CN113673315A - Face recognition method, device and equipment based on RFID and storage medium - Google Patents
Face recognition method, device and equipment based on RFID and storage medium Download PDFInfo
- Publication number
- CN113673315A CN113673315A CN202110776563.XA CN202110776563A CN113673315A CN 113673315 A CN113673315 A CN 113673315A CN 202110776563 A CN202110776563 A CN 202110776563A CN 113673315 A CN113673315 A CN 113673315A
- Authority
- CN
- China
- Prior art keywords
- data
- data set
- user
- face
- preset
- 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
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K7/00—Methods or arrangements for sensing record carriers, e.g. for reading patterns
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Collating Specific Patterns (AREA)
- Image Analysis (AREA)
Abstract
The invention is suitable for the technical field of face recognition, and provides a face recognition method, a face recognition device, face recognition equipment and a storage medium based on RFID, wherein the method comprises the following steps: the method comprises the steps of collecting facial data of a user in a preset range in real time through radio frequency identification equipment, dividing the collected facial data in real time to obtain first data, preprocessing the first data to obtain second data, and identifying the user through a pre-trained neural network according to the second data to identify the identity of the user, so that the influence degree of the environment on face identification is reduced, and the accuracy and the safety of the face identification are improved.
Description
Technical Field
The invention belongs to the technical field of face recognition, and particularly relates to a face recognition method, a face recognition device, face recognition equipment and a storage medium based on RFID.
Background
With the development of science and technology and the progress of society, the identification technology for identity authentication by using human biological characteristics has been rapidly developed in recent decades, and compared with other current biological characteristic identification technologies, the face identification technology has the characteristics of being direct, friendly and convenient, and in the face identification technology, face identification equipment can actively acquire face image information of a user without being in direct contact with the user, does not disturb the user, and is easily accepted by the user, so that the face identification technology has become one of hot directions of research in recent years, and is widely applied to the fields of public security, digital payment, equipment unlocking and the like.
Currently, the mainstream face recognition technology is mainly based on a camera and an image recognition algorithm, that is, the camera arranged in a suitable area is used for collecting face data of a passing person, and the data is compared with face data prestored in a system through the image recognition algorithm, so that identity verification and recognition judgment are realized, however, the face recognition technology based on the mode has the following problems:
firstly, the camera is sensitive to light rays, so that the camera without an infrared function cannot work at night, and the real-time performance of face recognition is reduced;
secondly, with the popularization of the application of the face recognition technology in various fields, face image information is easy to obtain, and the hidden danger of privacy disclosure of a user exists;
and thirdly, the face recognition method is easy to be subjected to 2D and 3D camouflage deception, such as 2D face photos and 3D face printing models, so that the accuracy of face recognition is reduced.
Therefore, a new face recognition scheme is needed to solve the above problems.
Disclosure of Invention
The invention aims to provide a face recognition method, a face recognition device, face recognition equipment and a storage medium based on RFID, and aims to solve the problem of low safety and accuracy of face recognition caused by the prior art.
In one aspect, the invention provides a face recognition method based on RFID, which comprises the following steps:
the method comprises the steps that face data of a user in a preset range are collected in real time through radio frequency identification equipment;
the collected face data is segmented in real time to obtain first data;
preprocessing the first data to obtain second data;
and identifying the user through a pre-trained neural network according to the second data so as to identify the identity of the user.
Preferably, the step of segmenting the acquired face data in real time includes:
performing variance calculation on the label signal intensity values in the face data in a sliding window with a preset size in real time to respectively obtain signal intensity variance values corresponding to each label in the radio frequency identification equipment;
and when the maximum value in the signal intensity variance values corresponding to each label is larger than a preset variance threshold value, acquiring the face data of the user in a preset time length as the first data.
Preferably, the step of preprocessing the first data includes:
adjusting phase values in the first data using an Unwarp algorithm;
filtering the adjusted first data by adopting a Kalman filter;
and carrying out normalization processing on the filtered first data to obtain second data.
Preferably, before the step of acquiring the face data of the user within the preset range in real time through the radio frequency identification device, the method further includes:
real-time segmentation is carried out on a training data set acquired through the radio frequency identification device to obtain a first training data set;
performing data enhancement processing on the first training data set to obtain a second training data set;
preprocessing the second training data set to obtain a third training data set;
training the neural network with the third training data set.
