CN113420582A - Anti-counterfeiting detection method and system for palm vein recognition - Google Patents
Anti-counterfeiting detection method and system for palm vein recognition Download PDFInfo
- Publication number
- CN113420582A CN113420582A CN202011215734.3A CN202011215734A CN113420582A CN 113420582 A CN113420582 A CN 113420582A CN 202011215734 A CN202011215734 A CN 202011215734A CN 113420582 A CN113420582 A CN 113420582A
- Authority
- CN
- China
- Prior art keywords
- palm
- palm vein
- region
- counterfeiting detection
- characteristic data
- 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
Images
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Collating Specific Patterns (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to an anti-counterfeiting detection method and an anti-counterfeiting detection system for palm vein identification. The method comprises the following steps: a region positioning step of positioning a palm position in the input palm vein image; a region extraction step of extracting a palm region from the palm position; a time sequence processing step, namely performing time sequence processing on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and an anti-counterfeiting detection step, which is used for realizing the anti-counterfeiting detection of the palm vein image by using a specified algorithm model based on the time-varying characteristic data. According to the invention, the anti-counterfeiting detection of the palm vein characteristic can be realized by detecting the time-varying characteristic of the infrared light reflection characteristic of the palm.
Description
Technical Field
The invention relates to a computer technology, in particular to an anti-counterfeiting detection method and an anti-counterfeiting detection system for palm vein identification.
Background
Although the palm vein recognition system is widely used in many fields, the vulnerability of the palm vein recognition system increases as it is popularized. Even if the metacarpal veins are hidden inside the human body and are invisible to the naked eye, they may still pose a safety risk due to the theft of the veins.
However, it has been found that the use of printed palm vein images can also cause an attack on the palm vein recognition system, which indicates that there is a need for improved vulnerability of the palm vein recognition system.
Disclosure of Invention
In view of the above problems, the present invention is directed to an anti-counterfeiting detection method for palm vein recognition and an anti-counterfeiting detection system for palm vein recognition, which can realize authenticity identification of palm veins.
The invention provides an anti-counterfeiting detection method for palm vein recognition, which is characterized by comprising the following steps:
a region positioning step of positioning a palm position in the input palm vein image;
a region extraction step of extracting a palm region from the palm position;
a time sequence processing step, namely performing time sequence processing on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and
and an anti-counterfeiting detection step, namely, based on the time-varying characteristic data, utilizing a specified algorithm model to realize anti-counterfeiting detection of the palm vein image.
Optionally, in the region positioning step, the positioning of the palm position is implemented by using a yolo network structure or a FastRCNN algorithm.
Optionally, in the region extracting step, a palm edge feature is extracted from the palm position, and a non-palm region is set to be black according to the palm edge feature, so as to generate a black-and-white picture;
and setting the non-palm area as black according to the original picture and the black and white picture by using the black and white picture as a mask so as to extract the palm area.
Optionally, in the region extraction step, extracting a palm edge feature from the palm position, obtaining a part similar to the palm gray scale according to the palm edge feature and the skin color, and setting a non-palm region to be black to generate a black-and-white picture; and setting the non-palm area as black according to the original picture and the black and white picture by using the black and white picture as a mask so as to extract the palm area.
Optionally, in the region extraction step, a Canny operator or a Sobel operator is used to extract the palm edge feature.
Optionally, in the region extracting step, a convolutional neural network FCN for semantic segmentation of the image is used to extract the palm region.
Optionally, before the area positioning step, the method further comprises:
and an image acquisition step, wherein a palm vein image is acquired and reflection characteristic data of the palm vein image is obtained.
Optionally, in the image acquiring step, infrared light reflection characteristic data of palm vein images in different time sequences are acquired respectively, the palm region is extracted after the acquired palm vein images are processed by the region positioning step and the region extracting step, and in the time sequence processing step, time sequence processing is performed on the infrared light reflection characteristic data in the palm region to obtain time-varying characteristic data.
Optionally, in the time-series processing step, time-varying characteristic data is obtained by using an image difference method.
Optionally, in the time-series processing step, a plurality of palm vein images at a predetermined time interval are acquired, and a difference between pixel values of two palm vein images having a largest sum of absolute values of pixel difference values in the plurality of palm vein images is determined as time-varying characteristic data.
Optionally, the prescribed algorithm model is formed based on a real photographed palm vein image and a forged palm vein image using deep neural network training.
Optionally, the prescribed algorithm model is formed based on a real captured palm vein image and a forged palm vein image trained using a convolutional neural network ImageNet.
The anti-counterfeiting detection system for palm vein recognition in one aspect of the invention is characterized by comprising:
the area positioning module is used for positioning the palm position in the input palm vein image;
the region extraction module is used for extracting a palm region from the palm position;
the time sequence processing module is used for carrying out time sequence processing on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and
and the anti-counterfeiting detection module is used for realizing anti-counterfeiting detection of the palm vein image by utilizing a specified algorithm model based on the time-varying characteristic data.
Optionally, the region positioning module uses a yolo network structure or a FastRCNN algorithm to position the palm position.
Optionally, the region extraction module extracts a palm edge feature from the palm position, sets the non-palm region to black according to the palm edge feature to generate a black-and-white picture, and sets the non-palm region to black by using the black-and-white picture as a mask according to the original picture and the black-and-white picture to extract the palm region.
Optionally, the region extraction module extracts a palm edge feature from the palm position, sets a non-palm region to black according to the palm edge feature and a portion obtained by skin color and similar to the palm gray level to generate a black-and-white picture, and sets the non-palm region to black according to the original picture and the black-and-white picture by using the black-and-white picture as a mask to extract the palm region.
Optionally, the region extraction module extracts the palm edge feature by using a Canny operator or a Sobel operator.
Optionally, the region extraction module extracts the palm region by using a convolutional neural network FCN for semantic segmentation of the image.
Optionally, further comprising:
and the image acquisition module is used for acquiring the palm vein image and acquiring the reflection characteristic data of the palm vein image to input the reflection characteristic data into the area positioning module.
Optionally, the image acquisition module respectively acquires infrared light reflection characteristic data of palm vein images with different time sequences, the palm vein images are processed by the region positioning module and the region extraction module to extract a palm region, and the time sequence processing module performs time sequence processing on the infrared light reflection characteristic data in the palm region to obtain time-varying characteristic data.
Optionally, the time-series processing module obtains the time-varying characteristic data by using an image difference method.
Optionally, the timing processing module acquires a plurality of palm vein images at a predetermined time interval, and determines a difference between pixel values of two palm vein images with the largest sum of absolute values of pixel differences in the plurality of palm vein images as time-varying characteristic data.
Optionally, the prescribed algorithm model is formed based on a real photographed palm vein image and a forged palm vein image using deep neural network training.
Optionally, the prescribed algorithm model is formed based on a real captured palm vein image and a forged palm vein image trained using a convolutional neural network ImageNet.
A computer-readable medium of an aspect of the present invention, on which a computer program is stored, is characterized in that the computer program, when being executed by a processor, implements the above-mentioned anti-counterfeiting detection method for palm vein identification.
The computer device of one aspect of the present invention includes a storage module, a processor, and a computer program stored on the storage module and executable on the processor, and is characterized in that the processor implements the above-mentioned anti-counterfeiting detection method for palm vein identification when executing the computer program.
According to the anti-counterfeiting detection method and the anti-counterfeiting detection system for the palm vein recognition, the anti-counterfeiting recognition of the palm vein feature is realized by detecting the time-varying characteristic of the infrared light reflection feature of the palm, the authenticity recognition of the palm vein feature can be completed in a short time, and a user does not need to perform special action coordination.
Drawings
Fig. 1 is a schematic flow chart showing an anti-counterfeit detection method for palm vein recognition according to an embodiment of the present invention.
Fig. 2 is a schematic diagram showing the positioning of the palm position in the palm vein image.
Fig. 3 is a schematic diagram showing extraction of a palm region from a palm position.
Fig. 4 is a schematic diagram showing a deep neural network structure.
Fig. 5 is a block diagram showing the configuration of an anti-counterfeit detection system for palm vein recognition according to an embodiment of the present invention.
Detailed Description
The following description is of some of the several embodiments of the invention and is intended to provide a basic understanding of the invention. It is not intended to identify key or critical elements of the invention or to delineate the scope of the invention.
For the purposes of brevity and explanation, the principles of the present invention are described herein with reference primarily to exemplary embodiments thereof. However, those skilled in the art will readily recognize that the same principles are equally applicable to all types of anti-counterfeiting detection methods for palm vein identification and anti-counterfeiting detection systems for palm vein identification, and that these same principles, as well as any such variations, may be implemented therein without departing from the true spirit and scope of the present patent application.
