CN106156727B - A kind of recognition methods and terminal of biological characteristic - Google Patents
A kind of recognition methods and terminal of biological characteristic Download PDFInfo
- Publication number
- CN106156727B CN106156727B CN201610472329.7A CN201610472329A CN106156727B CN 106156727 B CN106156727 B CN 106156727B CN 201610472329 A CN201610472329 A CN 201610472329A CN 106156727 B CN106156727 B CN 106156727B
- Authority
- CN
- China
- Prior art keywords
- feature
- sample
- terminal
- dimensionality reduction
- identification
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/50—Maintenance of biometric data or enrolment thereof
Abstract
The embodiment of the invention discloses a kind of recognition methods of biological characteristic and terminals, the speed for increasing terminal identification biological characteristic.The method comprise the steps that terminal determines target equal error rate and target accuracy of system identification；First primitive character dimensionality reduction of first sample is obtained fisrt feature according to target equal error rate by terminal, and the first primitive character dimensionality reduction is obtained second feature according to target accuracy of system identification, and first sample is the sample that terminal is stored in advance or acquires；Terminal stores fisrt feature and second feature；The Biological imaging that terminal obtains object to be identified generates the second sample and extracts the second primitive character of the second sample；Second primitive character dimensionality reduction is obtained third feature according to target equal error rate by terminal, and the second primitive character dimensionality reduction is obtained fourth feature according to target accuracy of system identification；Identification is compared with fisrt feature and identification is compared with second feature in fourth feature by terminal for third feature obtains recognition result.
Description
Technical field
The present invention relates to the recognition methods and terminal of field of terminal more particularly to a kind of biological characteristic.
Background technique
Biological identification technology is a kind of technology that authentication is carried out using human body biological characteristics, and biological characteristic is unique
(different from other people), can measure or can with the physiological property or behavior of automatic identification and verifying, be divided into physiological characteristic and
Behavioural characteristic.Biological recognition system is sampled feature, extracts its unique features and carries out authentication.Bio-identification
A kind of means of the technology as identification have unique advantage, have been increasingly becoming international research hotspot in recent years.Mesh
The mode of preceding comparative maturity and large-scale use is mainly fingerprint, iris, face, ear, palmmprint, private seal vein etc..And biological characteristic
Identification technology is usually according to scanning, digitized processing, analysis, feature extraction, storage, the several step process of matching classification.
At present in order to obtain higher reliability in living things feature recognition, can usually be mentioned as far as possible in feature extraction phases
Take high-dimensional feature, for example after face piecemeal 9x9 extracts LBP feature and calculates histogram, obtained feature is 9x9x256
Original local binary patterns (full name in English: Local Binary Patterns, abbreviation LBP) feature, if using unified LBP (English
Literary full name: Unified LBP) if, it is characterized in that 9x9x58；What the feature extraction algorithm based on key point punishment block obtained
Feature also can be very much, for example, 7 key points, the method for partition of 4x4 obtains at each key point Unified LBP histogram
Feature is 7x4x4x58=6496.
But the volume of high-dimensional feature is big, feature is more, and then causes terminal calculating process in matching identification complicated, into
And reduce the speed of terminal recognition biological characteristic.
Summary of the invention
The embodiment of the invention provides a kind of recognition methods of biological characteristic and terminals, special for increasing terminal identification biology
The speed of sign.
In a first aspect, the embodiment of the present invention provides a kind of recognition methods of biological characteristic, comprising:
Terminal determines target equal error rate and target accuracy of system identification；Terminal is according to the target equal error rate by first sample
The first primitive character dimensionality reduction obtain fisrt feature, and the first primitive character dimensionality reduction is obtained second according to the target accuracy of system identification
Feature, the first sample are the sample that the terminal is stored in advance or acquires；The terminal by the fisrt feature and the second feature into
Row storage；The Biological imaging that the terminal obtains object to be identified generates the second sample and extracts the second original spy of second sample
Sign；The second primitive character dimensionality reduction is obtained third feature according to the target equal error rate by the terminal, and is recognized according to the target
The second primitive character dimensionality reduction is obtained fourth feature by false rate；Knowledge is compared with the fisrt feature in the third feature by the terminal
Not and identification is compared with the second feature in the fourth feature and obtains recognition result.
In practical applications, terminal obtains the first primitive character dimensionality reduction of first sample according to the target equal error rate
Fisrt feature can be after the first primitive character dimensionality reduction is obtained second feature according to target accuracy of system identification by the terminal, the terminal
The second feature dimensionality reduction is obtained into the fisrt feature further according to target equal error rate, concrete mode is herein without limitation.
In a kind of possible implementation, which is in the first value range the fisrt feature not to be surpassed
The minimum equal error rate of the second value range is crossed, which is the value range of the target equal error rate, should
Second value range is the value range of the fisrt feature；
The target accuracy of system identification is that the minimum for making the second feature be no more than the 4th value range in third value range is recognized
False rate, the third value range are the value range of the target accuracy of system identification, and the 4th value range is the value of the second feature
Range.
In alternatively possible implementation, which is compared identification with the fisrt feature for the third feature and should
Fourth feature be compared with the second feature identification obtain recognition result include: the terminal judge the third feature and this first
Whether the similarity of feature is less than first threshold；
If the similarity of the third feature and the fisrt feature is not less than the first threshold, which judges the 4th spy
Whether sign is less than second threshold with the second feature；
If the similarity of the fourth feature and the second feature is not less than second threshold, which judges second sample
It is identical as the first sample.
In alternatively possible implementation, which judges whether the third feature and the similarity of the fisrt feature are less than
After first threshold, this method further include:, should if the similarity of the third feature and the fisrt feature is less than the first threshold
Terminal judges that second sample and the first sample be not identical.
In alternatively possible implementation, which judges whether the fourth feature and the second feature are less than second threshold
Later, this method further include: if the similarity of the fourth feature and the second feature is less than second threshold, terminal judgement should
Second sample and the first sample be not identical.
In alternatively possible implementation, which carries out storage for the fisrt feature and the second feature and includes:
The fisrt feature is quantified to obtain the first quantization characteristic according to relational expression by the terminal, and according to the relational expression by this
Two features are quantified to obtain the second quantization characteristic；
The terminal stores first quantization characteristic and second quantization characteristic；
Identification is compared with the fisrt feature and by the fourth feature and the second feature by the terminal for the third feature
It identification is compared obtains recognition result and include:
The terminal quantifies the third feature according to the relational expression to obtain third quantization characteristic, and should according to the relational expression
Fourth feature is quantified to obtain the 4th quantization characteristic；
The terminal third quantization characteristic is compared with the first quantization characteristic identification and by the 4th quantization characteristic with
Second quantization characteristic is compared identification and obtains recognition result.