Preferably, the step of performing data enhancement processing on the first training data set includes:
adding Gaussian noise to each training data in the first training data set to obtain a first extended data set;
carrying out weighted average calculation on the preset amount of data of the same user in the first expanded data set to obtain a second expanded data set;
performing scaling processing on the data in the first extended data set and the second extended data set within a preset scaling range to obtain a third extended data set;
turning over data of a preset part in the third expansion data set to obtain a fourth expansion data set;
and merging the first training data set and the fourth expansion data set to obtain the second training data set.
Preferably, the neural network is a hybrid neural network of a convolutional neural network-a long-short term memory network.
In another aspect, the present invention provides a face recognition device based on RFID, the device comprising:
the data acquisition unit is used for acquiring the facial data of the user in a preset range in real time through the radio frequency identification device;
the data segmentation unit is used for segmenting the acquired face data in real time to obtain first data;
the data preprocessing unit is used for preprocessing the first data to obtain second data; and
and the identity recognition unit is used for recognizing the user through a pre-trained neural network according to the second data so as to recognize the identity of the user.
Preferably, the data division unit includes:
the window variance calculation unit is used for carrying out variance calculation on the label signal intensity values in the face data in a sliding window with a preset size in real time to respectively obtain signal intensity variance values corresponding to each label in the radio frequency identification device; and
the first data acquisition unit is used for acquiring the face data of the user in a preset time length as the first data when the maximum value in the signal intensity variance value corresponding to each label is larger than a preset variance threshold value.
In another aspect, the present invention further provides a face recognition device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the RFID-based face recognition method when executing the computer program.
In another aspect, the present invention further provides a computer-readable storage medium, which stores a computer program, and the computer program, when executed by a processor, implements the steps of the RFID-based face recognition method.
The method and the device collect the facial data of the user in the preset range in real time through the radio frequency identification device, divide the collected facial data in real time to obtain the first data, preprocess the first data to obtain the second data, and identify the user through the pre-trained neural network according to the second data to identify the identity of the user, so that the influence degree of the environment on the face identification is reduced, and the accuracy and the safety of the face identification are improved.
Drawings
Fig. 1 is a flowchart illustrating an implementation of a face recognition method based on RFID according to an embodiment of the present invention;
fig. 2 is a flowchart of an implementation of a face recognition method based on RFID according to a second embodiment of the present invention;
fig. 3 is a schematic structural diagram of a face recognition device based on RFID according to a third embodiment of the present invention;
fig. 4 is a schematic structural diagram of a preferred structure of a face recognition device based on RFID according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a face recognition device based on RFID according to a fourth embodiment of the present invention;
fig. 6 is a schematic structural diagram of a preferred structure of a face recognition device based on RFID according to a fourth embodiment of the present invention; and
fig. 7 is a schematic structural diagram of a face recognition device according to a fifth embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
The following detailed description of specific implementations of the present invention is provided in conjunction with specific embodiments:
the first embodiment is as follows:
fig. 1 shows an implementation flow of a face recognition method based on RFID according to a first embodiment of the present invention, and for convenience of description, only the relevant parts related to the first embodiment of the present invention are shown, which are detailed as follows:
in step S101, facial data of a user within a preset range is collected in real time through a radio frequency identification device.
The embodiment of the invention is suitable for face recognition equipment based on Radio Frequency Identification (RFID) equipment (or system), and the RFID equipment consists of a Tag matrix (Tag matrix), a Reader (Reader) and an Antenna (Antenna). In the embodiment of the invention, the reader of the RFID equipment continuously emits electromagnetic waves to activate the tag, when the face of a user is positioned in a preset range, the antenna emits a radio frequency Signal to the face of the user, then the radio frequency Signal reflected by the face of the user is collected through the tag matrix, the radio frequency Signal comprises 2 Signal data of tag Signal Strength (RSS) and Phase (Phase), then the radio frequency Signal which is backscattered after the radio frequency Signal reflected by the face of the user is Received by the tag matrix is sent to the reader through the antenna, and finally the radio frequency Signal is analyzed through the reader to obtain the face data of the user.
Preferably, the tag matrix of the RFID device is formed by 5 rows and 7 columns of passive RFID tags, and the tag matrix is arranged perpendicular to the antenna, so that the influence of the face on the radio frequency signal is improved, and the accuracy of face recognition is further improved.
Preferably, the RFID device sampling time is set to 2 seconds and the data sampling frequency is set to 20 data points per tag per second, thereby increasing the availability of the collected data affected by the human face.
Preferably, the preset range is set to be within 20 cm of the face of the user to the tag matrix, thereby improving the accuracy of face recognition.
In step S102, the acquired face data is segmented in real time to obtain first data.
In the embodiment of the present invention, when the collected face data is segmented in real time, it is preferable that the segmentation of the face data in real time is realized by:
(1) and carrying out variance calculation on the label signal intensity values in the internal data of the sliding window with the preset size in real time to respectively obtain the signal intensity variance value corresponding to each label in the radio frequency identification device.
In the embodiment of the invention, the sliding window with the preset size is used for carrying out variance calculation on the RSS time sequence data in the face data in real time, and because each label in the radio frequency identification device collects the RSS time sequence data, the sliding window corresponding to each label can simultaneously carry out variance calculation on the RSS time sequence data corresponding to the label to obtain the signal intensity variance value corresponding to each label, for example, if the RFID device has 35 labels, 35 signal intensity variances are obtained.
Preferably, the size of the sliding window is set to 5, thereby reducing the false negative rate of the sliding window sampling.
(2) When the maximum value in the signal intensity variance values corresponding to each label is larger than a preset variance threshold value, collecting face data of a user in a preset time length as first data.
In the embodiment of the invention, the signal intensity variance values corresponding to the labels are compared, the maximum signal intensity variance value is taken as the integral variance value, the variance value of the RSS time sequence data acquired in real time at each moment is calculated in real time by using the sliding window along with time, when the integral variance value is larger than a preset variance threshold value, that is, a user is considered to be identified, the face data of the user in a preset time length is acquired as the first data, and the first data is the feature data influenced by the face.
Preferably, the preset variance threshold is 2, so as to improve the speed and the real-time performance of face recognition.
Preferably, the preset time is 2 seconds, so that the speed and the real-time performance of face recognition are improved.
The real-time segmentation of the face data is realized through the steps (1) to (2), so that whether a human face exists or not is detected through data fluctuation of RSS (received signal strength), whether a user identifies or not is determined, and the effectiveness of the collected first data is improved.
In step S103, the first data is preprocessed to obtain second data.
In the embodiment of the present invention, preferably, the preprocessing of the first data is implemented by:
(1) the phase values in the first data are adjusted using the Unwarp algorithm.
In the embodiment of the present invention, the number count1 of Phase occurring between (5,2 pi) and the number count2 of Phase occurring between (0,1.5) in the first data are calculated, when count1> -count 2, the Phase data between (0,1.5) is adjusted to 2 pi + Phase, and when count1< — count2, the Phase data between (5,2 pi) is adjusted to Phase-2 pi, so as to avoid oscillation of the Phase data collected by the tag between two values of 0 and 2 pi.
(2) And filtering the adjusted first data by adopting a Kalman filter.
In the embodiment of the invention, the first data after being adjusted is filtered by adopting the Kalman filter, so that the first data is smoothed, and the data fluctuation caused by environmental noise is reduced.
(3) And carrying out normalization processing on the filtered first data to obtain second data.
In the embodiment of the present invention, all phases in the filtered first data are divided by 2 pi, and all RSS in the filtered first data is divided by 50, so as to complete the normalization processing on the filtered first data.
Therefore, the preprocessing of the first data is realized through the steps (1) to (3), and the stability and the effectiveness of the data are improved.
In step S104, the user is identified through a pre-trained neural network according to the second data, so as to identify the identity of the user.
In the embodiment of the invention, the second data is input into a pre-trained neural network, the time and space characteristics of the user during face recognition are extracted through the neural network, and finally classification recognition is carried out to complete the recognition of the user identity.
Preferably, the Neural Network is a hybrid Neural Network of a Convolutional Neural Network (CNN) -Long Short Term Memory Network (LSTM), so as to improve the accuracy of classifying the face feature data.
In the embodiment of the invention, the face data of the user in the preset range is acquired in real time through the radio frequency identification device, the acquired face data is segmented in real time to obtain the first data, the first data is preprocessed to obtain the second data, and the user is identified through the pre-trained neural network according to the second data to identify the identity of the user, so that the influence degree of the environment on the face identification is reduced, and the accuracy and the safety of the face identification are improved.
Example two:
fig. 2 shows an implementation flow of a face recognition method based on RFID according to a second embodiment of the present invention, and for convenience of description, only the relevant parts related to the second embodiment of the present invention are shown, which are detailed as follows:
in step S201, a training data set acquired by a radio frequency identification device is segmented in real time to obtain a first training data set.
In the embodiment of the present invention, reference may be made to the description of step S102 in the first embodiment for a specific implementation of real-time segmentation of the training data set, and details are not repeated here.
In step S202, data enhancement processing is performed on the first training data set to obtain a second training data set.
In the embodiment of the present invention, preferably, the data enhancement processing on the first training data set is implemented by:
(1) gaussian noise is added to each training data in the first training data set to obtain a first extended data set.
In an embodiment of the present invention, a randomly generated mean of 0 and variance of σ is added to the RSS in each training data in the first training data setRSSGaussian noise of 0.5, Phase addition in each training data randomly generated mean 0 and variance σphaseA gaussian noise of 0.05, the first extended data set is composed of all noisy training data.
(2) And carrying out weighted average calculation on the preset amount of data of the same user in the first expanded data set to obtain a second expanded data set.
In the embodiment of the invention, a plurality of pieces of data from the same user are randomly taken out from the first expanded data set, and based on a Dynamic Time Warping (DTW) algorithm, a formula is adoptedCarrying out weighted average calculation on a plurality of pieces of data randomly taken from the same user to obtain new data, carrying out the weighted average calculation on the data of each user to obtain a second expanded data set, wherein T is a Phase and RSS time sequence data set collected in the face shaking process of the user, and w isiIs the weight of the ith piece of data, and N is the number of the taken data of the same user.
(3) And performing expansion and contraction processing on the data in the first expansion data set and the second expansion data set within a preset expansion and contraction range to obtain a third expansion data set.
In the embodiment of the present invention, the newly generated data in the first and second extended data sets are randomly elongated or compressed within a scaling range of (-30%, 30%), so as to obtain a third extended data set.
(4) And turning over the data of the preset part in the third expansion data set to obtain a fourth expansion data set.
In the embodiment of the invention, the formula phase is firstly passednew=phaseoriginalAlpha adjusts Phase in the third augmented data set and by the formula RSSnew=RSSoriginalAnd α adjusts the RSS in the third extended data set, and then turns over 20% of the randomly-selected and adjusted data in the third extended data set, for example, the data at the time i is changed into the data at the time N-i, where α is a random coefficient whose value range is (-5%, 5%), and N is the number of pieces of training data collected by each user.
(5) And merging the first training data set and the fourth expansion data set to obtain a second training data set.
In the embodiment of the present invention, the data in the fourth extended data set after the flipping process is merged with the training data in the first training data set to obtain the second training data set.
Therefore, data enhancement is realized through the steps (1) to (5), the phenomenon that the neural network model is over-trained and fitted due to small-batch training data is avoided, meanwhile, the complexity of training data acquisition is reduced, and the high recognition accuracy of the neural network model is kept.
In step S203, the second training data set is preprocessed to obtain a third training data set.
In the embodiment of the present invention, the description of step S103 in the first embodiment may be referred to for a specific implementation of preprocessing the second training data set, and is not repeated herein.
In step S204, the neural network is trained by a third training data set.
In the embodiment of the invention, the training data in the third training data set is input into the neural network to train the neural network, so that the neural network can extract the face features and perform classification and identification to identify the user identity.
In the embodiment of the invention, the training data set acquired by the radio frequency identification device is segmented in real time to obtain the first training data set, the first training data set is subjected to data enhancement processing to obtain the second training data set, the second training data set is preprocessed to obtain the third training data set, and the neural network is trained by the third training data set, so that the neural network model is prevented from being trained and over-fitted due to small-batch training data, the complexity of training data acquisition is reduced, and the accuracy of the neural network model for face identification is improved.