Moreover, in the following description, reference is made to the accompanying drawings that illustrate certain exemplary embodiments. Electrical, mechanical, logical, and structural changes may be made to these embodiments without departing from the spirit and scope of the invention. In addition, while a feature of the invention may have been disclosed with respect to only one of several implementations/embodiments, such feature may be combined with one or more other features of the other implementations/embodiments as may be desired and/or advantageous for any given or identified function. The following description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is defined by the appended claims and their equivalents.
Terms such as "comprising" and "comprises" mean that, in addition to having elements (modules) and modules directly and explicitly stated in the description and claims, the solution of the invention does not exclude the presence of other elements (modules) and modules not directly or explicitly stated.
Before describing the anti-counterfeiting detection method for palm vein recognition and the anti-counterfeiting detection system for palm vein recognition of the present invention, some technical terms in the field are described.
(1) Palm vein image and palm vein recognition
The palm vein recognition utilizes near infrared rays to irradiate the palm, and a sensor senses light reflected by the palm, wherein hemoglobin flowing into venous red blood cells absorbs the infrared rays, so that the reflection of the vein part is less, and a vein pattern, namely a palm vein image, can be generated on an image. The main characteristics of the palm vein recognition are non-invasive image acquisition, and the palm vein characteristics cannot be acquired under visible light, so that the palm vein recognition method has strong concealment and anti-counterfeiting performance.
（2）CNN（Convolution Neural Network）
A convolutional neural network comprising a feed-forward neural network of convolutional computation and having a depth structure.
（3）yolo（You Only Look Once）
A convolutional neural network for image object detection.
yolo is a target detection method, and the method has the characteristics of realizing rapid detection and simultaneously achieving higher accuracy. The method adopts a single neural network to directly predict the object boundary and the class probability, and realizes the end-to-end object detection. And yolo integrates target area prediction and target category prediction into a single neural network model, so that rapid target detection and identification are realized under the condition of high accuracy.
（4）FCN（Fully Convolutional Networks for Semantic Segmentation）
The full convolution neural network for image semantic segmentation only comprises a feedforward neural network with convolution structure calculation and depth structure.
（5）FastRCNN
FastRCNN is a region-based target detection algorithm.
(6) ImageNet (Picture data set)
ImageNet is a computer vision system recognition project name used in visual object recognition software research. The project has manually annotated over 1400 million images to indicate objects in the picture and provides a border in at least 100 million images. ImageNet contains 2 ten thousand typical categories.
(7) Canny operator
The Canny edge detection operator (Canny operator) is a multi-level edge detection algorithm developed by John f. Canny aims to find an optimal edge detection algorithm.
(8) Sobel operator (Sobel operator)
The sobel operator is an important processing method in the field of computer vision. The sobel operator is mainly used for edge detection. Technically, it is a discrete difference operator used to calculate the approximate value of the gradient of the image brightness function. Using this operator at any point in the image will produce the corresponding gradient vector or its normal vector. The Sobel operator not only produces a good detection effect, but also has a smooth suppression effect on noise.
The invention relates to an anti-counterfeiting detection method for palm vein recognition and a system thereof, and the inventive concept is based on the following contents: the palm is used as a human organ, complex metabolic activity occurs all the time, blood in the palm veins does not flow, the slight changes can affect the light reflecting characteristics of the palm veins and further reflect the light reflecting characteristics into the collected palm vein image, and the slight changes of the light reflecting characteristics have non-uniformity of spatial distribution and regularity in time, which cannot be counterfeited by means of laser printing or light source adjustment and the like. Therefore, the forgery of the absolute palm vein characteristics can be effectively avoided by identifying the micro change of the light reflection characteristics of the palm.
Fig. 1 is a schematic flow chart showing an anti-counterfeit detection method for palm vein recognition according to an embodiment of the present invention.
As shown in fig. 1, the anti-counterfeiting detection method for palm vein recognition according to one embodiment of the present invention includes the following steps:
step S100: collecting a palm vein image or inputting the palm vein image and obtaining the reflection characteristic data of the palm vein image;
step S200: positioning the palm position in the collected or input palm vein image;
step S300: extracting a palm area from the palm position;
step S400: time sequence processing is carried out on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and
step S500: and based on the time-varying characteristic data, utilizing a specified algorithm model to realize anti-counterfeiting detection of the palm vein image.
The steps are specifically described below.
Fig. 2 is a schematic diagram showing the positioning of the palm position in the palm vein image. As shown in fig. 2, in step S200, the palm position in the input palm vein image is located as shown in fig. 2.
Here, as an example, the positioning of the palm region may be implemented using a yolo network architecture. In the case of positioning a palm region by using a yolo network structure, the core idea of yolo is to use the whole graph as the input of the network, and directly return the position of a bounding box and the category to which the bounding box belongs in the output layer, for example, the following processes are included: the input image is divided into SxS bins, CNN extraction features and prediction, filtering bbox. The palm area positioning by using yolo has the advantages of high accuracy and high detection speed.
As yet another example, locating palm position can also be accomplished using the FastRCNN algorithm. Compared with the traditional algorithm, the Fast RCNN adopts the neural network for classification, so that the feature extraction network and the classification network can be trained simultaneously, higher accuracy is achieved, and the training speed is accelerated.
Fig. 3 is a schematic diagram showing extraction of a palm region from a palm position.
In step S300, the palm region in the palm vein image is extracted, and all the non-palm regions are set to black (255). As shown in fig. 3, by shielding the region other than the palm, the influence of the change in the background light on the discrimination effect is reduced. As an example, a palm edge feature is extracted from a palm region in a palm vein image, a black-and-white picture is generated by setting a non-palm region to black according to the palm edge feature, and the non-palm region is set to black by using the black-and-white picture as a mask according to an original picture (i.e., an acquired palm vein image) and the black-and-white picture to extract the palm region.
As another example, when the palm edge data is incomplete (the palm edge data is incomplete, for example, the palm edge line is not closed), a skin color and a gray scale may be added, the palm edge feature is extracted from the palm position, a part close to the palm gray scale is obtained according to the palm edge feature and the skin color, a non-palm region is set to be black to generate a black-and-white picture, and the non-palm region is set to be black by using the black-and-white picture as a mask to extract the palm region according to the original picture and the black-and-white picture.
In step S300, as a method for extracting a palm region, a Canny operator or a Sobel operator may be used to extract a palm edge feature. In addition, a convolutional neural network FCN or the like for semantic segmentation of an image may also be employed. The extraction of the palm edge features by using the Canny operator can be realized by the processes of graying, gaussian filtering, gradient value and direction calculation, non-maximum suppression, double-threshold selection, edge connection and the like.
The FCN classifies the image at the pixel level, so that the problem of image segmentation at the semantic level is solved, the classification of each pixel point of the image can be judged, accurate segmentation can be carried out, and the image semantic segmentation at the pixel level can be realized.
In step S400, time-series processing is performed on the reflection characteristic data in the palm region to obtain time-varying characteristic data, for example, in step S100, infrared ray reflection characteristic data of one palm vein image is collected at times T, T + S1.., T + Sn (where n is a natural number), and after the processing in steps S200 and S300, in step S400, time-series processing is performed on the infrared ray reflection characteristic data in the palm region by using an image difference method or the like to obtain time-varying characteristic data.
As the image difference method, images acquired at time T + Sp and time T + Sq may be cited, where the time difference between Sp and Sq may be within a reasonable time, and as an example, the time is 1 second, and then the difference Qpq = | Ip-Iq |, where Ip represents a palm vein image pixel value captured at time p, Iq represents a palm vein image pixel value captured at time q, Qpq represents the difference between two palm vein images, and two images with the largest sum of absolute values of pixel differences, i.e., the time-varying characteristic data, are selected as the output result.
In step S500, anti-forgery detection of the palm vein image is realized by using a predetermined algorithm model based on the time-varying characteristic data obtained in step S400. Here, as an example of specifying an algorithm model, for example, the following method can be adopted: inputting 5000 pairs of live palm vein images shot really and 5000 pairs of forged palm vein images (for example, palm vein images forged by a printing mode), generating time-varying characteristic data by adopting the same method in the step S400 for all the input palm vein images (specifically, assuming that 5000 pieces of palm vein data are collected, according to the step S400, a plurality of images are shot for each palm in a short time, then two images with the largest pixel difference value Qpq are selected from the plurality of images), differencing the palm vein images shot really (namely, each palm vein sample has two images collected at different moments, and directly differencing the pixel values of the two images to obtain a difference value image input to a network) and calibrating the difference value image as true (1), relatively differencing the forged palm vein images and calibrating the images as false (0), and zooming all the difference value images to 480 pixels (here, the deep neural network has requirements on the input image, the training program report is abnormal if the deep neural network does not meet the size requirement, the size of 640 x 480 pixels is an example, and other values can be selected), and the deep neural network is input to the deep neural network for training, and the structure of the deep neural network is shown in fig. 4.
Fig. 4 is a schematic diagram showing a deep neural network structure. As shown in fig. 4, the difference image "in fig. 4 is input to the deep neural network to perform convolution operation (conv), and finally avg pool (average pooling layer) is obtained. Deep neural networks are one technique in machine learning. As the deep neural network, for example, a convolutional neural network for image classification such as ImageNet or the like can be employed. The network loss is trained to be convergent, and the training can be set to be completed when the classification accuracy rate reaches more than 95%. The time-varying characteristic data acquired in step S400 is input to the algorithm model formed as described above, and the prediction result is output by the algorithm model as the authenticity judgment criterion.
The above describes one embodiment of the anti-counterfeit detection method for palm vein recognition according to the present invention. Next, an embodiment of the forgery prevention detection system for palm vein recognition according to the present invention will be described.
Fig. 5 is a block diagram showing the configuration of an anti-counterfeit detection system for palm vein recognition according to an embodiment of the present invention.
As shown in fig. 5, an anti-counterfeit detection system for palm vein recognition according to an embodiment of the present invention includes:
the image acquisition module 100 is used for acquiring a palm vein image and acquiring reflection characteristic data of the palm vein image;
a region locating module 200 for locating a palm position in the palm vein image input from the image capturing module 100;
a region extraction module 300, configured to extract a palm region from the palm position;
the time sequence processing module 400 is configured to perform time sequence processing on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and
and the anti-counterfeiting detection module 500 is used for realizing anti-counterfeiting detection of the palm vein image by utilizing a specified algorithm model based on the time-varying characteristic data.
The area positioning module 200 uses a yolo network structure or a FastRCNN algorithm to position the palm position. The region extraction module 200 extracts the palm edge feature from the palm position, generates a black-and-white picture according to the palm edge feature and a portion of the skin close to the palm gray scale, and sets the non-palm region to be black according to the original picture and the black-and-white picture by using the black-and-white picture as a mask to extract the palm region. As an example, the region extraction module 200 extracts the palm edge features using the Canny operator or Sobel operator.
The region extraction module 300 extracts the palm region by using a convolutional neural network FCN for semantic segmentation of the image.
As an example, the image capturing module 100 respectively captures infrared light reflection characteristic data of palm vein images with different time sequences, the palm vein images are processed by the region locating module 200 and the region extracting module 300 to extract a palm region, and the time sequence processing module 400 performs time sequence processing on the infrared light reflection characteristic data in the palm region to obtain time-varying characteristic data.
As one example, the timing processing module 400 obtains time-varying characteristic data using an image difference method.
In the anti-counterfeiting detection module 500, the prescribed algorithm model is formed based on a real photographed palm vein image and a forged palm vein image by using deep neural network training. As one example, the deep neural network may employ a convolutional neural network ImageNet.
As described above, since the inventor considers that biochemical activities exist in the palm and the palm slightly shakes during the palm vein data acquisition process, in the anti-counterfeiting detection method for palm vein recognition and the anti-counterfeiting detection system for palm vein recognition of the present invention, palm print images of a plurality of palm regions are acquired at different times, the same pixel in the palm print image changes with time (time change) and the change laws of different pixel points are different (spatial change), and the counterfeit data hardly simulates the time change and spatial change characteristics of a real palm, in view of such non-uniformity of spatial distribution and regularity of temporal distribution, the anti-counterfeiting identification of the palm vein characteristics can be realized by detecting the time-varying characteristic of the palm to the infrared light reflection characteristic, and the authenticity identification of the palm vein characteristics can be completed in a short time, and no special action coordination is required for the user.
The invention also provides a computer readable medium, on which a computer program is stored, wherein the computer program is executed by a processor to implement the above-mentioned anti-counterfeiting detection method for palm vein identification.
The invention also provides computer equipment which comprises a storage module, a processor and a computer program which is stored on the storage module and can run on the processor, and is characterized in that the processor executes the computer program to realize the anti-counterfeiting detection method for palm vein identification.
The above examples mainly illustrate the anti-counterfeiting detection method for palm vein identification and the anti-counterfeiting detection system for palm vein identification of the invention. Although only a few embodiments of the present invention have been described in detail, those skilled in the art will appreciate that the present invention may be embodied in many other forms without departing from the spirit or scope thereof. Accordingly, the present examples and embodiments are to be considered as illustrative and not restrictive, and various modifications and substitutions may be made therein without departing from the spirit and scope of the present invention as defined by the appended claims.
Claims (26)
1. An anti-counterfeiting detection method for palm vein identification is characterized by comprising the following steps:
a region positioning step of positioning a palm position in the input palm vein image;
a region extraction step of extracting a palm region from the palm position;
a time sequence processing step, namely performing time sequence processing on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and
and an anti-counterfeiting detection step, namely, based on the time-varying characteristic data, utilizing a specified algorithm model to realize anti-counterfeiting detection of the palm vein image.
2. The anti-counterfeiting detection method for palm vein identification according to claim 1,
in the area positioning step, positioning of the palm position is realized by adopting a yolo network structure or a FastRCNN algorithm.
3. The anti-counterfeiting detection method for palm vein recognition according to claim 1, wherein the region extraction step comprises:
extracting palm edge features from the palm position, and setting non-palm areas to be black according to the palm edge features to generate black and white pictures; and
and setting the non-palm area as black according to the original picture and the black and white picture by using the black and white picture as a mask so as to extract the palm area.
4. The anti-counterfeiting detection method for palm vein recognition according to claim 1, wherein the region extraction step comprises:
extracting palm edge features from the palm position, acquiring parts similar to the palm gray scale according to the palm edge features and skin color, and setting non-palm areas to be black so as to generate black and white pictures; and
and setting the non-palm area as black according to the original picture and the black and white picture by using the black and white picture as a mask so as to extract the palm area.
5. The anti-counterfeiting detection method for palm vein identification according to claim 3 or 4,
and in the region extraction step, extracting the palm edge features by adopting a Canny operator or a Sobel operator.
6. The anti-counterfeiting detection method for palm vein identification according to claim 3 or 4,
in the region extraction step, a convolution neural network (FCN) for image semantic segmentation is adopted to extract a palm region.
7. The anti-counterfeiting detection method for palm vein identification according to claim 1, wherein before the area positioning step, the method comprises:
and an image acquisition step, wherein a palm vein image is acquired and reflection characteristic data of the palm vein image is obtained.
8. The anti-counterfeiting detection method for palm vein identification according to claim 6,
in the image acquisition step, infrared light reflection characteristic data of palm vein images with different time sequences are respectively acquired, the palm region is extracted after the acquired palm vein images are processed by the region positioning step and the region extraction step, and time sequence processing is carried out on the infrared light reflection characteristic data in the palm region in the time sequence processing step to obtain time-varying characteristic data.
9. The anti-counterfeiting detection method for palm vein identification according to claim 1,
in the time-series processing step, time-varying characteristic data is obtained by using an image difference method.
10. The anti-counterfeiting detection method for palm vein identification according to claim 9,
in the time-series processing step, a plurality of palm vein images at a predetermined time interval are acquired, and the difference between the pixel values of two palm vein images having the largest sum of the absolute values of the pixel differences among the plurality of palm vein images is determined as time-varying characteristic data.
11. The anti-counterfeiting detection method for palm vein identification according to claim 1,
the prescribed algorithm model is formed based on a real shot palm vein image and a forged palm vein image by using deep neural network training.
12. The anti-counterfeiting detection method for palm vein identification according to claim 1,
the prescribed algorithm model is formed by training a convolutional neural network ImageNet based on a real shot palm vein image and a forged palm vein image.
13. An anti-counterfeiting detection system for palm vein identification, comprising:
the area positioning module is used for positioning the palm position in the input palm vein image;
the region extraction module is used for extracting a palm region from the palm position;
the time sequence processing module is used for carrying out time sequence processing on the light reflection characteristic data in the palm area to obtain time-varying characteristic data; and
and the anti-counterfeiting detection module is used for realizing anti-counterfeiting detection of the palm vein image by utilizing a specified algorithm model based on the time-varying characteristic data.
14. The anti-counterfeiting detection system for palm vein identification according to claim 13,
and the region positioning module realizes the positioning of the palm position by adopting a yolo network structure or a FastRCNN algorithm.
15. The anti-counterfeiting detection system for palm vein identification according to claim 13,
the region extraction module extracts palm edge features from the palm position, sets the non-palm region to be black according to the palm edge features to generate a black-and-white picture, and sets the non-palm region to be black by taking the black-and-white picture as a mask according to the original picture and the black-and-white picture to extract the palm region.
16. The anti-counterfeiting detection system for palm vein identification according to claim 13,
the region extraction module extracts the palm edge feature from the palm position, sets the non-palm region to be black according to the palm edge feature and the part which is obtained by skin color and is close to the palm gray scale so as to generate a black-white picture, and sets the non-palm region to be black according to the original picture and the black-white picture by using the black-white picture as a mask so as to extract the palm region.
17. The anti-counterfeiting detection system for palm vein identification according to claim 15 or 16,
and the region extraction module adopts a Canny operator or a Sobel operator to extract the palm edge characteristics.
18. The anti-counterfeiting detection system for palm vein identification according to claim 15 or 16,
the region extraction module adopts a convolutional neural network (FCN) for image semantic segmentation to extract a palm region.
19. The anti-counterfeiting detection system for palm vein identification according to claim 13, further comprising:
and the image acquisition module is used for acquiring the palm vein image and acquiring the reflection characteristic data of the palm vein image to input the reflection characteristic data into the area positioning module.
20. The anti-counterfeiting detection system for palm vein identification according to claim 19,
the image acquisition module respectively acquires infrared light reflection characteristic data of palm vein images with different time sequences, the palm vein images are processed by the region positioning module and the region extraction module to extract a palm region, and the time sequence processing module carries out time sequence processing on the infrared light reflection characteristic data in the palm region to obtain time-varying characteristic data.
21. The anti-counterfeiting detection system for palm vein identification according to claim 13,
the time sequence processing module adopts an image difference method to obtain time-varying characteristic data.
22. The anti-counterfeiting detection system for palm vein identification according to claim 21,
the time sequence processing module collects a plurality of palm vein images at a specified time interval, and determines that the difference between the pixel values of two palm vein images with the largest sum of the absolute values of the pixel differences in the plurality of palm vein images is time-varying characteristic data.
23. The anti-counterfeiting detection system for palm vein identification according to claim 13,
the prescribed algorithm model is formed based on a real shot palm vein image and a forged palm vein image by using deep neural network training.
24. The anti-counterfeiting detection system for palm vein identification according to claim 13,
the prescribed algorithm model is formed by training a convolutional neural network ImageNet based on a real shot palm vein image and a forged palm vein image.
25. A computer-readable medium, having stored thereon a computer program,
the computer program, when executed by a processor, implements the anti-counterfeiting detection method for palm vein recognition according to any one of claims 1 to 12.
26. A computer device comprising a storage module, a processor and a computer program stored on the storage module and executable on the processor, wherein the processor implements the anti-counterfeiting detection method for palm vein identification according to any one of claims 1 to 12 when executing the computer program.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011215734.3A CN113420582B (en) | 2020-11-04 | 2020-11-04 | Anti-fake detection method and system for palm vein recognition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011215734.3A CN113420582B (en) | 2020-11-04 | 2020-11-04 | Anti-fake detection method and system for palm vein recognition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113420582A true CN113420582A (en) | 2021-09-21 |
| CN113420582B CN113420582B (en) | 2023-09-05 |
Family
ID=77711725
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011215734.3A Active CN113420582B (en) | 2020-11-04 | 2020-11-04 | Anti-fake detection method and system for palm vein recognition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113420582B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115022605A (en) * | 2022-05-17 | 2022-09-06 | 深圳市当智科技有限公司 | Projection method, projector, and computer-readable storage medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984924A (en) * | 2014-05-08 | 2014-08-13 | 山东大学 | Reflection type finger vein recognition bioassay device and method |
| CN104778445A (en) * | 2015-03-17 | 2015-07-15 | 山东大学 | Living body recognizing device and method based on three-dimensional characteristics of finger venas |
| WO2018043796A1 (en) * | 2016-09-01 | 2018-03-08 | 상명대학교서울산학협력단 | Device and method for detecting forged bioinformation |
| WO2018061786A1 (en) * | 2016-09-27 | 2018-04-05 | 株式会社日立製作所 | Living body authentication device |
| CN109255305A (en) * | 2018-08-18 | 2019-01-22 | 深圳长元知识产权服务有限公司 | A kind of biometric devices and the shared device using the device |
| CN110321871A (en) * | 2019-07-11 | 2019-10-11 | 电子科技大学成都学院 | A kind of palm vein identification system and method based on LSTM |
