CN109558794A - Image-recognizing method, device, equipment and storage medium based on moire fringes - Google Patents
Image-recognizing method, device, equipment and storage medium based on moire fringes Download PDFInfo
- Publication number
- CN109558794A CN109558794A CN201811206940.0A CN201811206940A CN109558794A CN 109558794 A CN109558794 A CN 109558794A CN 201811206940 A CN201811206940 A CN 201811206940A CN 109558794 A CN109558794 A CN 109558794A
- Authority
- CN
- China
- Prior art keywords
- image
- moire fringes
- described image
- feature
- training
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 77
- 241000208340 Araliaceae Species 0.000 claims abstract description 18
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 claims abstract description 18
- 235000003140 Panax quinquefolius Nutrition 0.000 claims abstract description 18
- 235000008434 ginseng Nutrition 0.000 claims abstract description 18
- 238000012706 support-vector machine Methods 0.000 claims description 22
- 239000013598 vector Substances 0.000 claims description 14
- 208000011580 syndromic disease Diseases 0.000 claims description 13
- 238000004364 calculation method Methods 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 7
- 238000010606 normalization Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 5
- 238000003909 pattern recognition Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 230000004044 response Effects 0.000 claims description 4
- 238000000513 principal component analysis Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000005070 sampling Methods 0.000 claims description 3
- 241000894007 species Species 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 8
- 230000008569 process Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 4
- 230000008901 benefit Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 230000036760 body temperature Effects 0.000 description 2
- 238000013499 data model Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000001815 facial effect Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000003064 k means clustering Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- 230000007306 turnover Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
- G06V10/507—Summing image-intensity values; Histogram projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/467—Encoded features or binary features, e.g. local binary patterns [LBP]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
This application involves image identification technical field more particularly to a kind of image-recognizing method based on moire fringes, device, computer equipment and storage mediums.Image-recognizing method based on moire fringes includes: to obtain several images to be processed, is packaged into an image set, it includes the image with moire fringes and the image without moire fringes that described image, which is concentrated,;The moire fringes feature that any image that described image is concentrated is included is extracted, training sample is formed;Using the training sample as ginseng is entered, SVM training is carried out in SVM model, obtains image recognition sample, identify that sample carries out moire fringes identification to unknown images according to described image.The application identifies moire fringes using LBP algorithm there are problems that being cheated using reproduction picture during financial transaction, and passes through SVM model training, promotes speed and accuracy to reproduction photo array.
Description
Technical field
This application involves image identification technical field more particularly to a kind of image-recognizing method based on moire fringes, device,
Equipment and storage medium.
Background technique
During finance activities, generally require to carry out user recognition of face to determine identity.However carrying out people
In face identification process, certain people pretend to be true face using the facial image of reproduction to realize the purpose of fraud.Exist at present
Instruction action using by issuing some randomnesss at random main during reproduction photo array, such as blink, left and right are shaken
Head and that shakes the head etc up and down identify;Or human body temperature is identified using infrared ray.
But it results in the need for user there is the response time difference when being identified using stochastic instruction movement and repeatedly weighs
Double action work could complete identification, and infrared human body temperature is used to identify that the interference that then will receive ambient enviroment reduces the standard of identification
True rate.
Summary of the invention
In view of this, it is necessary to it is low for existing reproduction photo array degree, timely and effective user's identity cannot be carried out
The problem of identification, provides a kind of image-recognizing method based on moire fringes, device, equipment and storage medium.
A kind of image-recognizing method based on moire fringes, includes the following steps:
Several images to be processed are obtained, are packaged into an image set, it includes the image with moire fringes that described image, which is concentrated,
With the image without moire fringes;
The moire fringes feature that any image that described image is concentrated is included is extracted, training sample is formed;
Using the training sample as ginseng is entered, SVM training is carried out in support vector machines model, obtains image recognition
Model;
The unknown images uploaded are received, moire fringes identification is carried out to the unknown images according to described image identification model,
To determine whether described image is reproduction image.
It is described in one of the embodiments, to obtain several images to be processed, it is packaged into an image set, described image collection
In include the image with moire fringes and image without moire fringes, comprising:
When the distance of any image to image acquisition device screen is less than a certain distance threshold, described image is adopted
Collection;
Collected described image is subjected to size and color normalized, obtains normalized image;
All normalized images are ranked up according to the time is generated, are packaged into an image set.
The moire fringes feature extracted any image that described image is concentrated and included in one of the embodiments,
Form training sample, comprising:
By described image according to the sub-block for being horizontally and vertically divided into the sizes such as n × n, n is any positive integer;
Carry out multiple dimensioned local binary patterns LBP histogram calculation respectively to each sub-block, calculation method is as follows:
LBP of each pixel on certain scaleP,RValue are as follows:
Wherein, gcFor the gray value of pixel, gpFor with gcFor the center of circle, R is p pixel extracting on the circumference of radius
Gray value, S indicate impact factor, 2pIntermediate scheme species number, wherein p=0 ..., n;
According to the LBPP,RValue, establishes the LBP histogram of the sub-block;
Sub-block corresponding to ordinate maximum value in the LBP histogram is obtained, enters ginseng as training sample, wherein if
The LBP of sub-block corresponding to ordinate maximum value in the LBP histogramP,RBe zero or positive number then illustrate described image without
There are moire fringes, be negative, illustrates described image with moire fringes.
In one of the embodiments, it is described using the training sample as enter ginseng, in support vector machines model into
Row SVM training, obtains image recognition model, comprising:
Obtain the training sample;
Low-dimensional union feature matrix is obtained after the training sample is carried out principal component analysis PCA dimensionality reduction;
The low-dimensional union feature matrix is entered into ginseng and carries out classification based training to the SVM model, obtains several two classes identifications
Device, respectively moire fringes identifier and without moire fringes identifier;
Moire fringes identification is carried out to the training sample using the moire fringes identifier and the no moire fringes identifier,
Image recognition model is established according to moire fringes recognition result.
The moire fringes feature extracted any image that described image is concentrated and included in one of the embodiments,
Form training sample, comprising:
Described image is divided into nine trained subgraph blocks, and the trained subgraph block is divided into two groups, first group of total m block
As training subgraph block, second group of total 9-m block is as syndrome segment, wherein 1≤m≤8, and m is integer;
Extract the moire fringes feature in the trained subgraph block, wherein the instruction of moire fringes is had in the trained subgraph block
Practice subgraph block and be labeled as " 1 " labeled as " 2 ", without the training subgraph block of moire fringes, establishes training histogram,;
The moire fringes feature in the syndrome segment is extracted, the inspection subgraph of moire fringes is had in the syndrome segment
Block is labeled as " 1 " labeled as " 2 ", without the syndrome segment of moire fringes, establishes and examines histogram;
Summarize the trained histogram and histogram is examined to obtain image histogram, if there is one on described image histogram
Subgraph block is labeled as " 2 ", then described image has moire fringes feature, and otherwise described image is without moire fringes feature, and is formed
Training sample.
It is described in one of the embodiments, that collected described image is subjected to size and color normalized, it obtains
To normalized image, comprising:
Carry out that scale is w, non-lower sampling NSCT that direction is h is decomposed to the picture in the picture library, obtain one it is low
On frequency subgraph A0 and multiple and different scales and different directions high frequency subgraph A1,1;A1,2;...;Aw, 1;...;Aw, h },
High frequency subgraph Aw, it is w that h, which describes scale, and direction is the face texture information on h;
The size of each high frequency subgraph is obtained, arithmetic mean number is calculated and obtains the normalized picture of size;
Brightness Y and R that the normalized picture of the size is established according to the variation relation of RGB and YUV color space,
G, tri- color component corresponding relationships of B express the gray value of image, used formula are as follows: Y=aR+bG+cB, wherein a, b,
C is color parameter, and 0≤a, b, c≤1 expresses the gray value of image according to brightness value Y, realizes the normalization of gray scale.
It is described in one of the embodiments, that collected described image is subjected to size and color normalized, it obtains
To normalized image, comprising:
Extract the Two-Dimensional Gabor Wavelets textural characteristics in the gray scale I (x, y) of described image;
Convolution is filtered in each scale all directions to the Wavelet Texture,
Convolution Formula are as follows: Wu,v(x, y)=I (x, y) × ψu,v(x, y),
Wu,v(x, y) indicates the texture feature vector on scale v, direction u, Wu,vAmplitude contain image local energy
Variation, Wu,vReal and imaginary parts response including Gabor core, ψu,v(x, y) indicates frequency, and I (x, y) indicates gray value;
The speed of convolutional calculation, formula are promoted using FFT transform and inverse FFT transform are as follows:
W1 u,v(x, y)=F-1(ψu,v(x,y))·F(I(x,y)),
W1 u,vIndicate that improved texture feature vector, F () indicate FFT transform, F-1() indicates inverse FFT transform, is become
U × v feature vector after changing merges the normalization that these feature vectors realize dimension of picture and gray scale.
A kind of pattern recognition device based on moire fringes, including following module:
Image set forms module, is set as obtaining several images to be processed, is packaged into an image set, and described image is concentrated
Including the image with moire fringes and without the image of moire fringes;
Training sample generation module is set as extracting the moire fringes feature that any image that described image is concentrated is included,
Form training sample;
Identification model generation module, be set as using the training sample as enter ginseng, in support vector machines model into
Row SVM training, obtains image recognition model;
Unknown images identification module is set as receiving the unknown images uploaded, according to described image identification model to described
Unknown images carry out moire fringes identification, to determine whether described image is reproduction image.
A kind of computer equipment, including memory and processor are stored with computer-readable instruction in the memory, institute
When stating computer-readable instruction and being executed by the processor, so that the processor executes the above-mentioned image recognition based on moire fringes
The step of method.
A kind of storage medium being stored with computer-readable instruction, the computer-readable instruction are handled by one or more
When device executes, so that the step of one or more processors execute the above-mentioned image-recognizing method based on moire fringes.
The above-mentioned image-recognizing method based on moire fringes, device, computer equipment and storage medium, including obtain to be processed
Several images, be packaged into an image set, it includes the image with moire fringes and the figure without moire fringes that described image, which is concentrated,
Picture;The moire fringes feature that any image that described image is concentrated is included is extracted, training sample is formed;The training sample is made
To enter ginseng, SVM training is carried out in support vector machines model, obtains image recognition model;The unknown images uploaded are received,
Moire fringes identification is carried out to the unknown images according to described image identification model, to determine whether described image is reproduction figure
Picture.The technical program during financial transaction there are problems that being cheated using reproduction picture, using LBP algorithm pair
Moire fringes are identified, and pass through SVM model training, promote speed and accuracy to reproduction photo array.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the application
Limitation.
Fig. 1 is a kind of overall flow figure of the image-recognizing method based on moire fringes of the application;
Fig. 2 is the image set forming process schematic diagram in a kind of image-recognizing method based on moire fringes of the application;
Fig. 3 is that rubbing in a kind of image-recognizing method based on moire fringes of the application identifies sample generating process schematic diagram;
Fig. 4 is a kind of structure chart of the pattern recognition device based on moire fringes of the application.
Specific embodiment
It is with reference to the accompanying drawings and embodiments, right in order to which the objects, technical solutions and advantages of the application are more clearly understood
The application is further elaborated.It should be appreciated that specific embodiment described herein is only used to explain the application, and
It is not used in restriction the application.
Those skilled in the art of the present technique are appreciated that unless expressly stated, singular " one " used herein, " one
It is a ", " described " and "the" may also comprise plural form.It is to be further understood that being arranged used in the description of the present application
Diction " comprising " refer to that there are the feature, integer, step, operation, element and/or component, but it is not excluded that in the presence of or addition
Other one or more features, integer, step, operation, element, component and/or their group.
Fig. 1 is the flow chart of the image-recognizing method based on moire fringes in the application one embodiment, as shown in Figure 1,
A kind of image-recognizing method based on moire fringes, comprising the following steps:
S1 obtains several images to be processed, is packaged into an image set, it includes the figure with moire fringes that described image, which is concentrated,
Picture and image without moire fringes;
Specifically, the image acquisition device that detection terminal is set is called to carry out Image Acquisition respectively to two groups of images, wherein
First group of image is real human face, and second group of image is reproduction photo, and image acquisition device can be ccd image collector, is led
To be made of following components: front-end optical device, ccd image acquisition module, analog-to-digital conversion module, FPGA pre-process mould
Block, Flash program memory module, DSP image processing module, SDRAM data memory module, at image display and image
Manage device.
