN106447908A (en) * | 2016-09-20 | 2017-02-22 | 深圳怡化电脑股份有限公司 | Paper money authentic identification method and device |
| CN106447361A (en) * | 2016-10-28 | 2017-02-22 | 王友炎 | Anti-counterfeiting authentication and recording tracing system and method for paper-medium-based artworks |
| CN106780962A (en) * | 2016-11-10 | 2017-05-31 | 深圳怡化电脑股份有限公司 | A kind of method and device of paper money discrimination |
| CN107680245A (en) * | 2017-09-27 | 2018-02-09 | 深圳怡化电脑股份有限公司 | A kind of banknote image processing method, banknote image processing unit and electronic equipment |
Non-Patent Citations (2)
| Title |
|---|
| 方天红，陈庆虎，鄢煜尘，周前进: "《基于统计纹理特征的打印文档认证》", 《武汉大学学报(工学版)》 * |
| 项楠: "《票据诈骗案频发纸纹为票据"加把锁"》", 《防伪技术》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111680549A (en) * | 2020-04-28 | 2020-09-18 | 肯维捷斯（武汉）科技有限公司 | Paper pattern recognition method |
| CN111680549B (en) * | 2020-04-28 | 2023-12-05 | 肯维捷斯（武汉）科技有限公司 | Paper grain identification method |
| CN115257198A (en) * | 2022-07-26 | 2022-11-01 | 上海商米科技集团股份有限公司 | Thermal paper printing identification method and device and thermal printer |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110599665B (en) | 2021-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111191539B (en) | Certificate authenticity verification method and device, computer equipment and storage medium | |
| Tirunagari et al. | Detection of face spoofing using visual dynamics | |
| Chugh et al. | Fingerprint spoof detection using minutiae-based local patches | |
| US8744144B2 (en) | Feature point generation system, feature point generation method, and feature point generation program | |
| US10373014B2 (en) | Object detection method and image search system | |
| Peralta et al. | Minutiae filtering to improve both efficacy and efficiency of fingerprint matching algorithms | |
| CN111191568A (en) | Method, device, equipment and medium for identifying copied image | |
| CN112183296B (en) | Simulated bill image generation and bill image recognition method and device | |
| CN110599665B (en) | Paper pattern recognition method and device, computer equipment and storage medium | |
| CN110738222B (en) | Image matching method and device, computer equipment and storage medium | |
| CN113139536A (en) | Text verification code identification method and equipment based on cross-domain meta learning and storage medium | |
| US20230147685A1 (en) | Generalized anomaly detection | |
| CN111639667B (en) | Image recognition method, device, electronic equipment and computer readable storage medium | |
| Emam et al. | A robust detection algorithm for image Copy-Move forgery in smooth regions | |
| CN113903042A (en) | Trademark identification method and device, computer equipment and storage medium | |
| CN111340748A (en) | Battery defect identification method and device, computer equipment and storage medium | |
| CN117235820A (en) | Chip self-destruction circuit and destruction method based on same | |
| US20230069960A1 (en) | Generalized anomaly detection | |
| CN113723215B (en) | Training method of living body detection network, living body detection method and device | |
| CN115424001A (en) | Scene similarity estimation method and device, computer equipment and storage medium | |
| CN113762249A (en) | Image attack detection and image attack detection model training method and device | |
| Okokpujie et al. | A COMPARATIVE ANALYSIS PERFORMANCE OF DATA AUGMENTATION ON AGE-INVARIANT FACE RECOGNI-TION USING PRETRAINED RESIDUAL NEURAL NETWORK | |
| CN113705511A (en) | Gesture recognition method and device | |
| Li et al. | Fingerprint liveness detection based on binarized statistical image feature with sampling from Gaussian distribution | |
| CN113780131B (en) | Text image orientation recognition method, text content recognition method, device and equipment |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |