CN110069969B - Authentication fingerprint identification method based on pseudorandom integration - Google Patents
Authentication fingerprint identification method based on pseudorandom integration Download PDFInfo
- Publication number
- CN110069969B CN110069969B CN201810718434.3A CN201810718434A CN110069969B CN 110069969 B CN110069969 B CN 110069969B CN 201810718434 A CN201810718434 A CN 201810718434A CN 110069969 B CN110069969 B CN 110069969B
- Authority
- CN
- China
- Prior art keywords
- signal
- sample
- samples
- signal samples
- pseudo
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G06V40/12—Fingerprints or palmprints
- G06V40/1347—Preprocessing; Feature extraction
-
- 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
- G06V40/12—Fingerprints or palmprints
- G06V40/1365—Matching; Classification
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Image Analysis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses an authentication fingerprint identification method based on pseudorandom integration, which comprises the following steps: s1, a receiving device carries out signal acquisition and storage on a sending device to obtain N signal samples and stores the N signal samples in a sample library; s2, enhancing N signal samples in a sample library into K signal samples by adopting pseudo-random integration; s3, performing feature extraction on the generated signal sample; s4, repeating the operations from S1 to S3 for a plurality of different transmitting devices, and taking the device number as feedback; training the signals processed by each sending device by using a classification algorithm in machine learning to obtain a classifier; and S5, classifying the waveform to be detected by using a classifier, and judging the equipment to which the waveform belongs. The invention uses a data enhancement mode based on pseudo-random integration, and increases the training data amount for machine learning; meanwhile, machine learning is combined, and instability of the radio frequency fingerprint identification technology based on random integration and machine learning in the identification rate is reduced.
Description
Technical Field
The invention relates to the field of access authentication of a physical layer of wireless equipment, in particular to an authentication fingerprint identification method based on pseudorandom integration.
Background
Access authentication of a terminal node is an important and challenging problem for security of the internet of things. Radio frequency fingerprinting is a promising solution to this problem, where end node access authentication is performed by extracting fingerprints by extracting signal variations due to hardware manufacturing imperfections. Machine learning can be well accomplished by training of large amounts of data for radio frequency fingerprint identification.
The machine learning performance of the radio frequency fingerprint is greatly influenced by the training data volume, and a large amount of radio frequency fingerprint data is needed to obtain a better training and identification result. In some cases, the data is very valuable and not easy to acquire. Data enhancement is the most common method to obtain better results using limited data. Data enhancement, which has a large impact on the final classification performance and generalization capability of the system, generates additional pseudo-instances by applying a "label invariant" transformation to train the instances. Current research provides a variety of different data enhancement strategies that work mostly in the field of image recognition. A brief overview of the commonly used image enhancement strategy is as follows:
1. and (3) scale transformation: modifying the size of the image or scaling the image to a certain scale.
2. And (3) position conversion: flipping, rotating and translating, i.e. flipping the image in either the horizontal or vertical direction. The image is rotated at an angle. The image is moved to a certain distance.
3. Interception or deletion. A portion of the image is cropped and deleted or retained.
4. The color is dithered. The brightness, saturation or contrast of the image is changed.
5. PCA dithering. The principal component of the image is extracted by principal component analysis, and then a gaussian perturbation of (0,0.1) is added.
6. Noise. Pixel points are obtained randomly and then set to high brightness and low gray levels to simulate additive noise.
Strategies 1 to 4 are suitable for image enhancement, which cannot be used directly for one-dimensional sequences, such as sound or signals. Strategies 5 and 6 may be used for signal classification, but they add extra noise to the signal samples with high signal-to-noise ratio, which reduces the final classification accuracy. Existing research shows that when the signal-to-noise ratio is reduced, the accuracy of radio frequency fingerprint identification is rapidly reduced. Therefore, the current mainstream data enhancement method is not completely feasible in main signal processing and cannot be directly used for radio frequency fingerprint identification.
The Salamon and Bello study in 2017 provided a series of data enhancement strategies for sound classification that were closely related to the signal, as follows: and (3) prolonging the time: keeping the frequency unchanged, and increasing or decreasing the speed of the sample; frequency change: keeping the speed unchanged, and increasing or decreasing the frequency of the sample; dynamic range compression: some small signals are amplified as required while large signals remain unchanged. These strategies can indeed be used for signal processing. However, the approximation of the signal in radio frequency fingerprinting is very high. The difference between a signal of the same type and its transformed signal will be greater than the difference between the signal and another signal. One key concept of data enhancement is that the distortion applied to the data does not change the meaning of the tag, and therefore these strategies are not applicable to radio frequency fingerprinting either.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides an authentication fingerprint identification method based on pseudo-random integration, which increases the training data amount for machine learning by using a data enhancement mode based on the pseudo-random integration; meanwhile, machine learning is combined, and instability of the radio frequency fingerprint identification technology based on random integration and machine learning in the identification rate is reduced.
The purpose of the invention is realized by the following technical scheme: an authentication fingerprint identification method based on pseudo-random integration comprises the following steps:
s1, a receiving device carries out signal acquisition and storage on a sending device to obtain N signal samples and stores the N signal samples in a sample library;
s2, enhancing N signal samples in a sample library into K signal samples by adopting pseudo-random integration;
s3, performing feature extraction on the generated signal sample;
s4, for a plurality of different transmitting devices, repeating the operations from S1 to S3, and taking the device number as feedback; training the signals processed by each sending device by using a classification algorithm in a machine learning algorithm to obtain a classifier;
and S5, classifying the waveform to be detected by using the classifier obtained by training, and judging the equipment to which the waveform belongs.
Specifically, the step S1 includes:
s101, detecting the starting point position of a received starting transient signal by receiving equipment;
s102, collecting M starting transient signal sample points of a sending device from a starting point position as a signal sample;
s103, numbering starting transient signal sample points in the signal samples, and defining a starting transient signal sample point amplitude function f by using the corresponding amplitude of each starting transient signal sample pointi：
fi＝{amp:(1,2,……,M)}
＝{amp1,amp2,……,ampM}；
S104, collecting N signal samples according to the steps S101-S103 and storing the N signal samples in a sample library.
Specifically, the step S2 performs coherent or non-coherent accumulation on multiple power-on transient signal sample point amplitude functions of the same transmitting device through pseudo-random selection, and generates a large amount of training data for machine learning by using limited signal samples, including:
s201, randomly selecting N signal samples from a sample library, wherein N is less than N;
s202. starting from the second selection of signal samples, it is detected whether the selected n samples are identical to the previous selection: if yes, jumping to step S204; if not, go to step S203;
s203, the n samples are subjected to coherent accumulation or non-coherent accumulation and stored in a database:
wherein f isrIs a sample of the original signal in the library, faTo the generated signal samples;
s204, removing the signal samples participating in integration from a sample library;
s205, repeatedly executing the steps S201 to S204 until the number of samples in the signal library is less than n;
s206, initializing a signal library, and resetting the signal library to be in an original state;
s207, repeatedly executing S201 to S206 until the number of signal samples in the database reaches K;
and S208, outputting the signal samples in the database.
Specifically, the step S3 includes:
s301, for each signal sample in the database, firstly carrying out folding processing, namely taking an absolute value of an amplitude value:
favi＝|fi|；
s302, normalization processing is carried out on the folded signal sample:
in the formula, ampiRepresents the ith sample point amplitude in the signal sample, i 1,2maxRepresenting the maximum sample point amplitude, amp, in the signal sampleminRepresenting the smallest sample point amplitude in the signal sample;
s303, performing feature extraction on the signal sample after normalization processing;
s304, repeating the steps S301 to S303 until the K signal samples in the database are processed, and obtaining the characteristic data of the K signal samples.
Preferably, in step S1, the subsequent operation speed may be increased by vectorization, that is, the collected signal samples are stored in a matrix form:
preferably, in step S4, to increase the operation rate, the feedback value may also be vectorized, that is:
wherein, yiIs a sample fiAnd Y is a feedback storage matrix.
Preferably, in step S4, for the convenience of machine learning, the feature data and the feedback value may be stored together, that is:
wherein, train is a machine learning training matrix.
Preferably, the classification algorithm in the machine learning algorithm includes a k-nearest neighbor algorithm, a naive bayes algorithm, an SVM algorithm and a decision tree algorithm.
Preferably, in step S101, the method for detecting the position of the start point includes absolute amplitude value detection and slope detection.
Preferably, the feature extraction method in step S3 includes amplitude-based feature extraction, amplitude and phase combination-based feature extraction, frequency-based feature extraction, phase shift-based feature extraction, wavelet transform-based feature extraction, and multi-resolution analysis-based feature extraction; wherein the selection of the mother wavelet in the multi-resolution analysis comprises
haar, dB2, bior, and morl.
The invention has the beneficial effects that: (1) the invention utilizes a random integration method to enhance the data of the radio frequency fingerprint, and generates a large number of signal samples by utilizing limited signal samples, thereby improving the available training data for machine learning, reducing overfitting and improving the identification accuracy to a certain extent. (2) The invention uses the pseudo-random selection method, reduces the randomness of selecting the radio frequency fingerprint sample and reduces the instability of the identification accuracy. (3) The invention carries out data acquisition and feature extraction on a plurality of transmitting devices, and takes the device number as feedback to realize the training of the classifier and the judgment of the waveform to be detected, so that each device can keep a higher identification accuracy rate, and the reliability of the identification result is improved.
Drawings
FIG. 1 is a flow chart of a method of the present invention;
FIG. 2 is a flow chart of data enhancement based on pseudo-random integration;
FIG. 3 is a graph comparing the recognition rate of pseudo-random integration, random integration and signals without data enhancement at various signal-to-noise ratios in the present invention;
fig. 4 is a graph comparing the recognition rate of signals to individual devices without data enhancement, pseudo-random integration, and random integration in accordance with the present invention.
Detailed Description
The technical solutions of the present invention are further described in detail below with reference to the accompanying drawings, but the scope of the present invention is not limited to the following.
As shown in fig. 1, an authenticated fingerprint identification method based on pseudo-random integration includes the following steps:
s1, a receiving device carries out signal acquisition and storage on a sending device to obtain N signal samples and stores the N signal samples in a sample library;
s2, enhancing N signal samples in a sample library into K signal samples by adopting pseudo-random integration;
s3, performing feature extraction on the generated signal sample;
s4, for a plurality of different transmitting devices, repeating the operations from S1 to S3, and taking the device number as feedback; training the signals processed by each sending device by using a classification algorithm in a machine learning algorithm to obtain a classifier;
and S5, classifying the waveform to be detected by using the classifier obtained by training, and judging the equipment to which the waveform belongs.
Specifically, the step S1 includes:
s101, detecting the starting point position of a received starting transient signal by receiving equipment;
s102, collecting M starting transient signal sample points of a sending device from a starting point position as a signal sample;
for example, when M is 800, 800 start-up transient signal sample points may be collected from the starting point position as one signal sample; in some other embodiments, starting from 100 positions before the starting point, 100 starting transient signal sample points before the starting point and 700 starting transient signal sample points after the starting point are collected as one signal sample.
S103, numbering starting transient signal sample points in the signal samples, and defining a starting transient signal sample point amplitude function f by using the corresponding amplitude of each starting transient signal sample pointi：
fi＝{amp:(1,2,……,M)}
＝{amp1,amp2,……,ampM}；
S104, collecting N signal samples according to the steps S101-S103 and storing the N signal samples in a sample library.
As shown in fig. 2, the step S2 performs coherent or non-coherent accumulation on multiple power-on transient signal sample point amplitude functions of the same transmitting device through pseudo-random selection, and generates a large amount of training data for machine learning by using limited signal samples, including:
s201, randomly selecting N signal samples from a sample library, wherein N is less than N;
s202. starting from the second selection of signal samples, it is detected whether the selected n samples are identical to the previous selection: if yes, jumping to step S204; if not, go to step S203;
s203, the n samples are subjected to coherent accumulation or non-coherent accumulation and stored in a database:
wherein f isrIs a sample of the original signal in the library, faTo the generated signal samples;
s204, removing the signal samples participating in integration from a sample library;
s205, repeatedly executing the steps S201 to S204 until the number of samples in the signal library is less than n;
s206, initializing a signal library, and resetting the signal library to be in an original state;
s207, repeatedly executing S201 to S206 until the number of signal samples in the database reaches K;
and S208, outputting the signal samples in the database.
Specifically, the step S3 includes:
s301, for each signal sample in the database, firstly carrying out folding processing, namely taking an absolute value of an amplitude value:
favi＝|fi|；
s302, normalization processing is carried out on the folded signal sample:
in the formula, ampiRepresents the ith sample point amplitude in the signal sample, i 1,2maxRepresenting the maximum sample point amplitude, amp, in the signal sampleminRepresenting the smallest sample point amplitude in the signal sample;
s303, performing feature extraction on the signal sample after normalization processing;
s304, repeating the steps S301 to S303 until the K signal samples in the database are processed, and obtaining the characteristic data of the K signal samples.
In step S101, the method for detecting the starting point position includes absolute amplitude value detection and slope detection; in the embodiments of the present application, absolute amplitude value detection is employed: the threshold is set to 0.003, and when the absolute value of the signal amplitude is greater than 0.003, 3000 sample points (M3000) 800 before and 2200 after the point are sampled, respectively. In this embodiment, the subsequent operation rate may be increased by vectorization, that is, the collected signal samples are uniformly stored in a matrix form:
the feature extraction method in step S3 includes amplitude-based feature extraction, amplitude and phase combination-based feature extraction, frequency-based feature extraction, phase shift-based feature extraction, wavelet transform-based feature extraction, and multi-resolution analysis-based feature extraction; wherein the selection of mother wavelets in the multi-resolution analysis includes haar, dB2, bior, and morl.
In the embodiment of the patent, two-level multi-resolution analysis is adopted, a dB2 waveform function is taken as a mother wavelet, and two discrete wavelet transforms are sequentially carried out on a signal waveform:
fa1i＝DWT(favi,dB2)
fa2i＝DWT(fa1i,dB2)
in particular, the amount of the solvent to be used,
in the embodiment of the present application, to increase the operation rate, the feedback value of step S4 may also be vectorized, that is:
wherein, yiIs a sample fiCorresponding feedback values, Y is a feedback storage matrix; further, for the convenience of machine learning, the feature data and the feedback value can be correspondingly and uniformly stored, namely:
wherein, train is a machine learning training matrix.
The classification algorithm in the machine learning algorithm comprises a k-nearest neighbor algorithm, a naive Bayes algorithm, an SVM algorithm and a decision tree algorithm.
In the embodiment of the present patent, step S4 uses a gaussian kernel SVM to generate a classifier, specifically: first, inputting a training matrix with feedback values
Second, first l of each row in train matrixtrainEach element is a feature, the last element in each row is a feedback value, and a Gaussian kernel SVM is used for training, specifically:
s.t.,αi≥0,i＝1,...,n
wherein the content of the first and second substances,
the method is a Gaussian kernel function, and the SVM is matched with training data by adjusting the size of a Gaussian kernel scale sigma, so that under-fitting or over-fitting is prevented.
To verify the difference between the pseudo-random integration and the traditional random integration, four groups of samples are generated for a radio frequency fingerprint identification experiment:
(a) acquiring 20 samples (N-20) by each transmitting device, selecting 3 samples for integration (N-3) each time, and generating 500 samples (K-500) by using random integration;
(b) acquiring 20 samples (N is 20) by each transmitting device, selecting 3 samples for integration (N is 3) each time, and generating 500 samples (K is 500) by using pseudorandom integration;
(c) each sending device collected 20 samples for comparison. In order to ensure the reliability of the experiment, the comparison sample adopts ordered coherent accumulation (n is 3). This is because coherent accumulation has a noise reduction function, which will improve the signal-to-noise ratio to the same extent;
(d) each transmitting device collects 100 samples and generates 100 samples for comparison using ordered coherent accumulation (n-3).
The experimental results are shown in fig. 3 and 4, specifically:
1) under low signal-to-noise ratio, the recognition result of random integration (a) is better than the ordered coherent accumulation (c) of 20 samples, and is similar to the ordered coherent accumulation (d) of 100 samples;
2) at high snr, the classification accuracy of random integration (a) becomes unstable and accuracy of integration order (c) (d) cannot be achieved;
3) although the pseudo-random integration (b) appears jittering in the classification accuracy curve, the effect is better. Under the condition of high signal-to-noise ratio, the classification precision is stabilized to be more than 99 percent;
4) for a single device, the recognition rate of the pseudo-random integration (b) for 10 devices reaches more than 98%, and the stability is better than that of the other 3 methods (a), (c) and (d).
In summary, the signal of the sending device of the present invention is collected, and a large amount of training data is generated by using a small amount of collected signal samples. Training by using the generated training data based on a classification algorithm in machine learning to obtain a classifier, and testing the performance of the classifier by using a test set. The method is suitable for radio frequency fingerprint identification scenes with insufficient training data, and has the advantages of high identification accuracy, stability and reliability.
The foregoing is a preferred embodiment of the present invention, it is to be understood that the invention is not limited to the form disclosed herein, but is not to be construed as excluding other embodiments, and is capable of other combinations, modifications, and environments and is capable of changes within the scope of the inventive concept as expressed herein, commensurate with the above teachings, or the skill or knowledge of the relevant art. And that modifications and variations may be effected by those skilled in the art without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (7)
1. An authentication fingerprint identification method based on pseudo-random integration is characterized in that: the method comprises the following steps:
s1, a receiving device carries out signal acquisition and storage on a sending device to obtain N signal samples and stores the N signal samples in a sample library;
s2, enhancing N signal samples in a sample library into K signal samples by adopting pseudo-random integration;
in step S2, performing coherent or non-coherent accumulation on multiple start-up transient signal sample point amplitude functions of the same transmission device through pseudo-random selection, and generating a large amount of training data for machine learning by using limited signal samples, including:
s201, randomly selecting N signal samples from a sample library, wherein N is less than N;
s202. starting from the second selection of signal samples, it is detected whether the selected n samples are identical to the previous selection:
if yes, jumping to step S204;
if not, go to step S203;
s203, the n samples are subjected to coherent accumulation or non-coherent accumulation and stored in a database:
wherein f isrIs a sample of the original signal in the library, faTo the generated signal samples;
s204, removing the signal samples participating in integration from a sample library;
s205, repeatedly executing the steps S201 to S204 until the number of samples in the signal library is less than n;
s206, initializing a signal library, and resetting the signal library to be in an original state;
s207, repeatedly executing S201 to S206 until the number of signal samples in the database reaches K;
s208, outputting a signal sample in the database;
s3, performing feature extraction on the generated signal sample;
s4, for a plurality of different transmitting devices, repeating the operations from S1 to S3, and taking the device number as feedback; training the signals processed by each sending device by using a classification algorithm in a machine learning algorithm to obtain a classifier;
and S5, classifying the waveform to be detected by using the classifier obtained by training, and judging the equipment to which the waveform belongs.
2. The authentication fingerprint identification method based on the pseudo-random integration according to claim 1, wherein: the step S1 includes:
s101, detecting the starting point position of a received starting transient signal by receiving equipment;
s102, collecting M starting transient signal sample points of a sending device from a starting point position as a signal sample;
s103, numbering starting transient signal sample points in the signal samples, and defining a starting transient signal sample point amplitude function f by using the corresponding amplitude of each starting transient signal sample pointi，：
fi＝{amp:(1,2,……,M)}
＝{amp1,amp2,……,ampM}；
S104, collecting N signal samples according to the steps S101-S103 and storing the N signal samples in a sample library.
3. The authentication fingerprint identification method based on the pseudo-random integration according to claim 2, wherein: the step S3 includes:
s301, for each signal sample in the database, firstly carrying out folding processing, namely taking an absolute value of an amplitude value:
favi＝|fi|；
s302, normalization processing is carried out on the folded signal sample:
in the formula, ampiRepresents the ith sample point amplitude in the signal sample, i 1,2maxRepresenting the maximum sample point amplitude, amp, in the signal sampleminRepresenting the smallest sample point amplitude in the signal sample;
s303, performing feature extraction on the signal sample after normalization processing;
s304, repeating the steps S301 to S303 until the K signal samples in the database are processed, and obtaining the characteristic data of the K signal samples.
4. The authentication fingerprint identification method based on the pseudo-random integration according to claim 1, wherein: the classification algorithm in the machine learning algorithm comprises a k-nearest neighbor algorithm, a naive Bayes algorithm, an SVM algorithm and a decision tree algorithm.
5. The authentication fingerprint identification method based on the pseudo-random integration according to claim 2, wherein: in step S101, the method for detecting the start point position includes absolute amplitude value detection and slope detection.
6. The authentication fingerprint identification method based on the pseudo-random integration according to claim 1, wherein: the feature extraction method in step S3 includes amplitude-based feature extraction, amplitude and phase combination-based feature extraction, frequency-based feature extraction, phase shift-based feature extraction, wavelet transform-based feature extraction, and multi-resolution analysis-based feature extraction.
7. The authentication fingerprint identification method based on the pseudo-random integration according to claim 6, wherein: the selection of the mother wavelet in the multiresolution analysis includes haar, dB2, bior, and morl.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810718434.3A CN110069969B (en) | 2018-07-03 | 2018-07-03 | Authentication fingerprint identification method based on pseudorandom integration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810718434.3A CN110069969B (en) | 2018-07-03 | 2018-07-03 | Authentication fingerprint identification method based on pseudorandom integration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110069969A CN110069969A (en) | 2019-07-30 |
| CN110069969B true CN110069969B (en) | 2021-10-19 |
Family
ID=67365740
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810718434.3A Active CN110069969B (en) | 2018-07-03 | 2018-07-03 | Authentication fingerprint identification method based on pseudorandom integration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110069969B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111027623A (en) * | 2019-12-10 | 2020-04-17 | 深圳供电局有限公司 | Data-enhanced intelligent terminal security level classification method and system |
| CN111586689B (en) * | 2020-04-28 | 2023-04-18 | 福建师范大学 | Multi-attribute lightweight physical layer authentication method based on principal component analysis algorithm |
| CN113347637B (en) * | 2021-04-19 | 2023-07-18 | 厦门大学 | Radio Frequency (RF) fingerprint identification method and device based on embedded wireless equipment |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9443298B2 (en) * | 2012-03-02 | 2016-09-13 | Authentect, Inc. | Digital fingerprinting object authentication and anti-counterfeiting system |
| US9235748B2 (en) * | 2013-11-14 | 2016-01-12 | Wacom Co., Ltd. | Dynamic handwriting verification and handwriting-based user authentication |
| US10289822B2 (en) * | 2016-07-22 | 2019-05-14 | Nec Corporation | Liveness detection for antispoof face recognition |
| CN108230233A (en) * | 2017-05-16 | 2018-06-29 | 北京市商汤科技开发有限公司 | Data enhancing, treating method and apparatus, electronic equipment and computer storage media |
| CN107392123B (en) * | 2017-07-10 | 2021-02-05 | 电子科技大学 | Radio frequency fingerprint feature extraction and identification method based on coherent accumulation noise elimination |
-
2018
- 2018-07-03 CN CN201810718434.3A patent/CN110069969B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110069969A (en) | 2019-07-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Lopes et al. | Data-free knowledge distillation for deep neural networks | |
| Chen et al. | Marginalized denoising auto-encoders for nonlinear representations | |
| CN110069969B (en) | Authentication fingerprint identification method based on pseudorandom integration | |
| Shekhar et al. | Analysis sparse coding models for image-based classification | |
| Shen et al. | AdaBoost Gabor feature selection for classification | |
| Chandrasekhar et al. | Comparison of local feature descriptors for mobile visual search | |
| Qin et al. | Perceptual image hashing with selective sampling for salient structure features | |
| Schuch et al. | Survey on the impact of fingerprint image enhancement | |
| US20120321125A1 (en) | Image processing method and apparatus | |
| CN106663195B (en) | Improved method, apparatus and system for fingerprint matching and camera identification | |
| Kasban et al. | An efficient approach for forgery detection in digital images using Hilbert–Huang transform | |
| Caldelli et al. | PRNU-based image classification of origin social network with CNN | |
| Juefei-Xu et al. | Fastfood dictionary learning for periocular-based full face hallucination | |
| Sutthiwan et al. | New developments in color image tampering detection | |
| Cozzolino et al. | Towards universal gan image detection | |
| Zois et al. | Parsimonious coding and verification of offline handwritten signatures | |
| Manikantan et al. | DWT-based illumination normalization and feature extraction for enhanced face recognition | |
| CN111259792A (en) | Face living body detection method based on DWT-LBP-DCT characteristics | |
| Guan et al. | Detecting steganography in JPEG images recompressed with the same quantization matrix | |
| Zuo et al. | A novel slant transform-based image feature extract algorithm | |
| Thanki et al. | Compressive sensing based multiple watermarking technique for biometric template protection | |
| CN109598216B (en) | Convolution-based radio frequency fingerprint feature extraction method | |
| Ustubioglu et al. | A novel keypoint based forgery detection method based on local phase quantization and SIFT | |
| Zhang et al. | Global contrast enhancement detection via deep multi-path network | |
| Setyawan et al. | Spatio-temporal digital video hashing using edge orientation histogram and discrete cosine transform |
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 |