When carrying out Image Acquisition, photosensitive element is set on ccd image collector, environmental light intensity is incuded, rear
It holds in the processor of PC machine and is provided with Intensity threshold, Intensity threshold is set as 200lx under normal conditions, when environmental light intensity is less than
Start flash lamp when 200lx, does not have to flash lamp then when greater than 200lx.
S2 extracts the moire fringes feature that any image that described image is concentrated is included, and forms training sample;
Specifically, carry out moire fringes feature extraction when extracted using LBP algorithm, local binary patterns (English:
Local binary patterns, abbreviation: LBP) it is a kind of feature that classification is used in field of machine vision, in quilt in 1994
It proposes.Local binary patterns are a very powerful feature in Texture classification problem;If local binary patterns feature with
Histograms of oriented gradients combines, then can highly effective promotion detection effect.Local binary patterns be one simply but very
Effective texture operator.The most important attribute of LBP is the robustness to grey scale change caused by illumination variation etc., it
Another key property is that its calculating is simple, this analyzes it in real time image.
S3 carries out SVM training using the training sample as ginseng is entered in support vector machines model, obtains image knowledge
Other model;
Wherein, SVM (Support Vector Machine) is a kind of supporting vector machine model, is a kind of common differentiation
Method.It is the learning model for having supervision in machine learning field, commonly used to carry out pattern-recognition, classification and recurrence
Analysis.
It is linear can a point situation analyzed, the case where for linearly inseparable, by using Nonlinear Mapping
The sample of low-dimensional input space linearly inseparable is converted high-dimensional feature space by algorithm makes its linear separability, so that higher-dimension
Feature space carries out linear analysis using nonlinear characteristic of the linear algorithm to sample and is possibly realized;And it is based on Structural risk minization
Change and construct optimal hyperlane in feature space on theory, so that learner obtains global optimization, and in entire sample
The expectation in space meets certain upper bound with some probability.
S4, the unknown images uploaded are received, moire fringes knowledge is carried out to the unknown images according to described image identification model
Not, to determine whether described image is reproduction image.
Specifically, image recognition model (i.e. trained SVM model) to be written to spare in SDK program, calling write-in
The SDK program of image recognition model carries out screenshot to unknown images, and screenshot (i.e. unknown images) is entered ginseng and is carried out into SVM model
Moire fringes identification, is reproduction image if identifying and having moire fringes in screenshot (i.e. unknown images), is otherwise real human face
Image.
In the present embodiment, moire fringes feature is identified by LBP algorithm and is trained and can effectively mention by SVM model
The efficiency for rising moire fringes identification, more rapidly and effectively identifies reproduction picture.
Fig. 2 is the image set forming process schematic diagram in a kind of image-recognizing method based on moire fringes of the application, is such as schemed
It is shown, it is described to obtain several images to be processed, it is packaged into an image set, it includes the image with moire fringes that described image, which is concentrated,
With the image without moire fringes, comprising:
S101, when any described image to image acquisition device screen distance be less than a certain distance threshold, to image carry out
Acquisition;
Specifically, can be acquired using the mode of Fixed Time Interval, i.e., when being acquired to image to needs
Each image of Image Acquisition is taken pictures twice, takes the high image of middle clarity of taking pictures twice as images to be recognized.
The angle of rotating image collector is from different perspectives to the figure of the moire fringes of being carried out identification during taking pictures twice
As being shot, image is clapped with the mode at the image for the Image Acquisition of being carried out right angle in 90 ° for example, taking for the first time
Acquisition is taken the photograph, second rotating image collector, 30 °~45° angle carries out shooting to the image of the moire fringes of being carried out identification and adopts
Collection.It, can be by different angles of incidence of light to acquired image gray scale using the Image Acquisition mode of two kinds of different angles
It has an impact, and then so that the gray scale of picture is generated difference during normalizing picture, to prevent single angle shot figure figure
The moire fringes feature as caused by does not protrude, and influences the accuracy that moire fringes feature is identified using LBP algorithm.
S102, collected described image is subjected to size and color normalized, obtains normalized image;
Specifically, the size of image is compared with pre-set dimension, obtains the figure when carrying out size normalized
As the pre-set dimension that size most carries out, image is stretched or squeezed according to this pre-set dimension reaches normalized effect.
S103, all normalized images are ranked up according to the time is generated, are packaged into an image set.
It is marked specifically, assigning normalized image with the time, the image that the same time generates is packaged into a figure in advance
As group is then packaged into an image set in the common recognition node established between each image group.
In the present embodiment, by the way that image is normalized, so that the speed and efficiency of moire fringes identification mention significantly
It rises.
The moire fringes feature extracted any image that described image is concentrated and included in one of the embodiments,
Form training sample, comprising:
By described image according to the sub-block for being horizontally and vertically divided into the sizes such as n × n, n is any positive integer;
Carry out multiple dimensioned local binary patterns LBP histogram calculation respectively to each sub-block, calculation method is as follows:
LBP of each pixel on certain scaleP,RValue are as follows:
Wherein, gcFor the gray value of pixel, gpFor with gcFor the center of circle, R is p pixel extracting on the circumference of radius
Gray value, S indicate impact factor, 2pIntermediate scheme species number, wherein p=0 ..., n;
According to the LBPP,RValue, establishes the LBP histogram of the sub-block;
Sub-block corresponding to ordinate maximum value in the LBP histogram is obtained, enters ginseng as training sample, wherein if
The LBP of sub-block corresponding to ordinate maximum value in the LBP histogramP,RBe zero or positive number then illustrate described image without
There are moire fringes, be negative, illustrates described image with moire fringes.
In the present embodiment, by being split into multiple sub-blocks to image, and each sub-block is subjected to the solution of LBP value,
It can preferably identify the moire fringes feature in image.
Fig. 3 is that rubbing in a kind of image-recognizing method based on moire fringes of the application identifies sample generating process schematic diagram,
As shown, carrying out SVM training in support vector machines model using the training sample as ginseng is entered, obtaining image recognition
Model, comprising:
S301, the training sample is obtained;
S302, low-dimensional union feature matrix is obtained after the training sample is carried out principal component analysis PCA dimensionality reduction;
Wherein, PCA is commonly used in the exploration and visualization of High Dimensional Data Set, can be also used for data compression, data are located in advance
Reason etc..PCA can be the low-dimensional variable that may have the higher-dimension variable synthesizing linear of correlation unrelated, referred to as principal component
(principal components).New low-dimensional data rally retains the variable of initial data as far as possible.
S303, the low-dimensional union feature matrix is entered to ginseng to SVM model progress classification based training, obtains several two classes
Identifier, respectively moire fringes identifier and without moire fringes identifier;
Wherein, Classification and Identification is carried out to training sample using moire fringes identifier and without moire fringes identifier, further according to knowledge
Other result is voted, and is counted to the gained vote sum of two classifications, and the classification more than poll is the training sample classification;It adopts
Sample is counted to reduce with the mode of ballot and omits significant data in statistic processes, if as many, can adopt
The foundation for using other parameters as ballot is voted until selecting classification.
S304, moire fringes are carried out to the training sample using the moire fringes identifier and the no moire fringes identifier
Identification, establishes image recognition model according to moire fringes recognition result.
Specifically, can be adopted when being trained using moire fringes identifier or without moire fringes identifier to training sample
With being first trained using moire fringes identifier to each training sample, identifies the image with moire fringes, then make
It is identified with image of no moire fringes identifier to unidentified moire fringes out;Or using to each figure in training sample
Piece first uses moire fringes identifier to identify, then is directly identified with no moire fringes identifier, then summarizes recognition result.
In the present embodiment, by carrying out dimension-reduction treatment to training sample, the efficiency of moire fringes identification can be promoted.
In one embodiment, the moire fringes feature extracted any image that described image is concentrated and included, forms
Training sample, comprising:
Described image is divided into nine trained subgraph blocks, and the trained subgraph block is divided into two groups, first group of total m block
As training subgraph block, second group of total 9-m block is as syndrome segment, wherein 1≤m≤8, and m is integer;
Specifically, dividing an image into nine sub- segments, Preferable scheme is that first group is 5 pieces, second group is 4 pieces, with
Convenient for determining the location of moire fringes, position is generated to different types of reproduction picture moire fringes and carries out position statistics, is established
The linear statistical data model of moire fringes position, and linear statistical data model is trained to obtain different types of turn over
Clap the common location that picture moire fringes occur.When being identified to new image, can first to a certain segment region into
The identification of row moire fringes, does not need to identify other regions if the region recognition goes out moire fringes.It can using such mode
To promote the efficiency of moire fringes identification, for ccd image collector high-definition, LBP feature extraction is being carried out
It does not need to extract all pixels point the time for having saved pixel extraction and gray proces during pixel.
Extract the moire fringes feature in the trained subgraph block, wherein the instruction of moire fringes is had in the trained subgraph block
Practice subgraph block and be labeled as " 1 " labeled as " 2 ", without the training subgraph block of moire fringes, establishes training histogram,;
Specifically, in histogram, the quantity of the class in each region is by input assignment grid when carrying out statistics with histogram
It determines;If specified figure layer, the symbol device of figure layer define the quantity of class;If specified data set.
The moire fringes feature in the syndrome segment is extracted, the inspection subgraph of moire fringes is had in the syndrome segment
Block is labeled as " 1 " labeled as " 2 ", without the syndrome segment of moire fringes, establishes and examines histogram;
Summarize the trained histogram and histogram is examined to obtain image histogram, if there is one on described image histogram
Subgraph block is labeled as " 2 ", then described image has moire fringes feature, and otherwise described image is without moire fringes feature, and is formed
Training sample.
In the present embodiment, further division can also be carried out to training segment, i.e., drawn each training segment homalographic
It is divided into nine trained subgraph blocks, LBP feature is extracted to each training subgraph block, equally establishes histogram, counting statistics
Histogram, the label in subgraph block with moire fringes are 2 ", are labeled as " 1 " without moire fringes, pass through the straight of statistics subgraph block
Fang Tu.
In one embodiment, described that collected described image is subjected to size and color normalized, returned
One image changed, comprising:
Carry out that scale is w, non-lower sampling NSCT that direction is h is decomposed to the picture in the picture library, obtain one it is low
On frequency subgraph A0 and multiple and different scales and different directions high frequency subgraph A1,1;A1,2;...;Aw, 1;...;Aw, h },
High frequency subgraph Aw, it is w that h, which describes scale, and direction is the face texture information on h;
The size of each high frequency subgraph is obtained, arithmetic mean number is calculated and obtains the normalized picture of size;
Brightness Y and R that the normalized picture of the size is established according to the variation relation of RGB and YUV color space,
G, tri- color component corresponding relationships of B express the gray value of image, used formula are as follows: Y=aR+bG+cB, wherein a, b,
C is color parameter, 0≤a, b, c≤1, that is, any one of a, b, c are the number between 0 to 1, express image according to brightness value Y
Gray value, realize the normalization of gray scale.
Specifically, a large amount of image descriptor (such as SIFT, HOG can be extracted at random during normalized picture
Deng), each image descriptor is a vector, is clustered using K-means clustering algorithm to these image descriptors, is obtained
To K classification (K is adjustable parameter, representative value 1024,2048,10000 etc.).Cluster centre is referred to as " word ", gathers
The all categories that class obtains form one " code book ";
For piece image, feature descriptor (such as SIFT, HOG) is extracted in dense mode;For each description
Symbol, searches for most like cluster centre (namely word) in the codebook.The frequency that different words occur in the images is counted, forms one
A histogram.L1 normalization is made to the histogram, obtains the image texture characteristic based on bag of words to the end.
In the present embodiment, being handled by size to picture and pixel can reduce what moire fringes identification process was subject to
Interference.
In one embodiment, described that collected described image is subjected to size and color normalized, returned
One image changed, comprising:
Extract the Two-Dimensional Gabor Wavelets textural characteristics in the gray scale I (x, y) of described image;
Convolution is filtered in each scale all directions to the Wavelet Texture,
Convolution Formula are as follows: Wu,v(x, y)=I (x, y) × ψu,v(x, y),
Wu,v(x, y) indicates the texture feature vector on scale v, direction u, Wu,vAmplitude contain image local energy
Variation, Wu,vReal and imaginary parts response including Gabor core, ψu,v(x, y) indicates frequency, and I (x, y) indicates gray value;
The speed of convolutional calculation, formula are promoted using FFT transform and inverse FFT transform are as follows:
W1 u,v(x, y)=F-1(ψu,v(x,y))·F(I(x,y)),
W1 u,vIndicate that improved texture feature vector, F () indicate FFT transform, F-1() indicates inverse FFT transform, is become
U × v feature vector after changing merges the normalization that these feature vectors realize dimension of picture and gray scale.
In the present embodiment, facial image is normalized by the method for Gabor wavelet kernel function, is further mentioned
Rise the precision of moire fringes identification.
In one embodiment it is proposed that a kind of pattern recognition device based on moire fringes, as shown in figure 4, including as follows
Module:
Image set forms module, is set as obtaining several images to be processed, is packaged into an image set, and described image is concentrated
Including the image with moire fringes and without the image of moire fringes;
Training sample generation module is set as extracting the moire fringes feature that any image that described image is concentrated is included,
Form training sample;
Identification model generation module, be set as using the training sample as enter ginseng, in support vector machines model into
Row SVM training, obtains image recognition model;
Unknown images identification module is set as receiving the unknown images uploaded, according to described image identification model to described
Unknown images carry out moire fringes identification, to determine whether described image is reproduction image.
A kind of computer equipment, including memory and processor are stored with computer-readable instruction in the memory, institute
When stating computer-readable instruction and being executed by the processor, so that the processor described being based on of executing in the various embodiments described above
The step of image-recognizing method of moire fringes.
A kind of storage medium being stored with computer-readable instruction, the computer-readable instruction are handled by one or more
When device executes, so that one or more processors execute the image-recognizing method based on moire fringes in the various embodiments described above
The step of.The storage medium can be non-volatile memory medium.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage
Medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random
Access Memory), disk or CD etc..
Each technical characteristic of embodiment described above can be combined arbitrarily, for simplicity of description, not to above-mentioned reality
It applies all possible combination of each technical characteristic in example to be all described, as long as however, the combination of these technical characteristics is not deposited
In contradiction, all should be considered as described in this specification.
The some exemplary embodiments of the application above described embodiment only expresses, the description thereof is more specific and detailed, but
It cannot be understood as the limitations to the application the scope of the patents.It should be pointed out that for the ordinary skill people of this field
For member, without departing from the concept of this application, various modifications and improvements can be made, these belong to the application's
Protection scope.Therefore, the scope of protection shall be subject to the appended claims for the application patent.
Claims (10)
1. a kind of image-recognizing method based on moire fringes, which comprises the steps of:
Several images to be processed are obtained, are packaged into an image set, it includes image with moire fringes and not that described image, which is concentrated,
Image with moire fringes;
The moire fringes feature that any image that described image is concentrated is included is extracted, training sample is formed;
Using the training sample as ginseng is entered, SVM training is carried out in support vector machines model, obtains image recognition model;
The unknown images uploaded are received, moire fringes identification are carried out to the unknown images according to described image identification model, with true
Determine whether described image is reproduction image.
2. the image-recognizing method according to claim 1 based on moire fringes, which is characterized in that the acquisition is to be processed
Several images, are packaged into an image set, and it includes the image with moire fringes and the image without moire fringes that described image, which is concentrated,
Include:
When the distance of any image to image acquisition device screen is less than a certain distance threshold, described image is acquired;
Collected described image is subjected to size and color normalized, obtains normalized image;
All normalized images are ranked up according to the time is generated, are packaged into an image set.
3. the image-recognizing method according to claim 1 based on moire fringes, which is characterized in that the extraction described image
The moire fringes feature that any image of concentration is included forms training sample, comprising:
By described image according to the sub-block for being horizontally and vertically divided into the sizes such as n × n, n is any positive integer;
Carry out multiple dimensioned local binary patterns LBP histogram calculation respectively to each sub-block, calculation method is as follows:
LBP of each pixel on certain scaleP,RValue are as follows:
Wherein, gcFor the gray value of pixel, gpFor with gcFor the center of circle, R is the ash of the p pixel extracted on the circumference of radius
Angle value, S expression impact factor, 2pIntermediate scheme species number, wherein p=0 ..., n;
According to the LBPP,RValue, establishes the LBP histogram of the sub-block;
Sub-block corresponding to ordinate maximum value in the LBP histogram is obtained, enters ginseng as training sample, wherein if described
The LBP of sub-block corresponding to ordinate maximum value in LBP histogramP,RIt is zero or positive number then illustrates described image without rubbing
That line, is negative, and illustrates described image with moire fringes.
4. the image-recognizing method according to claim 1 based on moire fringes, which is characterized in that described by the trained sample
This conduct enters ginseng, and SVM training is carried out in support vector machines model, obtains image recognition model, comprising:
Obtain the training sample;
Low-dimensional union feature matrix is obtained after the training sample is carried out principal component analysis PCA dimensionality reduction;
The low-dimensional union feature matrix is entered into ginseng and carries out classification based training to the SVM model, obtains several two classes identifiers, point
It Wei not moire fringes identifier and without moire fringes identifier;
Moire fringes identification is carried out to the training sample using the moire fringes identifier and the no moire fringes identifier, according to
Moire fringes recognition result establishes image recognition model.
5. the image-recognizing method according to claim 1 based on moire fringes, which is characterized in that the extraction described image
The moire fringes feature that any image of concentration is included forms training sample, comprising:
Described image is divided into nine trained subgraph blocks, and the trained subgraph block is divided into two groups, first group of total m block conduct
Training subgraph block, second group of total 9-m block is as syndrome segment, wherein 1≤m≤8, and m is integer;
Extract the moire fringes feature in the trained subgraph block, wherein training in the trained subgraph block with moire fringes
Segment is labeled as " 1 " labeled as " 2 ", without the training subgraph block of moire fringes, establishes training histogram;
The moire fringes feature in the syndrome segment is extracted, the syndrome segment mark of moire fringes is had in the syndrome segment
It is denoted as " 2 ", is labeled as " 1 " without the syndrome segment of moire fringes, establish and examine histogram;
Summarize the trained histogram and histogram is examined to obtain image histogram, if there is a subgraph on described image histogram
Block is labeled as " 2 ", then described image has moire fringes feature, and otherwise described image is without moire fringes feature, and forms training
Sample.
6. the image-recognizing method according to claim 2 based on moire fringes, which is characterized in that described by collected institute
It states image and carries out size and color normalized, obtain normalized image, comprising:
The non-lower sampling NSCT that scale is w, direction is h is carried out to the picture in the picture library to decompose, and obtains low frequency
Scheme on A0 and multiple and different scales and different directions high frequency subgraph A1,1;A1,2;...;Aw, 1;...;Aw, h }, high frequency
Subgraph Aw, it is w that h, which describes scale, and direction is the face texture information on h;
The size of each high frequency subgraph is obtained, arithmetic mean number is calculated and obtains the normalized picture of size;
The brightness Y and R, G, B that the normalized picture of the size is established according to the variation relation of RGB and YUV color space
Three color component corresponding relationships express the gray value of image, used formula are as follows: Y=aR+bG+cB, wherein a, b, c are
Color parameter, 0≤a, b, c≤1 express the gray value of image according to brightness value Y, realize the normalization of gray scale.
7. the image-recognizing method according to claim 2 based on moire fringes, which is characterized in that described by collected described image
Size and color normalized are carried out, normalized image is obtained, comprising:
Extract the Two-Dimensional Gabor Wavelets textural characteristics in the gray scale I (x, y) of described image;
Convolution is filtered in each scale all directions to the Wavelet Texture,
Convolution Formula are as follows: Wu,v(x, y)=I (x, y) × ψu,v(x, y),
Wu,v(x, y) indicates the texture feature vector on scale v, direction u, Wu,vAmplitude contain the change of image local energy
Change, Wu,vReal and imaginary parts response including Gabor core, ψu,v(x, y) indicates frequency, and I (x, y) indicates gray value;
The speed of convolutional calculation, formula are promoted using FFT transform and inverse FFT transform are as follows:
W1 u,v(x, y)=F-1(ψu,v(x,y))·F(I(x,y)),
W1 u,vIndicate that improved texture feature vector, F () indicate FFT transform, F-1() indicates inverse FFT transform, after being converted
U × v feature vector, merge these feature vectors realize dimension of picture and gray scale normalization.
8. a kind of pattern recognition device based on moire fringes, which is characterized in that including following module:
Image set forms module, is set as obtaining several images to be processed, is packaged into an image set, and described image concentration includes
Image with moire fringes and the image without moire fringes;
Training sample generation module is set as extracting the moire fringes feature that any image that described image is concentrated is included, be formed
Training sample;
Identification model generation module is set as carrying out in support vector machines model using the training sample as ginseng is entered
SVM training, obtains image recognition model;
Unknown images identification module is set as receiving the unknown images uploaded, according to described image identification model to described unknown
Image carries out moire fringes identification, to determine whether described image is reproduction image.
9. a kind of computer equipment, including memory and processor, it is stored with computer-readable instruction in the memory, it is described
When computer-readable instruction is executed by the processor, so that the processor executes such as any one of claims 1 to 7 right
It is required that the step of image-recognizing method based on moire fringes.
10. a kind of storage medium for being stored with computer-readable instruction, the computer-readable instruction is handled by one or more
When device executes, so that one or more processors are executed is based on moire fringes as described in any one of claims 1 to 7 claim
Image-recognizing method the step of.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811206940.0A CN109558794B (en) | 2018-10-17 | 2018-10-17 | Moire-based image recognition method, device, equipment and storage medium |
PCT/CN2018/124583 WO2020077866A1 (en) | 2018-10-17 | 2018-12-28 | Moire-based image recognition method and apparatus, and device and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811206940.0A CN109558794B (en) | 2018-10-17 | 2018-10-17 | Moire-based image recognition method, device, equipment and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109558794A true CN109558794A (en) | 2019-04-02 |
CN109558794B CN109558794B (en) | 2024-06-28 |
Family
ID=65865032
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811206940.0A Active CN109558794B (en) | 2018-10-17 | 2018-10-17 | Moire-based image recognition method, device, equipment and storage medium |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN109558794B (en) |
WO (1) | WO2020077866A1 (en) |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110263708A (en) * | 2019-06-19 | 2019-09-20 | 郭玮强 | Image sources recognition methods, equipment and computer readable storage medium |
CN110348511A (en) * | 2019-07-08 | 2019-10-18 | 创新奇智(青岛)科技有限公司 | A kind of picture reproduction detection method, system and electronic equipment |
CN110569716A (en) * | 2019-07-26 | 2019-12-13 | 浙江工业大学 | Goods shelf image copying detection method |
CN110705608A (en) * | 2019-09-12 | 2020-01-17 | 杭州惠合信息科技有限公司 | Retail terminal display shelf reproduction identification method and device |
CN110781823A (en) * | 2019-10-25 | 2020-02-11 | 北京字节跳动网络技术有限公司 | Screen recording detection method and device, readable medium and electronic equipment |
CN110992244A (en) * | 2019-10-31 | 2020-04-10 | 上海扩博智能技术有限公司 | Picture generation method, system, equipment and storage medium with Moire patterns |
CN111062922A (en) * | 2019-12-14 | 2020-04-24 | 创新奇智(北京)科技有限公司 | Method and system for judging copied image and electronic equipment |
CN111368944A (en) * | 2020-05-27 | 2020-07-03 | 支付宝(杭州)信息技术有限公司 | Method and device for recognizing copied image and certificate photo and training model and electronic equipment |
CN111798380A (en) * | 2019-04-09 | 2020-10-20 | Tcl集团股份有限公司 | Image processing method, image processing device and terminal equipment |
CN111814551A (en) * | 2020-06-06 | 2020-10-23 | 中信百信银行股份有限公司 | Human face recognition method, device and equipment based on small program and storage medium |
CN112085094A (en) * | 2020-09-08 | 2020-12-15 | 中国平安财产保险股份有限公司 | Document image reproduction detection method and device, computer equipment and storage medium |
TWI716111B (en) * | 2019-09-23 | 2021-01-11 | 大陸商北京集創北方科技股份有限公司 | Image acquisition quality evaluation method and system |
CN112257709A (en) * | 2020-10-23 | 2021-01-22 | 北京云杉世界信息技术有限公司 | Signboard photo auditing method and device, electronic equipment and readable storage medium |
CN112562028A (en) * | 2020-12-03 | 2021-03-26 | 科大讯飞股份有限公司 | Method, device and equipment for constructing image with moire fringes and storage medium |
CN113486861A (en) * | 2021-08-03 | 2021-10-08 | 北京百度网讯科技有限公司 | Moire pattern picture generation method and device |
CN113554252A (en) * | 2020-04-23 | 2021-10-26 | 广西民族师范学院 | Laboratory management system |
CN113657327A (en) * | 2021-08-24 | 2021-11-16 | 平安科技(深圳)有限公司 | Non-living attack discrimination method, device, equipment and medium suitable for image |
CN113723516A (en) * | 2021-08-31 | 2021-11-30 | 平安科技(深圳)有限公司 | Moire pattern removing method, device, equipment and medium based on neural network |
CN114677769A (en) * | 2022-04-08 | 2022-06-28 | 中国平安人寿保险股份有限公司 | Method and device for identifying copied certificate, computer equipment and storage medium |
CN116109878A (en) * | 2023-04-10 | 2023-05-12 | 南京掌控网络科技有限公司 | Image reproduction identification method, system, device and storage medium |
CN116168038A (en) * | 2023-04-26 | 2023-05-26 | 创新奇智(青岛)科技有限公司 | Image reproduction detection method and device, electronic equipment and storage medium |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111860498B (en) * | 2020-07-01 | 2023-12-19 | 广州大学 | Method, device and storage medium for generating antagonism sample of license plate |
CN111899232B (en) * | 2020-07-20 | 2023-07-04 | 广西大学 | Method for nondestructive detection of bamboo-wood composite container bottom plate by image processing |
CN112070714B (en) * | 2020-07-29 | 2024-02-20 | 西安工业大学 | Method for detecting flip image based on local ternary counting feature |
CN112070116B (en) * | 2020-08-05 | 2023-06-16 | 湖北工业大学 | Automatic artistic drawing classification system and method based on support vector machine |
CN112330648B (en) * | 2020-11-12 | 2024-01-05 | 深圳大学 | Non-reference image quality evaluation method and device based on semi-supervised learning |
CN112633082B (en) * | 2020-12-04 | 2023-08-18 | 西安理工大学 | Multi-feature fusion weed detection method |
CN113111888B (en) * | 2021-04-15 | 2024-04-26 | 广州图匠数据科技有限公司 | Picture discrimination method and device |
CN113538369B (en) * | 2021-07-14 | 2024-02-09 | 安徽炬视科技有限公司 | Pressing plate state detection algorithm based on ellipse measurement learning |
CN113723515B (en) * | 2021-08-31 | 2023-08-18 | 平安科技(深圳)有限公司 | Moire pattern recognition method, device, equipment and medium based on image recognition |
CN114005019B (en) * | 2021-10-29 | 2023-09-22 | 北京有竹居网络技术有限公司 | Method for identifying flip image and related equipment thereof |
CN114022383B (en) * | 2021-11-05 | 2024-07-26 | 瞬联软件科技(北京)有限公司 | Method and device for removing mole patterns of text and image and electronic equipment |
CN116309454B (en) * | 2023-03-16 | 2023-09-19 | 首都师范大学 | Intelligent pathological image recognition method and device based on lightweight convolution kernel network |
CN117372283B (en) * | 2023-11-06 | 2024-06-04 | 上海衡亮电子科技股份有限公司 | Method and device for removing moire |
CN117611849A (en) * | 2023-11-23 | 2024-02-27 | 朋友电力科技有限公司 | Connection point positioning method and system for realizing wire clamps based on artificial intelligence |
CN117333762B (en) * | 2023-12-02 | 2024-03-15 | 深圳爱莫科技有限公司 | Image reproduction identification method based on multi-feature fusion |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105118048A (en) * | 2015-07-17 | 2015-12-02 | 北京旷视科技有限公司 | Method and device for identifying copying certificate image |
CN106778525A (en) * | 2016-11-25 | 2017-05-31 | 北京旷视科技有限公司 | Identity identifying method and device |
CN106951869A (en) * | 2017-03-22 | 2017-07-14 | 腾讯科技(深圳)有限公司 | A kind of live body verification method and equipment |
CN107103266A (en) * | 2016-02-23 | 2017-08-29 | 中国科学院声学研究所 | The training of two-dimension human face fraud detection grader and face fraud detection method |
CN108229120A (en) * | 2017-09-07 | 2018-06-29 | 北京市商汤科技开发有限公司 | Face unlock and its information registering method and device, equipment, program, medium |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104268539B (en) * | 2014-10-17 | 2017-10-31 | 中国科学技术大学 | A kind of high performance face identification method and system |
CN107798281B (en) * | 2016-09-07 | 2021-10-08 | 北京眼神科技有限公司 | Face living body detection method and device based on LBP (local binary pattern) characteristics |
CN106599872A (en) * | 2016-12-23 | 2017-04-26 | 北京旷视科技有限公司 | Method and equipment for verifying living face images |
CN108197543A (en) * | 2017-12-22 | 2018-06-22 | 深圳云天励飞技术有限公司 | Image filtering method and device, embedded device and storage medium |
-
2018
- 2018-10-17 CN CN201811206940.0A patent/CN109558794B/en active Active
- 2018-12-28 WO PCT/CN2018/124583 patent/WO2020077866A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105118048A (en) * | 2015-07-17 | 2015-12-02 | 北京旷视科技有限公司 | Method and device for identifying copying certificate image |
CN107103266A (en) * | 2016-02-23 | 2017-08-29 | 中国科学院声学研究所 | The training of two-dimension human face fraud detection grader and face fraud detection method |
CN106778525A (en) * | 2016-11-25 | 2017-05-31 | 北京旷视科技有限公司 | Identity identifying method and device |
CN106951869A (en) * | 2017-03-22 | 2017-07-14 | 腾讯科技(深圳)有限公司 | A kind of live body verification method and equipment |
CN108229120A (en) * | 2017-09-07 | 2018-06-29 | 北京市商汤科技开发有限公司 | Face unlock and its information registering method and device, equipment, program, medium |
Cited By (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111798380A (en) * | 2019-04-09 | 2020-10-20 | Tcl集团股份有限公司 | Image processing method, image processing device and terminal equipment |
CN110263708B (en) * | 2019-06-19 | 2020-03-13 | 郭玮强 | Image source identification method, device and computer readable storage medium |
CN110263708A (en) * | 2019-06-19 | 2019-09-20 | 郭玮强 | Image sources recognition methods, equipment and computer readable storage medium |
CN110348511A (en) * | 2019-07-08 | 2019-10-18 | 创新奇智(青岛)科技有限公司 | A kind of picture reproduction detection method, system and electronic equipment |
CN110569716A (en) * | 2019-07-26 | 2019-12-13 | 浙江工业大学 | Goods shelf image copying detection method |
CN110705608A (en) * | 2019-09-12 | 2020-01-17 | 杭州惠合信息科技有限公司 | Retail terminal display shelf reproduction identification method and device |
TWI716111B (en) * | 2019-09-23 | 2021-01-11 | 大陸商北京集創北方科技股份有限公司 | Image acquisition quality evaluation method and system |
CN110781823A (en) * | 2019-10-25 | 2020-02-11 | 北京字节跳动网络技术有限公司 | Screen recording detection method and device, readable medium and electronic equipment |
CN110781823B (en) * | 2019-10-25 | 2022-07-26 | 北京字节跳动网络技术有限公司 | Screen recording detection method and device, readable medium and electronic equipment |
CN110992244A (en) * | 2019-10-31 | 2020-04-10 | 上海扩博智能技术有限公司 | Picture generation method, system, equipment and storage medium with Moire patterns |
CN110992244B (en) * | 2019-10-31 | 2023-10-13 | 上海扩博智能技术有限公司 | Picture generation method, system, device and storage medium with mole pattern |
CN111062922A (en) * | 2019-12-14 | 2020-04-24 | 创新奇智(北京)科技有限公司 | Method and system for judging copied image and electronic equipment |
CN111062922B (en) * | 2019-12-14 | 2024-02-20 | 创新奇智(北京)科技有限公司 | Method and system for distinguishing flip image and electronic equipment |
CN113554252A (en) * | 2020-04-23 | 2021-10-26 | 广西民族师范学院 | Laboratory management system |
CN111368944A (en) * | 2020-05-27 | 2020-07-03 | 支付宝(杭州)信息技术有限公司 | Method and device for recognizing copied image and certificate photo and training model and electronic equipment |
CN111814551A (en) * | 2020-06-06 | 2020-10-23 | 中信百信银行股份有限公司 | Human face recognition method, device and equipment based on small program and storage medium |
CN112085094B (en) * | 2020-09-08 | 2024-04-05 | 中国平安财产保险股份有限公司 | Document image reproduction detection method, device, computer equipment and storage medium |
CN112085094A (en) * | 2020-09-08 | 2020-12-15 | 中国平安财产保险股份有限公司 | Document image reproduction detection method and device, computer equipment and storage medium |
CN112257709B (en) * | 2020-10-23 | 2024-05-07 | 北京云杉世界信息技术有限公司 | Signboard photo auditing method and device, electronic equipment and readable storage medium |
CN112257709A (en) * | 2020-10-23 | 2021-01-22 | 北京云杉世界信息技术有限公司 | Signboard photo auditing method and device, electronic equipment and readable storage medium |
CN112562028A (en) * | 2020-12-03 | 2021-03-26 | 科大讯飞股份有限公司 | Method, device and equipment for constructing image with moire fringes and storage medium |
CN113486861A (en) * | 2021-08-03 | 2021-10-08 | 北京百度网讯科技有限公司 | Moire pattern picture generation method and device |
CN113657327B (en) * | 2021-08-24 | 2024-03-08 | 平安科技(深圳)有限公司 | Non-living body attack discrimination method, device, equipment and medium suitable for image |
CN113657327A (en) * | 2021-08-24 | 2021-11-16 | 平安科技(深圳)有限公司 | Non-living attack discrimination method, device, equipment and medium suitable for image |
CN113723516A (en) * | 2021-08-31 | 2021-11-30 | 平安科技(深圳)有限公司 | Moire pattern removing method, device, equipment and medium based on neural network |
CN113723516B (en) * | 2021-08-31 | 2024-05-14 | 平安科技(深圳)有限公司 | Moire removing method, device, equipment and medium based on neural network |
CN114677769A (en) * | 2022-04-08 | 2022-06-28 | 中国平安人寿保险股份有限公司 | Method and device for identifying copied certificate, computer equipment and storage medium |
CN116109878A (en) * | 2023-04-10 | 2023-05-12 | 南京掌控网络科技有限公司 | Image reproduction identification method, system, device and storage medium |
CN116109878B (en) * | 2023-04-10 | 2023-09-08 | 南京掌控网络科技有限公司 | Image reproduction identification method, system, device and storage medium |
CN116168038A (en) * | 2023-04-26 | 2023-05-26 | 创新奇智(青岛)科技有限公司 | Image reproduction detection method and device, electronic equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
WO2020077866A1 (en) | 2020-04-23 |
CN109558794B (en) | 2024-06-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109558794A (en) | Image-recognizing method, device, equipment and storage medium based on moire fringes | |
CN105518668B (en) | Content-based image retrieval | |
Wang et al. | Visual saliency by selective contrast | |
Liu et al. | Evaluation of LBP and deep texture descriptors with a new robustness benchmark | |
Hasija et al. | Fish species classification using graph embedding discriminant analysis | |
Gomez et al. | A fast hierarchical method for multi-script and arbitrary oriented scene text extraction | |
Lin et al. | Live Face Verification with Multiple Instantialized Local Homographic Parameterization. | |
CN111274883B (en) | Synthetic sketch face recognition method based on multi-scale HOG features and deep features | |
Geetha et al. | Gender classification from face images by mixing the classifier outcome of prime, distinct descriptors | |
Wang et al. | Research on face recognition technology based on PCA and SVM | |
CN116188956A (en) | Method and related equipment for detecting deep fake face image | |
CN110222660B (en) | Signature authentication method and system based on dynamic and static feature fusion | |
Alphonse et al. | A novel maximum and minimum response-based Gabor (MMRG) feature extraction method for facial expression recognition | |
Shi et al. | Region contrast and supervised locality-preserving projection-based saliency detection | |
Zhang et al. | Scene categorization by deeply learning gaze behavior in a semisupervised context | |
Hsieh et al. | Video-based human action and hand gesture recognition by fusing factored matrices of dual tensors | |
Wei et al. | A novel magnification-robust network with sparse self-attention for micro-expression recognition | |
Dhar et al. | Fish image classification by XgBoost based on Gist and GLCM Features | |
Tian et al. | Improved bag-of-words model for person re-identification | |
CN112183422A (en) | Human face living body detection method and device based on space-time characteristics, electronic equipment and storage medium | |
Duan et al. | Bio-inspired visual attention model and saliency guided object segmentation | |
Chakraborty et al. | Scene text detection using modified histogram of oriented gradient approach | |
CN117456284B (en) | Image classification method, device, equipment and storage medium | |
Tong et al. | Wide or narrow? a visual attention inspired model for view-type classification | |
Awad et al. | A new hybrid texture-perceptual descriptor: application cbir |
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 |