In practical applications, which can also store personal identification corresponding with the fisrt feature and the second feature
Password (full name in English: Personal Identification Number, referred to as: PIN).
In alternatively possible implementation, the relational expression are as follows:
Wherein the V is the feature value after sample dimensionality reduction, the VminFor the minimal characteristic value after sample dimensionality reduction, the VmaxFor
Maximum feature value after sample dimensionality reduction, the value of N are 255 or 65535.
Second aspect, the embodiment of the present invention provide a kind of terminal, comprising:
Determining module, for determining target equal error rate and target accuracy of system identification；
First dimensionality reduction module, the target equal error rate for being determined according to the determining module is by the first of first sample
Primitive character dimensionality reduction obtains fisrt feature, and is dropped first primitive character according to the target accuracy of system identification that the determining module determines
Dimension obtains second feature, which is the sample that the terminal is stored in advance or acquires；
Memory module, the fisrt feature and the second feature for obtaining the dimensionality reduction module dimensionality reduction store；
Module is obtained, the Biological imaging for obtaining object to be identified generates the second sample and extracts the of second sample
Two primitive characters；
Second dimensionality reduction module, the target equal error rate for being determined according to the determining module obtain the acquisition module
The second primitive character dimensionality reduction target accuracy of system identification for obtaining third feature, and being determined according to the determining module by the acquisition mould
The second primitive character dimensionality reduction that block obtains obtains fourth feature；
Identification module, the third feature and the first dimensionality reduction module dimensionality reduction for obtaining the second dimensionality reduction module dimensionality reduction
The fourth feature and first drop for identifying and obtaining the second dimensionality reduction module dimensionality reduction is compared in the obtained fisrt feature
The second feature that dimension module dimensionality reduction obtains is compared identification and obtains recognition result.
In a kind of possible implementation, which is in the first value range the fisrt feature not to be surpassed
The minimum equal error rate of the second value range is crossed, which is the value range of the target equal error rate, should
Second value range is the value range of the fisrt feature；
The target accuracy of system identification is that the minimum for making the second feature be no more than the 4th value range in third value range is recognized
False rate, the third value range are the value range of the target accuracy of system identification, and the 4th value range is the value of the second feature
Range.
In alternatively possible implementation, which includes:
First judging unit, for judging whether the third feature and the similarity of the fisrt feature are less than first threshold；
Second judgment unit, if judging the similarity of the third feature and the fisrt feature not for first judging unit
Less than the first threshold, then judge whether the fourth feature and the second feature are less than second threshold；
First recognition unit, if second judging that the fourth feature and the similarity of the second feature are not less than for this
Second threshold then judges that second sample is identical as the first sample.
In alternatively possible implementation, the identification module further include:
Second recognition unit, if judging that the similarity of the third feature and the fisrt feature is small for first judging unit
In the first threshold, then judge that second sample and the first sample be not identical.
In alternatively possible implementation, the identification module further include:
Third recognition unit, if judging that the similarity of the fourth feature and the second feature is small for the second judgment unit
In second threshold, then the terminal judges that second sample and the first sample be not identical.
In alternatively possible implementation, which includes:
First quantifying unit obtains the first quantization characteristic for quantifying the fisrt feature according to relational expression, and according to this
Relational expression is quantified the second feature to obtain the second quantization characteristic；
Storage unit, first quantization characteristic and second quantization characteristic for quantifying the quantifying unit carry out
Storage；
The identification module includes:
Second quantifying unit, for quantifying to obtain third quantization characteristic for the third feature according to the relational expression, and according to
The relational expression is quantified the fourth feature to obtain the 4th quantization characteristic；
4th recognition unit, for the third quantization characteristic to be compared to identification with the first quantization characteristic and by the 4th
Quantization characteristic is compared identification with second quantization characteristic and obtains recognition result.
In alternatively possible implementation, the relational expression are as follows:
Wherein the V is the feature value after sample dimensionality reduction, the VminFor the minimal characteristic value after sample dimensionality reduction, the VmaxFor
Maximum feature value after sample dimensionality reduction, the value of N are 255 or 65535.
The third aspect, the embodiment of the present invention provide a kind of terminal, comprising:
Transceiver, processor, memory and bus；
The transceiver, the processor are connected with the memory by the bus；
The processor has following function: determining target equal error rate and target accuracy of system identification；It is true according to the determining module
First primitive character dimensionality reduction of first sample is obtained fisrt feature by the fixed target equal error rate, and according to the determining module
The first primitive character dimensionality reduction is obtained second feature by the determining target accuracy of system identification, which is that the terminal is stored in advance
Or the sample of acquisition；
The memory has following function: fisrt feature and the second feature that the dimensionality reduction module dimensionality reduction is obtained carry out
Storage；
The transceiver have following function: obtain object to be identified Biological imaging generate the second sample and extract this second
Second primitive character of sample；
The processor has following function: the target equal error rate determined according to the determining module is by the acquisition module
The the second primitive character dimensionality reduction obtained obtains third feature, and is obtained this according to the target accuracy of system identification that the determining module determines
The second primitive character dimensionality reduction that modulus block obtains obtains fourth feature；The third that the second dimensionality reduction module dimensionality reduction is obtained is special
Sign is compared with the fisrt feature that the first dimensionality reduction module dimensionality reduction obtains to be identified and obtains the second dimensionality reduction module dimensionality reduction
The fourth feature identification be compared with the second feature that the first dimensionality reduction module dimensionality reduction obtains obtain recognition result.
As can be seen from the above technical solutions, the embodiment of the present invention has the advantage that terminal according to target equal error
Rate and target accuracy of system identification carry out that dimensionality reduction obtains fisrt feature and second feature is gone forward side by side to the first primitive character of first sample respectively
Row storage；When the Biological imaging that terminal obtains object to be identified obtains the second sample and extracts the second original spy of the second sample
After sign, dimensionality reduction is carried out to the second primitive character respectively also according to target equal error rate and target accuracy of system identification and obtains third feature
And fourth feature；During the second sample and first sample matching identification, third feature is compared with fisrt feature, the
Four features are compared with second feature, due to reducing the biological characteristic of first sample and the second sample, to reduce end
The complexity of calculating process during matching identification is held, and then accelerates the speed of terminal recognition biological characteristic.
Detailed description of the invention
Fig. 1 is one embodiment schematic diagram of the recognition methods of biological characteristic in the embodiment of the present invention；
Fig. 2 is one embodiment schematic diagram of terminal in the embodiment of the present invention；
Fig. 3 is another embodiment schematic diagram of terminal in the embodiment of the present invention.
Specific embodiment
The embodiment of the invention provides a kind of recognition methods of biological characteristic and terminals, special for increasing terminal identification biology
The speed of sign.
In order to enable those skilled in the art to better understand the solution of the present invention, below in conjunction in the embodiment of the present invention
Attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is only
The embodiment of a part of the invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
The model that the present invention protects all should belong in member's every other embodiment obtained without making creative work
It encloses.
It below by specific embodiment, is described in detail respectively, referring to Fig. 1, the embodiment of the present invention provides a kind of life
The recognition methods of object feature, comprising:
101, terminal determines target equal error rate and target accuracy of system identification.
The value range of feature after the pre-set high-dimensional primitive character dimensionality reduction of user, the value model of equal error rate
Enclose the value range with accuracy of system identification；Then it selects to make according to after equal error rate dimensionality reduction in the value range of equal error rate
Feature value be no more than preparatory value range minimum equal error rate be target equal error rate, then in accuracy of system identification
Selection can make the minimum accuracy of system identification for being no more than preparatory value range according to the value of the feature after accuracy of system identification dimensionality reduction in value range
For target accuracy of system identification.
In the present embodiment, by taking the primitive character dimension of face is 6496 as an example, if after according to equal error rate dimensionality reduction
The value range of feature is [10,20], is [200,400] according to the value range of the feature after accuracy of system identification dimensionality reduction, and equal mistake
Accidentally the value range of rate is the value range [0.000001,0.000003] of [0.03,0.05] and accuracy of system identification, and practical application
In, if the terminal can be by primitive character dimensionality reduction to 10 dimensions, when equal error rate value when equal error rate value is 0.05
When being 0.03, which can be by primitive character dimensionality reduction to 20 dimensions, then the terminal selects 0.03 as target equal error rate, together
If reason, when accuracy of system identification value is 0.000001, which can be by primitive character dimensionality reduction to 400 dimensions, when accuracy of system identification value is
When 0.000003, which can be by primitive character dimensionality reduction to 200 dimensions, then the terminal selects 0.000001 as the equal mistake of target
Accidentally rate.
102, the first primitive character dimensionality reduction of first sample is obtained fisrt feature according to target equal error rate by terminal, and
The first primitive character dimensionality reduction is obtained into second feature according to target accuracy of system identification.
The terminal gets the first original spy for needing the first sample for being stored in advance or acquiring and extracting the first sample
Sign.First primitive character progress dimensionality reduction is obtained fisrt feature according to the target equal error rate determined by the terminal, according to
First primitive character progress dimensionality reduction is obtained second feature by the target accuracy of system identification determined.
In practical applications, when terminal extracts the first primitive character of face, various ways, such as face can be used
After piecemeal 9x9 extracts LBP feature and calculates histogram, obtained feature is 9x9x256 original LBP features, if using
If Unified LBP, it is characterized in that 9x9x58；The feature that feature extraction algorithm based on key point punishment block obtains also can
Very much, for example, 7 key points, the Unified LBP histogram feature that the method for partition of 4x4 obtains at each key point is
7x4x4x58=6496, the extracting mode of the primitive character of sample is herein without limitation.The present embodiment is with face primitive character
For 6496, dimensionality reduction is carried out to the face according to the target equal error rate and target accuracy of system identification selected in step 101 and is obtained
Fisrt feature be 20 dimension, second feature be 400 dimension.
103, terminal stores fisrt feature and second feature.
Terminal by after dimensionality reduction fisrt feature and second feature store.
In practical applications, which can also determine PIN corresponding with the fisrt feature and the second feature, and will
The PIN is stored simultaneously with the fisrt feature and the second feature, can be directly read in this way in specific compare this
One feature and the second feature.
In practical applications, if the memory space of terminal is little, which can also be according to relational expression:Fisrt feature and second feature are quantified to obtain the first quantization characteristic and the second quantization characteristic,
For example fisrt feature is 20 dimensions, the value of one of dimension is 0.5, and the value of another dimension is 2, and the value of maximum dimension is 3,
The value that the value of the smallest dimension is 0.2, N is 255, then the quantized value difference of the feature quantified according to relational expression is as follows:
The quantized value that the quantized value that the quantized value that 0.5 quantized value is 27,2 is 163,3 is 255,0.2 is 0, i.e. the fisrt feature can be with
It is indicated with 0 to 255 without 8 integers of symbol, while these signless 8 integers can carry out table with a byte
Show, then first quantization characteristic only takes up 20 bytes when carrying out storing first quantization characteristic, greatly reduced occupancy
Memory space.Simultaneously in practical application, when quantization, can also select -128 to 127 8 signed integers or 0 to 65535
16 signless integers or -32768 to 32,767 16 signed integers indicate that feature, concrete mode do not limit herein
It is fixed.
104, the Biological imaging that terminal obtains object to be identified generates the second sample and the second of the second sample of extraction is original
Feature.
Terminal during use, obtains the Biological imaging of object to be identified, such as face, get Biological imaging it
Afterwards by as the second sample identified, and extract the second primitive character of the Biological imaging.
In practical applications, the Biological imaging for the object to be identified that terminal obtains can be fingerprint, iris, face, ear, the palm
Line, private seal vein etc., while the mode of the Biological imaging of terminal acquisition object to be identified can be through camera, infrared scan
The mode that the second primitive character is extracted Deng, terminal is also possible to extract LBP feature and calculates histogram or punished based on key point
The feature extraction algorithm of block obtains feature, specifically herein without limitation.
105, the second primitive character dimensionality reduction is obtained third feature according to target equal error rate by terminal, and according to the target
The second primitive character dimensionality reduction is obtained fourth feature by accuracy of system identification.
Second primitive character progress dimensionality reduction is obtained third feature according to the target equal error rate determined by the terminal,
Second primitive character progress dimensionality reduction is obtained into fourth feature according to the target accuracy of system identification determined.
In practical applications, if terminal in the case where memory space is small by the first sample for being stored in advance or acquiring
Fisrt feature and second feature store again after being quantified, then the terminal also need the third feature of second sample and
Fourth feature is quantified to obtain quantization characteristic in the same way.
106, third feature is compared terminal with fisrt feature identifies and compares fourth feature and second feature
Recognition result is obtained to identification.
The first threshold of the similarity of the terminal profile third feature and the fisrt feature, and set the fourth feature with
The second threshold of the similarity of the second feature, then the terminal judge the third feature and the fisrt feature similarity whether
Less than first threshold；If the similarity of the third feature and the fisrt feature is not less than the first threshold, terminal judgement should
Whether fourth feature and the second feature are less than second threshold；If the similarity of the fourth feature and the second feature is not less than the
Two threshold values, then the terminal judges that second sample is identical as the first sample.If the third feature is similar to the fisrt feature
Degree is less than the first threshold, then the terminal judges that second sample and the first sample be not identical, while can not have to comparing
Identify the similarity between fourth feature and second feature；If the similarity of the third feature and the fisrt feature not less than this
When the similarity of one threshold value and the fourth feature and the second feature is less than the second threshold, which equally judges second sample
This is not identical as the first sample.
For example first threshold is 0.8, second threshold 0.9, the similarity of third feature and fisrt feature is 0.85, this
When four features and second feature similarity are 0.9, which judges that the second sample and first sample are same people, if the third is special
The similarity of sign and fisrt feature is 0.79, then terminal judges the second sample and first sample is not that same people or the third are special
The similarity of sign and fisrt feature is 0.85, then the similarity of the fourth feature and second feature is 0.89, then the terminal is same
Judge that second sample and the first sample are not same people.
In practical applications, if terminal is special by the first of the pre-stored first sample in the case where memory space is small
It seeks peace after second feature is quantified and stores again, then the terminal is then needed when identification is compared by third quantization characteristic
Known with the second quantization characteristic using above-mentioned same matching identification mode with the first quantization characteristic and the 4th quantization characteristic
Not.
For ease of understanding, the recognition methods of biological characteristic in the present embodiment is carried out with a practical application scene below detailed
Thin to describe, in the present embodiment, terminal is including but not limited to mobile phone, punched-card machine etc., and terminal is by taking punched-card machine as an example.
If the punched-card machine uses one-to-many identification method, it assumes that tri- people of company personnel A, B, C, the punched-card machine need
Sample A is prestored, sample B, sample C, sample standard deviation is face herein.The primitive character of sample A, sample B and sample C are 6496
Dimension, then according to equal error rate 0.03 respectively by sample A, sample B, sample C carries out dimensionality reduction and obtains characteristic set the punched-card machine
(20A, 20B, 20C)；Further according to accuracy of system identification 0.000001 respectively by sample A, sample B, sample C carries out dimensionality reduction and obtains feature set
It closes (400A, 400B, 400C).The punched-card machine again respectively by characteristic set (20A, 20B, 20C) and characteristic set (400A, 400B,
It 400C) is stored, i.e., characteristic set (20A, 20B, 20C) occupies 3*20 byte, and characteristic set (400A, 400B, 400C) accounts for
With 3*400 byte.If employee A checks card, punched-card machine gets the face image of employee A as knowledge by camera
Very this simultaneously extracts primitive character, then in the same way by the primitive character dimensionality reduction of employee A quantify to obtain feature 20a and
Feature 400a.The characteristic set of storage (20A, 20B, 20C) is first loaded into memory before matching identification by the punched-card machine, then will
20a is compared to obtain following result respectively with 20A, 20B, 20C: 20a is 0.85,20a and 20B similarity with 20A similarity
Be 0.8,20a and 20C similarity be 0.75, punched-card machine judgement herein is that the condition of same people is that similarity is not less than 0.8, this
When similarity not less than 0.8 the first quantization characteristic be (20A, 20B)；The punched-card machine again by the characteristic set of storage (400A,
400B) be loaded into memory, then 400a is compared to obtain with 400A, 400B following result respectively: 400a is with 400A similarity
0.95,400a with 400B similarity is 0.5, and punched-card machine judgement herein is that the condition of same people is similarity not less than 0.9.Due to
The face image Sample Similarity of the only feature of sample A and employee A reaches requirement, then punched-card machine thinks that sample A is that employee A is
Same people simultaneously shows.If in practical applications the punched-card machine 400a is compared respectively obtain with 400A, 400B as
Lower result: 400a is 0.95,400a with 400A similarity and 400B similarity is 0.9, and punched-card machine judgement herein is same people
Condition is similarity not less than 0.9.Then the punched-card machine may determine that sample A is same people with employee A with sample B and shows
Show；Or can be only considered that the sample A and employee A that similarity is 0.95 is same people and shows, i.e., only select similarity
Highest sample is final result；Or two similarities again can be compared third threshold value, third threshold herein
Value is 0.95, then the punched-card machine finally thinks that sample A and employee A is same people and shows.
Simultaneously punched-card machine can also use one-to-one verification mode, then the punched-card machine storage sample A feature (20A,
When 400A), it can determine whether in the memory for being saved in the punched-card machine together with a PIN corresponding with the feature of sample A.Employee A
When needing to check card, the PIN is inputted, which reads from memory inputs the corresponding feature of the PIN, i.e., (20A, 400A) is carried
Enter memory, while acquiring the face image of employee A, extracts feature and dimensionality reduction obtains (20a, 400a)；Punched-card machine comparison (20A,
20a), it if its similarity is less than first threshold 0.8, checks card unsuccessfully；If similarity is not less than 0.8, punched-card machine further compares
To (400A, 400a), if its similarity not less than success of checking card if second threshold 0.9, the mistake if similarity is checked card less than 0.9
It loses.
Terminal is original to the first of first sample respectively according to target equal error rate and target accuracy of system identification in the present embodiment
Feature carries out dimensionality reduction and obtains fisrt feature and second feature and stored；When the Biological imaging that terminal obtains object to be identified obtains
To the second sample and after extracting the second primitive character of the second sample, also according to target equal error rate and target accuracy of system identification
Dimensionality reduction is carried out to the second primitive character respectively and obtains third feature and fourth feature；When the second sample and first sample matching identification
During, third feature is compared with fisrt feature, and fourth feature is compared with second feature, due to reducing first
The biological characteristic of sample and the second sample, to reduce the complexity of terminal calculating process during matching identification, in turn
Accelerate the speed of terminal recognition biological characteristic.Fisrt feature and second feature can be subjected to quantization storage simultaneously, in this way may be used
Effectively to save memory space.
The recognition methods of the biological characteristic in the embodiment of the present invention is described above, is described below in the embodiment of the present invention
Terminal, referring to Fig. 2, one embodiment of the terminal in the embodiment of the present invention includes:
Determining module 201, for determining target equal error rate and target accuracy of system identification；
First dimensionality reduction module 202, the target equal error rate for being determined according to the determining module is by first sample
First primitive character dimensionality reduction obtains fisrt feature, and the target accuracy of system identification determined according to the determining module is by the first original spy
Sign dimensionality reduction obtains second feature, which is the sample that the terminal is stored in advance or acquires；
Memory module 203, the fisrt feature and the second feature for obtaining the dimensionality reduction module dimensionality reduction store；
Module 204 is obtained, the Biological imaging for obtaining object to be identified generates the second sample and extracts second sample
The second primitive character；
Second dimensionality reduction module 205, the target equal error rate for being determined according to the determining module is by the acquisition module
The the second primitive character dimensionality reduction obtained obtains third feature, and is obtained this according to the target accuracy of system identification that the determining module determines
The second primitive character dimensionality reduction that modulus block obtains obtains fourth feature；
Identification module 206, the third feature and the first dimensionality reduction module for obtaining the second dimensionality reduction module dimensionality reduction
The fisrt feature that dimensionality reduction obtains be compared identify and the fourth feature that obtains the second dimensionality reduction module dimensionality reduction and this
The second feature that one dimensionality reduction module dimensionality reduction obtains is compared identification and obtains recognition result.
Optionally, the target equal error rate is in the first value range the fisrt feature to be no more than in the present embodiment
The minimum equal error rate of second value range, first value range are the value range of the target equal error rate, this
Two value ranges are the value range of the fisrt feature；
The target accuracy of system identification is that the minimum for making the second feature be no more than the 4th value range in third value range is recognized
False rate, the third value range are the value range of the target accuracy of system identification, and the 4th value range is the value of the second feature
Range.
Optionally, which includes:
First judging unit, for judging whether the third feature and the similarity of the fisrt feature are less than first threshold；
Second judgment unit, if judging the similarity of the third feature and the fisrt feature not for first judging unit
Less than the first threshold, then judge whether the fourth feature and the second feature are less than second threshold；
First recognition unit, if second judging that the fourth feature and the similarity of the second feature are not less than for this
Second threshold then judges that second sample is identical as the first sample.
Optionally, the identification module 206 further include:
Second recognition unit, if judging that the similarity of the third feature and the fisrt feature is small for first judging unit
In the first threshold, then judge that second sample and the first sample be not identical.
Optionally, the identification module 206 further include:
Third recognition unit, if judging that the similarity of the fourth feature and the second feature is small for the second judgment unit
In second threshold, then the terminal judges that second sample and the first sample be not identical.
Optionally, which includes:
First quantifying unit obtains the first quantization characteristic for quantifying the fisrt feature according to relational expression, and according to this
Relational expression is quantified the second feature to obtain the second quantization characteristic；
Storage unit, first quantization characteristic and second quantization characteristic for quantifying the quantifying unit carry out
Storage；
The identification module 206 includes:
Second quantifying unit, for quantifying to obtain third quantization characteristic for the third feature according to the relational expression, and according to
The relational expression is quantified the fourth feature to obtain the 4th quantization characteristic；
4th recognition unit, for the third quantization characteristic to be compared to identification with the first quantization characteristic and by the 4th
Quantization characteristic is compared identification with second quantization characteristic and obtains recognition result.
Optionally, the relational expression are as follows:
Wherein the V is the feature value after sample dimensionality reduction, the VminFor the minimal characteristic value after sample dimensionality reduction, the VmaxFor
Maximum feature value after sample dimensionality reduction, the value of N are 255 or 65535.
The target equal error rate and target that the first dimensionality reduction module 202 is determined according to determining module 201 in the present embodiment are recognized
False rate carries out dimensionality reduction to the first primitive character of first sample respectively and obtains fisrt feature and second feature and by storage storage
Module 203 is stored；The second original that Biological imaging obtains the second sample and extracts the second sample is obtained when obtaining module 204
After beginning feature, the second dimensionality reduction module 205 is also according to target equal error rate and target accuracy of system identification respectively to the second primitive character
It carries out dimensionality reduction and obtains third feature and fourth feature；During the second sample and first sample matching identification, identification module
206 third feature is compared with fisrt feature, and fourth feature is compared with second feature, due to reducing first sample
With the biological characteristic of the second sample, to reduce the complexity of terminal calculating process during matching identification, and then accelerate
The speed of terminal recognition biological characteristic.
Referring specifically to Fig. 3, another embodiment of terminal in the embodiment of the present invention, comprising:
Transceiver 301, processor 302, bus 303, memory 304；
The transceiver 301, the processor 302 are connected with the memory 304 by the bus 303；
Bus 303 can be Peripheral Component Interconnect standard (peripheral component interconnect, abbreviation
PCI) bus or expanding the industrial standard structure (extended industry standard architecture, abbreviation EISA)
Bus etc..The bus can be divided into address bus, data/address bus, control bus etc..For convenient for indicating, in Fig. 3 only with one slightly
Line indicates, it is not intended that an only bus or a type of bus.
Processor 302 can be central processing unit (central processing unit, abbreviation CPU), network processing unit
The combination of (network processor, abbreviation NP) or CPU and NP.
Processor 302 can further include hardware chip.Above-mentioned hardware chip can be specific integrated circuit
(application-specific integrated circuit, abbreviation ASIC), programmable logic device
(programmable logic device, abbreviation PLD) or combinations thereof.Above-mentioned PLD can be Complex Programmable Logic Devices
(complex programmable logic device, abbreviation CPLD), field programmable gate array (field-
Programmable gate array, abbreviation FPGA), Universal Array Logic (generic array logic, abbreviation GAL) or
Any combination thereof.
The memory 304 may include volatile memory (volatile memory), such as random access memory
(random-access memory, abbreviation RAM)；Memory also may include nonvolatile memory (non-volatile
), such as flash memory (flash memory), hard disk (hard disk drive, abbreviation HDD) or solid state hard disk memory
(solid-state drive, abbreviation SSD)；Memory 304 can also include the combination of the memory of mentioned kind.
Optionally, memory 304 can be also used for storage program instruction, and processor 302 calls to be stored in the memory 304
Program instruction, the one or more steps or in which optional embodiment in embodiment illustrated in fig. 1 can be executed, realize
The function of terminal behavior in the above method.
The processor 302 has following function: determining target equal error rate and target accuracy of system identification；According to the determination mould
First primitive character dimensionality reduction of first sample is obtained fisrt feature by the target equal error rate that block determines, and according to the determination
The first primitive character dimensionality reduction is obtained second feature by the target accuracy of system identification that module determines, which is that the terminal is preparatory
The sample of storage or acquisition；
The memory 304, with following function: fisrt feature and the second feature that the dimensionality reduction module dimensionality reduction is obtained
It is stored；
The transceiver 301, with following function: the Biological imaging for obtaining object to be identified, which generates the second sample and extracts, to be somebody's turn to do
Second primitive character of the second sample；
The processor 302, with following function: the target equal error rate determined according to the determining module is by the acquisition
The second primitive character dimensionality reduction that module obtains obtains third feature, and will according to the target accuracy of system identification that the determining module determines
The second primitive character dimensionality reduction that the acquisition module obtains obtains fourth feature；The second dimensionality reduction module dimensionality reduction is obtained this
Identification is compared with the fisrt feature that the first dimensionality reduction module dimensionality reduction obtains and by the second dimensionality reduction module dimensionality reduction for three features
The obtained fourth feature is compared identification with the second feature that the first dimensionality reduction module dimensionality reduction obtains and obtains recognition result.
Optionally, which is that the fisrt feature is made to be no more than the second value model in the first value range
The minimum equal error rate enclosed, first value range are the value range of the target equal error rate, second value range
For the value range of the fisrt feature；
The target accuracy of system identification is that the minimum for making the second feature be no more than the 4th value range in third value range is recognized
False rate, the third value range are the value range of the target accuracy of system identification, and the 4th value range is the value of the second feature
Range.
Optionally, the processor 302 specifically also has following function: judging the phase of the third feature with the fisrt feature
Whether it is less than first threshold like degree；If first judging unit judges that the third feature and the similarity of the fisrt feature are not less than
The first threshold, then judge whether the fourth feature and the second feature are less than second threshold；If this second judge this
The similarity of four features and the second feature is not less than second threshold, then judges that second sample is identical as the first sample.
Optionally, the processor 302 specifically also has following function: if first judging unit judges the third feature
It is less than the first threshold with the similarity of the fisrt feature, then judges that second sample and the first sample be not identical.
Optionally, the processor 302 specifically also has following function: if the second judgment unit judges the fourth feature
It is less than second threshold with the similarity of the second feature, then the terminal judges that second sample and the first sample be not identical.
Optionally, the memory 304 specifically also has following function: quantifying to obtain by the fisrt feature according to relational expression
First quantization characteristic, and quantified the second feature according to the relational expression to obtain the second quantization characteristic；By the quantifying unit
Quantify obtained first quantization characteristic and second quantization characteristic is stored；
The processor 302 specifically also has following function: being quantified the third feature according to the relational expression to obtain third amount
Change feature, and is quantified the fourth feature according to the relational expression to obtain the 4th quantization characteristic；By the third quantization characteristic with
First quantization characteristic, which is compared identification and identification is compared with second quantization characteristic in the 4th quantization characteristic, to be known
Other result.
Optionally, the relational expression are as follows:
Wherein the V is the feature value after sample dimensionality reduction, the VminFor the minimal characteristic value after sample dimensionality reduction, the VmaxFor
Maximum feature value after sample dimensionality reduction, the value of N are 255 or 65535.
Processor 302 is according to target equal error rate and target accuracy of system identification respectively to the first of first sample in the present embodiment
Primitive character carries out dimensionality reduction and obtains fisrt feature and second feature and stored by memory 304；When transceiver 301 obtains
After the Biological imaging of object to be identified obtains the second sample and extracts the second primitive character of the second sample, processor 302 is same
Sample carries out dimensionality reduction to the second primitive character respectively according to target equal error rate and target accuracy of system identification and obtains third feature and the 4th
Feature；During the second sample and first sample matching identification, processor 302 compares third feature and fisrt feature
Right, fourth feature is compared with second feature, due to reducing the biological characteristic of first sample and the second sample, to reduce
The complexity of terminal calculating process during matching identification, and then accelerate the speed of terminal recognition biological characteristic.
It is apparent to those skilled in the art that for convenience and simplicity of description, the system of foregoing description,
The specific work process of device and unit, can refer to corresponding processes in the foregoing method embodiment, and details are not described herein.
In several embodiments provided herein, it should be understood that disclosed system, device and method can be with
It realizes by another way.For example, the apparatus embodiments described above are merely exemplary, for example, the unit
It divides, only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple units or components
It can be combined or can be integrated into another system, or some features can be ignored or not executed.Another point, it is shown or
The mutual coupling, direct-coupling or communication connection discussed can be through some interfaces, the indirect coupling of device or unit
It closes or communicates to connect, can be electrical property, mechanical or other forms.
The unit as illustrated by the separation member may or may not be physically separated, aobvious as unit
The component shown may or may not be physical unit, it can and it is in one place, or may be distributed over multiple
In network unit.It can select some or all of unit therein according to the actual needs to realize the mesh of this embodiment scheme
's.
It, can also be in addition, the functional units in various embodiments of the present invention may be integrated into one processing unit
It is that each unit physically exists alone, can also be integrated in one unit with two or more units.Above-mentioned integrated list
Member both can take the form of hardware realization, can also realize in the form of software functional units.
If the integrated unit is realized in the form of SFU software functional unit and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, technical solution of the present invention is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a storage medium, including some instructions are used so that a computer
Equipment (can be personal computer, server or the network equipment etc.) executes the complete of each embodiment the method for the present invention
Portion or part steps.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only
Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can store journey
The medium of sequence code.
The above, the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although referring to before
Stating embodiment, invention is explained in detail, those skilled in the art should understand that: it still can be to preceding
Technical solution documented by each embodiment is stated to modify or equivalent replacement of some of the technical features；And these
It modifies or replaces, the spirit and scope for technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution.
Claims (13)
1. a kind of recognition methods of biological characteristic characterized by comprising
Terminal determines that target equal error rate and target accuracy of system identification, the target equal error rate are to make in the first value range
Fisrt feature is no more than the minimum equal error rate of the second value range, and first value range is the target equal error
The value range of rate, second value range are the value range of the fisrt feature；The target accuracy of system identification takes for third
Second feature is made to be no more than the minimum accuracy of system identification of the 4th value range within the scope of value, the third value range is the target
The value range of accuracy of system identification, the 4th value range are the value range of the second feature；
First primitive character dimensionality reduction of first sample is obtained fisrt feature according to the target equal error rate by terminal, and according to
The first primitive character dimensionality reduction is obtained second feature by the target accuracy of system identification, and the first sample is that the terminal is deposited in advance
The sample of storage or acquisition；
The terminal stores the fisrt feature and the second feature；
The Biological imaging that the terminal obtains object to be identified generates the second sample and extracts the second original of second sample
Feature；
The second primitive character dimensionality reduction is obtained third feature according to the target equal error rate by the terminal, and according to institute
It states target accuracy of system identification and the second primitive character dimensionality reduction is obtained into fourth feature；
Identification is compared with the fisrt feature and by the fourth feature and described the by the terminal for the third feature
Two features are compared identification and obtain recognition result.
2. the method according to claim 1, wherein the terminal is by the third feature and the fisrt feature
Identification is compared and with the second feature identification is compared in the fourth feature obtains recognition result and includes:
The terminal judges whether the third feature and the similarity of the fisrt feature are less than first threshold；
If the similarity of the third feature and the fisrt feature is not less than the first threshold, described in the terminal judgement
Whether fourth feature and the second feature are less than second threshold；
If the similarity of the fourth feature and the second feature is not less than second threshold, the terminal judges described second
Sample is identical as the first sample.
3. according to the method described in claim 2, it is characterized in that, the terminal judges the third feature and first spy
Whether the similarity of sign is less than after first threshold, the method also includes:
If the similarity of the third feature and the fisrt feature is less than the first threshold, the terminal judges described the
Two samples and the first sample be not identical.
4. according to the method described in claim 2, it is characterized in that, the terminal judges the fourth feature and second spy
Whether sign is less than after second threshold, the method also includes:
If the similarity of the fourth feature and the second feature is less than second threshold, the terminal judges second sample
This is not identical as the first sample.
5. method according to claim 1 to 4, which is characterized in that the terminal by the fisrt feature and
The second feature carries out storage
The fisrt feature is quantified to obtain the first quantization characteristic according to relational expression by the terminal, and according to the relational expression by institute
Second feature is stated to be quantified to obtain the second quantization characteristic；
The terminal stores first quantization characteristic and second quantization characteristic；
Identification is compared with the fisrt feature and by the fourth feature and described the by the terminal for the third feature
Two features, which are compared identification and obtain recognition result, includes:
The terminal quantifies the third feature according to the relational expression to obtain third quantization characteristic, and according to the relational expression
The fourth feature is quantified to obtain the 4th quantization characteristic；
Identification is compared with the first quantization characteristic and by the 4th quantization characteristic by the terminal for the third quantization characteristic
Identification is compared with second quantization characteristic and obtains recognition result.
6. according to the method described in claim 5, it is characterized in that, the relational expression are as follows:
Wherein the V is the feature value after sample dimensionality reduction, the VminFor the minimal characteristic value after sample dimensionality reduction, the Vmax
For the maximum feature value after sample dimensionality reduction, the value of N is 255 or 65535.
7. a kind of terminal characterized by comprising
Determining module, for determining that target equal error rate and target accuracy of system identification, the target equal error rate are the first value
Fisrt feature is made to be no more than the minimum equal error rate of the second value range in range, first value range is the mesh
The value range of equal error rate is marked, second value range is the value range of the fisrt feature；The target recognizes vacation
Rate is the minimum accuracy of system identification for making second feature be no more than the 4th value range in third value range, the third value range
For the value range of the target accuracy of system identification, the 4th value range is the value range of the second feature；
First dimensionality reduction module, the target equal error rate for being determined according to the determining module is by the first of first sample
Primitive character dimensionality reduction obtains fisrt feature, and original by described first according to the target accuracy of system identification that the determining module determines
Feature Dimension Reduction obtains second feature, and the first sample is the sample that the terminal is stored in advance or acquires；
Memory module, the fisrt feature and the second feature for obtaining the dimensionality reduction module dimensionality reduction store；
Module is obtained, the Biological imaging for obtaining object to be identified generates the second sample and extracts the second of second sample
Primitive character；
Second dimensionality reduction module, the target equal error rate for being determined according to the determining module obtain the acquisition module
The the second primitive character dimensionality reduction taken obtains third feature, and will according to the target accuracy of system identification that the determining module determines
The the second primitive character dimensionality reduction for obtaining module acquisition obtains fourth feature；
Identification module, the third feature and the first dimensionality reduction module for obtaining the second dimensionality reduction module dimensionality reduction drop
Tie up the obtained fisrt feature be compared identify and the fourth feature that obtains the second dimensionality reduction module dimensionality reduction with
The second feature that the first dimensionality reduction module dimensionality reduction obtains is compared identification and obtains recognition result.
8. terminal according to claim 7, which is characterized in that the identification module includes:
First judging unit, for judging whether the third feature and the similarity of the fisrt feature are less than first threshold；
Second judgment unit, if judging the similarity of the third feature Yu the fisrt feature for first judging unit
Not less than the first threshold, then judge whether the fourth feature and the second feature are less than second threshold；
First recognition unit, if judging that the similarity of the fourth feature and the second feature is not small for described second
In second threshold, then judge that second sample is identical as the first sample.
9. terminal according to claim 8, which is characterized in that the identification module further include:
Second recognition unit, if judging the similarity of the third feature Yu the fisrt feature for first judging unit
Less than the first threshold, then judge that second sample and the first sample be not identical.
10. terminal according to claim 8, which is characterized in that the identification module further include:
Third recognition unit, if judging the similarity of the fourth feature Yu the second feature for the second judgment unit
Less than second threshold, then the terminal judges that second sample and the first sample be not identical.
11. terminal according to any one of claims 7 to 10, which is characterized in that the memory module includes:
First quantifying unit obtains the first quantization characteristic for quantifying the fisrt feature according to relational expression, and according to described
Relational expression is quantified the second feature to obtain the second quantization characteristic；
Storage unit, first quantization characteristic and second quantization characteristic for quantifying the quantifying unit into
Row storage；
The identification module includes:
Second quantifying unit, for quantifying to obtain third quantization characteristic for the third feature according to the relational expression, and according to
The relational expression is quantified the fourth feature to obtain the 4th quantization characteristic；
4th recognition unit, for the third quantization characteristic to be compared to identification with the first quantization characteristic and by the described 4th
Quantization characteristic is compared identification with second quantization characteristic and obtains recognition result.
12. terminal according to claim 11, which is characterized in that the relational expression are as follows:
Wherein the V is the feature value after sample dimensionality reduction, the VminFor the minimal characteristic value after sample dimensionality reduction, the Vmax
For the maximum feature value after sample dimensionality reduction, the value of N is 255 or 65535.
13. a kind of terminal characterized by comprising
Transceiver, processor, memory and bus；
The transceiver, the processor are connected with the memory by the bus；
The processor has following function: determining target equal error rate and target accuracy of system identification, the target equal error rate
To make fisrt feature be no more than the minimum equal error rate of the second value range, the first value model in the first value range
It encloses for the value range of the target equal error rate, second value range is the value range of the fisrt feature；Institute
Stating target accuracy of system identification is the minimum accuracy of system identification for making second feature be no more than the 4th value range in third value range, described the
Three value ranges are the value range of the target accuracy of system identification, and the 4th value range is the value model of the second feature
It encloses；The first primitive character dimensionality reduction of first sample is obtained into fisrt feature according to the target equal error rate, and according to described
The first primitive character dimensionality reduction is obtained second feature by target accuracy of system identification, the first sample be the terminal be stored in advance or
The sample of acquisition；
The memory has following function: the fisrt feature and the second feature are stored；
The transceiver has following function: the Biological imaging for obtaining object to be identified generates the second sample and extracts described second
Second primitive character of sample；
The processor has following function: the second primitive character dimensionality reduction being obtained the according to the target equal error rate
Three features, and the second primitive character dimensionality reduction is obtained by fourth feature according to the target accuracy of system identification；By the third feature
Identification is compared with the fisrt feature and identification is compared with the second feature in the fourth feature and is identified
As a result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610472329.7A CN106156727B (en) | 2016-06-24 | 2016-06-24 | A kind of recognition methods and terminal of biological characteristic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610472329.7A CN106156727B (en) | 2016-06-24 | 2016-06-24 | A kind of recognition methods and terminal of biological characteristic |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106156727A CN106156727A (en) | 2016-11-23 |
| CN106156727B true CN106156727B (en) | 2019-07-09 |
Family
ID=57349308
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610472329.7A Active CN106156727B (en) | 2016-06-24 | 2016-06-24 | A kind of recognition methods and terminal of biological characteristic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106156727B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101021900A (en) * | 2007-03-15 | 2007-08-22 | 上海交通大学 | Method for making human face posture estimation utilizing dimension reduction method |
| CN101526997A (en) * | 2009-04-22 | 2009-09-09 | 无锡名鹰科技发展有限公司 | Embedded infrared face image identifying method and identifying device |
| CN103927529A (en) * | 2014-05-05 | 2014-07-16 | 苏州大学 | Acquiring method, application method and application system of final classifier |
| CN105138972A (en) * | 2015-08-11 | 2015-12-09 | 北京天诚盛业科技有限公司 | Face authentication method and device |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8219571B2 (en) * | 2006-02-21 | 2012-07-10 | Samsung Electronics Co., Ltd. | Object verification apparatus and method |
| US20070230754A1 (en) * | 2006-03-30 | 2007-10-04 | Jain Anil K | Level 3 features for fingerprint matching |
| JP2009163555A (en) * | 2008-01-08 | 2009-07-23 | Omron Corp | Face collation apparatus |
-
2016
- 2016-06-24 CN CN201610472329.7A patent/CN106156727B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101021900A (en) * | 2007-03-15 | 2007-08-22 | 上海交通大学 | Method for making human face posture estimation utilizing dimension reduction method |
| CN101526997A (en) * | 2009-04-22 | 2009-09-09 | 无锡名鹰科技发展有限公司 | Embedded infrared face image identifying method and identifying device |
| CN103927529A (en) * | 2014-05-05 | 2014-07-16 | 苏州大学 | Acquiring method, application method and application system of final classifier |
| CN105138972A (en) * | 2015-08-11 | 2015-12-09 | 北京天诚盛业科技有限公司 | Face authentication method and device |
Non-Patent Citations (1)
| Title |
|---|
| "基于最小分类错误率和Parzen窗的降维方法";贺邓超 等;《计算机工程与应用》;20140831;第50卷(第14期);185-188 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106156727A (en) | 2016-11-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20150007295A1 (en) | Biometric-based authentication method, apparatus and system | |
| CN110378301B (en) | Pedestrian re-identification method and system | |
| CN106056083B (en) | A kind of information processing method and terminal | |
| CN111861240A (en) | Suspicious user identification method, device, equipment and readable storage medium | |
| CN111931548B (en) | Face recognition system, method for establishing face recognition data and face recognition method | |
| CN110751069A (en) | Face living body detection method and device | |
| CN108108711A (en) | Face supervision method, electronic equipment and storage medium | |
| CN110647895B (en) | Phishing page identification method based on login box image and related equipment | |
| CN110929244A (en) | Digital identity identification method, device, equipment and storage medium | |
| EP3371739A1 (en) | High speed reference point independent database filtering for fingerprint identification | |
| CN110991231B (en) | Living body detection method and device, server and face recognition equipment | |
| WO2019062588A1 (en) | Information recognition method and apparatus, and electronic device | |
| CN111079585B (en) | Pedestrian re-identification method combining image enhancement with pseudo-twin convolutional neural network | |
| CN106156727B (en) | A kind of recognition methods and terminal of biological characteristic | |
| CN112766162A (en) | Living body detection method, living body detection device, electronic apparatus, and computer-readable storage medium | |
| CN112633281A (en) | Vehicle identity authentication method and system based on Hash algorithm | |
| CN110874602A (en) | Image identification method and device | |
| Aizi et al. | Remote multimodal biometric identification based on the fusion of the iris and the fingerprint | |
| CN114170652A (en) | Face image detection method and device, terminal equipment and storage medium | |
| Sehgal | Palm recognition using LBP and SVM | |
| CN107194362B (en) | Biological identification method and programmable device | |
| CN112200120B (en) | Identity recognition method, living body recognition device and electronic equipment | |
| CN102693415A (en) | Online biology feature identification system | |
| WO2016000509A1 (en) | Data filtering method, and data filter construction method and apparatus | |
| WO2021098772A1 (en) | Assessment method and system for facial verification, and computer storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20170824 Address after: 361008, Xiamen three software park, Fujian Province, 8 North Street, room 2001 Applicant after: Xiamen Central Intelligent Information Technology Co., Ltd. Address before: 361000 Fujian province Xiamen software park two sunrise Road No. 32 403 unit 02 District Applicant before: XIAMEN ZHONGKONG BIOLOGICAL RECOGNITION INFORMATION TECHNOLOGY CO., LTD. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: Room 1301, No.132, Fengqi Road, phase III, software park, Xiamen City, Fujian Province Patentee after: Xiamen Entropy Technology Co., Ltd Address before: 361008 room 8, 2001 North Avenue, Xiamen Software Park, Fujian, three Patentee before: XIAMEN ZKTECO BIOMETRIC IDENTIFICATION TECHNOLOGY Co.,Ltd. |
|
| CP03 | Change of name, title or address |