Example three:
fig. 3 shows a structure of a face recognition device based on RFID according to a third embodiment of the present invention, and for convenience of description, only the parts related to the third embodiment of the present invention are shown, which include:
the data acquisition unit 31 is used for acquiring the facial data of the user in a preset range in real time through the radio frequency identification device;
the data segmentation unit 32 is used for segmenting the acquired face data in real time to obtain first data;
the data preprocessing unit 33 is configured to preprocess the first data to obtain second data; and
and the identity recognition unit 34 is used for recognizing the user through the pre-trained neural network according to the second data so as to recognize the identity of the user.
As shown in fig. 4, the data dividing unit 32 of the embodiment of the present invention preferably includes:
the window variance calculating unit 321 is configured to perform variance calculation on the label signal intensity values in the internal data of the sliding window with the preset size in real time to obtain a signal intensity variance value corresponding to each label in the radio frequency identification device; and
the first data obtaining unit 322 is configured to, when a maximum value in the signal strength variance values corresponding to each tag is greater than a preset variance threshold, collect face data of a user in a preset time period as first data.
The data preprocessing unit 33 includes:
a phase adjusting unit 331 for adjusting a phase value in the first data using an Unwarp algorithm;
a data filtering unit 332, configured to filter the adjusted first data by using a kalman filter; and
and the data normalization unit 333 is configured to perform normalization processing on the filtered first data to obtain second data.
In the embodiment of the present invention, each unit of the RFID-based face recognition device may be implemented by a corresponding hardware or software unit, and each unit may be an independent software or hardware unit, or may be integrated into a software or hardware unit, which is not limited herein. Specifically, the implementation of each unit can refer to the description of the first embodiment, and is not repeated herein.
Example four:
fig. 5 shows a structure of a face recognition device based on RFID according to a fourth embodiment of the present invention, and for convenience of description, only the parts related to the embodiment of the present invention are shown, which include:
a training data segmentation unit 51, configured to perform real-time segmentation on a training data set acquired by a radio frequency identification device to obtain a first training data set;
a training data enhancement unit 52, configured to perform data enhancement processing on the first training data set to obtain a second training data set;
a training data processing unit 53, configured to perform preprocessing on the second training data set to obtain a third training data set; and
and a network training unit 54 for training the neural network by the third training data set.
As shown in fig. 6, preferably, the training data enhancing unit 52 of the embodiment of the present invention includes:
a noise adding unit 521, configured to add gaussian noise to each training data in the first training data set to obtain a first extended data set;
a data weighting unit 522, configured to perform weighted average calculation on a preset number of data of the same user in the first augmented data set to obtain a second augmented data set;
a data scaling unit 523, configured to scale data in the first expansion data set and the second expansion data set within a preset scaling range to obtain a third expansion data set;
a data flipping unit 524, configured to flip data of a preset portion in the third extended data set, so as to obtain a fourth extended data set; and
the data merging unit 525 is configured to merge the first training data set and the fourth extended data set to obtain a second training data set.
In the embodiment of the present invention, each unit of the RFID-based face recognition device may be implemented by a corresponding hardware or software unit, and each unit may be an independent software or hardware unit, or may be integrated into a software or hardware unit, which is not limited herein. Specifically, the implementation of each unit can refer to the description of the first and second embodiments, and is not repeated herein.
Example five:
fig. 7 shows a structure of a face recognition apparatus according to a fifth embodiment of the present invention, and for convenience of description, only the portions related to the embodiment of the present invention are shown.
The face recognition apparatus 7 of the embodiment of the present invention includes a processor 70, a memory 71, and a computer program 72 stored in the memory 71 and executable on the processor 70. The processor 70, when executing the computer program 72, implements the steps in the above-described RFID-based face recognition method embodiments, such as steps S101 to S104 shown in fig. 1. Alternatively, the processor 70, when executing the computer program 72, implements the functions of the units in the above-described apparatus embodiments, such as the functions of the units 31 to 34 shown in fig. 3.
In the embodiment of the invention, the face data of the user in the preset range is acquired in real time through the radio frequency identification device, the acquired face data is segmented in real time to obtain the first data, the first data is preprocessed to obtain the second data, and the user is identified through the pre-trained neural network according to the second data to identify the identity of the user, so that the influence degree of the environment on the face identification is reduced, and the accuracy and the safety of the face identification are improved.
The face recognition device of the embodiment of the present invention may be a computing device based on an RFID device, for example, a personal computer or a server, or may be an RFID device with computing capability. The steps implemented when the processor 70 executes the computer program 72 in the face recognition device 7 to implement the RFID-based face recognition method can refer to the description of the foregoing method embodiments, and are not described herein again.
Example six:
in an embodiment of the present invention, a computer-readable storage medium is provided, which stores a computer program, and the computer program, when executed by a processor, implements the steps in the above-mentioned RFID-based face recognition method embodiment, for example, steps S101 to S104 shown in fig. 1. Alternatively, the computer program may be adapted to perform the functions of the units of the above-described device embodiments, such as the functions of the units 31 to 34 shown in fig. 3, when executed by the processor.
In the embodiment of the invention, the face data of the user in the preset range is acquired in real time through the radio frequency identification device, the acquired face data is segmented in real time to obtain the first data, the first data is preprocessed to obtain the second data, and the user is identified through the pre-trained neural network according to the second data to identify the identity of the user, so that the influence degree of the environment on the face identification is reduced, and the accuracy and the safety of the face identification are improved.
The computer readable storage medium of the embodiments of the present invention may include any entity or device capable of carrying computer program code, a recording medium, such as a ROM/RAM, a magnetic disk, an optical disk, a flash memory, or the like.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A face recognition method based on RFID is characterized by comprising the following steps:
the method comprises the steps that face data of a user in a preset range are collected in real time through radio frequency identification equipment;
the collected face data is segmented in real time to obtain first data;
preprocessing the first data to obtain second data;
and identifying the user through a pre-trained neural network according to the second data so as to identify the identity of the user.
2. The method of claim 1, wherein the step of segmenting the acquired face data in real-time comprises:
performing variance calculation on the label signal intensity values in the face data in a sliding window with a preset size in real time to respectively obtain signal intensity variance values corresponding to each label in the radio frequency identification equipment;
and when the maximum value in the signal intensity variance values corresponding to each label is larger than a preset variance threshold value, acquiring the face data of the user in a preset time length as the first data.
3. The method of claim 1, wherein the step of preprocessing the first data comprises:
adjusting phase values in the first data using an Unwarp algorithm;
filtering the adjusted first data by adopting a Kalman filter;
and carrying out normalization processing on the filtered first data to obtain second data.
4. The method of claim 1, wherein prior to the step of acquiring in real time facial data of the user within the predetermined range via the radio frequency identification device, the method further comprises:
real-time segmentation is carried out on a training data set acquired through the radio frequency identification device to obtain a first training data set;
performing data enhancement processing on the first training data set to obtain a second training data set;
preprocessing the second training data set to obtain a third training data set;
training the neural network with the third training data set.
5. The method of claim 4, wherein the step of data enhancing the first training data set comprises:
adding Gaussian noise to each training data in the first training data set to obtain a first extended data set;
carrying out weighted average calculation on the preset amount of data of the same user in the first expanded data set to obtain a second expanded data set;
performing scaling processing on the data in the first extended data set and the second extended data set within a preset scaling range to obtain a third extended data set;
turning over data of a preset part in the third expansion data set to obtain a fourth expansion data set;
and merging the first training data set and the fourth expansion data set to obtain the second training data set.
6. The method of claim 1, in which the neural network is a hybrid neural network of convolutional neural network-long short term memory network.
7. An RFID-based face recognition device, the device comprising:
the data acquisition unit is used for acquiring the facial data of the user in a preset range in real time through the radio frequency identification device;
the data segmentation unit is used for segmenting the acquired face data in real time to obtain first data;
the data preprocessing unit is used for preprocessing the first data to obtain second data; and
and the identity recognition unit is used for recognizing the user through a pre-trained neural network according to the second data so as to recognize the identity of the user.
8. The apparatus of claim 7, wherein the data partitioning unit comprises:
the window variance calculation unit is used for carrying out variance calculation on the label signal intensity values in the face data in a sliding window with a preset size in real time to respectively obtain signal intensity variance values corresponding to each label in the radio frequency identification device; and
the first data acquisition unit is used for acquiring the face data of the user in a preset time length as the first data when the maximum value in the signal intensity variance value corresponding to each label is larger than a preset variance threshold value.
9. A face recognition device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that the processor implements the steps of the method according to any of claims 1 to 6 when executing the computer program.
10. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110776563.XA CN113673315B (en) | 2021-07-09 | 2021-07-09 | Face recognition method, device, equipment and storage medium based on RFID |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110776563.XA CN113673315B (en) | 2021-07-09 | 2021-07-09 | Face recognition method, device, equipment and storage medium based on RFID |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113673315A true CN113673315A (en) | 2021-11-19 |
CN113673315B CN113673315B (en) | 2023-10-17 |
Family
ID=78539038
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110776563.XA Active CN113673315B (en) | 2021-07-09 | 2021-07-09 | Face recognition method, device, equipment and storage medium based on RFID |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113673315B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116933167A (en) * | 2023-08-03 | 2023-10-24 | 天津大学 | Mobile scene liquid sensing method based on RFID |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107977633A (en) * | 2017-12-06 | 2018-05-01 | 平安科技(深圳)有限公司 | Age recognition methods, device and the storage medium of facial image |
CN109409220A (en) * | 2018-09-19 | 2019-03-01 | 平安科技(深圳)有限公司 | Business bootstrap technique, device and storage medium based on recognition of face |
-
2021
- 2021-07-09 CN CN202110776563.XA patent/CN113673315B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107977633A (en) * | 2017-12-06 | 2018-05-01 | 平安科技(深圳)有限公司 | Age recognition methods, device and the storage medium of facial image |
CN109409220A (en) * | 2018-09-19 | 2019-03-01 | 平安科技(深圳)有限公司 | Business bootstrap technique, device and storage medium based on recognition of face |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116933167A (en) * | 2023-08-03 | 2023-10-24 | 天津大学 | Mobile scene liquid sensing method based on RFID |
Also Published As
Publication number | Publication date |
---|---|
CN113673315B (en) | 2023-10-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106934359B (en) | Multi-view gait recognition method and system based on high-order tensor subspace learning | |
Qin et al. | Region growth–based feature extraction method for finger-vein recognition | |
Tome et al. | The 1st competition on counter measures to finger vein spoofing attacks | |
CN102629320B (en) | Ordinal measurement statistical description face recognition method based on feature level | |
Hsu et al. | Personal authentication through dorsal hand vein patterns | |
Lee et al. | Dorsal hand vein recognition based on directional filter bank | |
Trabelsi et al. | Hand vein recognition system with circular difference and statistical directional patterns based on an artificial neural network | |
CN102214299A (en) | Method for positioning facial features based on improved ASM (Active Shape Model) algorithm | |
Chaa et al. | Score-level fusion of two-dimensional and three-dimensional palmprint for personal recognition systems | |
El-Alfy et al. | Automated gait-based gender identification using fuzzy local binary patterns with tuned parameters | |
Srivastava | Personal identification using iris recognition system, a review | |
Nigam et al. | Finger knuckleprint based recognition system using feature tracking | |
CN113673315B (en) | Face recognition method, device, equipment and storage medium based on RFID | |
Jingade et al. | DOG-ADTCP: A new feature descriptor for protection of face identification system | |
Mehboob et al. | Live fingerprint detection using magnitude of perceived spatial stimuli and local phase information | |
Kumar et al. | Rank level integration of face based biometrics | |
Chouhan et al. | Analysis of statistical feature extraction for Iris Recognition System using Laplacian of Gaussian filter | |
Lu et al. | Fingerprint liveness detection based on pore analysis | |
Lin et al. | A combination recognition method of palmprint and palm vein based on gray surface matching | |
Hassan et al. | An information-theoretic measure for face recognition: Comparison with structural similarity | |
Oueslati et al. | Identity verification through dorsal hand vein texture based on NSCT coefficients | |
Jeon et al. | Performance evaluation of teeth image recognition system based on difference image entropy | |
Chouhan et al. | Comparative analysis of robust iris recognition system using log gabor wavelet and Laplacian of Gaussian filter | |
Majidpour et al. | Unreadable offline handwriting signature verification based on generative adversarial network using lightweight deep learning architectures | |
Deorankar et al. | Proposing SVM and hog techniques for effective face recognition in video surveillance |
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 |