| CN110321870A (en) * | 2019-07-11 | 2019-10-11 | 西北民族大学 | A kind of vena metacarpea recognition methods based on LSTM |
| CN110866235A (en) * | 2019-10-21 | 2020-03-06 | 上海交通大学 | Identity recognition method and device for simultaneously capturing human pulse and vein images |
| CN111462379A (en) * | 2020-03-17 | 2020-07-28 | 广东网深锐识科技有限公司 | Access control management method, system and medium containing palm vein and face recognition |
-
2020
- 2020-11-04 CN CN202011215734.3A patent/CN113420582B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103984924A (en) * | 2014-05-08 | 2014-08-13 | 山东大学 | Reflection type finger vein recognition bioassay device and method |
| CN104778445A (en) * | 2015-03-17 | 2015-07-15 | 山东大学 | Living body recognizing device and method based on three-dimensional characteristics of finger venas |
| WO2018043796A1 (en) * | 2016-09-01 | 2018-03-08 | 상명대학교서울산학협력단 | Device and method for detecting forged bioinformation |
| WO2018061786A1 (en) * | 2016-09-27 | 2018-04-05 | 株式会社日立製作所 | Living body authentication device |
| CN109255305A (en) * | 2018-08-18 | 2019-01-22 | 深圳长元知识产权服务有限公司 | A kind of biometric devices and the shared device using the device |
| CN110321871A (en) * | 2019-07-11 | 2019-10-11 | 电子科技大学成都学院 | A kind of palm vein identification system and method based on LSTM |
| CN110321870A (en) * | 2019-07-11 | 2019-10-11 | 西北民族大学 | A kind of vena metacarpea recognition methods based on LSTM |
| CN110866235A (en) * | 2019-10-21 | 2020-03-06 | 上海交通大学 | Identity recognition method and device for simultaneously capturing human pulse and vein images |
| CN111462379A (en) * | 2020-03-17 | 2020-07-28 | 广东网深锐识科技有限公司 | Access control management method, system and medium containing palm vein and face recognition |
Non-Patent Citations (3)
| Title |
|---|
| JEN-CHUN LEE: "A novel biometric system based on palm vein image", PATTERN RECOGNITION LETTERS, vol. 33, no. 12, XP028503876, DOI: 10.1016/j.patrec.2012.04.007 * |
| 刘霞: "基于深度学习的手指静脉特征提取与防伪检测算法研究", CNKI优秀硕士学位论文全文库, no. 2020 * |
| 郑哲章;罗海波;张建华;: "基于指静脉生物特性的身份识别系统研究与设计", 闽江学院学报, no. 02 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115022605A (en) * | 2022-05-17 | 2022-09-06 | 深圳市当智科技有限公司 | Projection method, projector, and computer-readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113420582B (en) | 2023-09-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109165593B (en) | Feature extraction and matching and template update for biometric authentication | |
| US7327860B2 (en) | Conjunctival scans for personal identification | |
| Qiu et al. | Finger vein presentation attack detection using total variation decomposition | |
| CN111967427A (en) | Fake face video identification method, system and readable storage medium | |
| CN102419819B (en) | Method and system for recognizing human face image | |
| KR101993266B1 (en) | method for designing and learning varying model and method for detecting video forgeries therewith | |
| CN109784130B (en) | Pedestrian re-identification method, device and equipment thereof | |
| CN111695406B (en) | Face recognition anti-spoofing method, system and terminal based on infrared ray | |
| Choudhary et al. | Signature Recognition & Verification System Using Back Propagation Neural Network | |
| Du et al. | A new approach to iris pattern recognition | |
| CN111767879A (en) | Living body detection method | |
| CN115131880A (en) | Multi-scale attention fusion double-supervision human face in-vivo detection method | |
| CN113420582B (en) | Anti-fake detection method and system for palm vein recognition | |
| Kim et al. | Facial manipulation detection based on the color distribution analysis in edge region | |
| CN111798404B (en) | Iris image quality evaluation method and system based on deep neural network | |
| CN111832464A (en) | Living body detection method and device based on near-infrared camera | |
| CN116229528A (en) | Living body palm vein detection method, device, equipment and storage medium | |
| Zaidan et al. | A novel hybrid module of skin detector using grouping histogram technique for Bayesian method and segment adjacent-nested technique for neural network | |
| KR101767051B1 (en) | Method and apparatus for extracting finger vein image based on fuzzy inference | |
| CN115620405A (en) | Face fake video detection method based on rPPG heart rate characteristics | |
| CN113989870A (en) | Living body detection method, door lock system and electronic equipment | |
| CN112906508A (en) | Face living body detection method based on convolutional neural network | |
| CN114694196A (en) | Living body classifier establishing method, human face living body detection method and device | |
| CN112084922A (en) | Abnormal behavior crowd detection method based on gestures and facial expressions | |
| Han et al. | Real-time face detection using AdaBoot algorithm |
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 |