CN104809428A - Biont sign authentication method - Google Patents
Biont sign authentication method Download PDFInfo
- Publication number
- CN104809428A CN104809428A CN201510136324.2A CN201510136324A CN104809428A CN 104809428 A CN104809428 A CN 104809428A CN 201510136324 A CN201510136324 A CN 201510136324A CN 104809428 A CN104809428 A CN 104809428A
- Authority
- CN
- China
- Prior art keywords
- vein
- hand
- finger
- picture
- angle
- 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.)
- Pending
Links
Landscapes
- Collating Specific Patterns (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a biont signal authentication method. The method comprises that two firm handles are arranged in the middle portion of the horizontal axis, the middle finger is inserted into space between the handles, digital finger images of the contact portion of the middle finger, the index finger, the little finger with the back of the hand, the hand-back authentication end sending the digital hand back image and a server are all connected with a communication line, and hand back vein authentication includes a hand back authentication system. A subcutaneous vein image of the back of the hand is extracted, and used after confirmation of a user; an incomplete finger image and decrease of system performance caused by trauma, ornament or change of the external environment are minimized, the system performance can be stable, and the method for identifying hand-back vein images ensures the convenience of users to the largest degree; and the system flexibility and the error rejection rage caused by movement of the hand of the user are determined according to the micro change of external environment and a gripping image, and the error authentication and recognition rate is reduced.
Description
Technical field
The present invention relates to a kind of biological sign authentication method.
Background technology
Generally, guarantee control of access and the security personnel of extensive user or small-scale user, need security system and the Verification System of person identification, this system uses the magnetic card key of magnetic card, RF card, smart card etc.But, use the person identification mode of magnetic card key fundamentally not preventing card loss, to usurp and misuse etc.In addition, when having large-scale user, the credit card quantity paid due to individual is many, and shortcoming is that financial burden is large.For improving this shortcoming, certification user, develops and uses physiology recognition system, can automatically identify, the unique characteristics of certification user, confirms that user's identity, authenticating identity and control security personnel come in and go out.
Physiology recognition system is divided into the signature identification identifying people's behavioural characteristic, the fingerprint recognition being familiar with physical trait, face recognition, the shape recognition of hand, the identification etc. of hand back vein sample substantially, the representatively content such as the system utilizing these to identify.
In physiology recognition system, it is well known that fingerprint recognition system, because every user has fixing fingerprint, as recognition system, there is sufficient advantage, identify the sample fingerprint be positioned on human epidermal, by good fingerprint picture, carry out personal identification.But, the method of fingerprint recognition system extracts supracutaneous fingerprint characteristic, when extracting the fingerprint of the laborer of building-site and soldier by the method, extract fingerprint well very difficult, this is not the main cause causing system performance lowly unique, due to hand perspiration, injured, external contamination, it is also the main cause causing system performance low.
Secondly, in physiology recognition system, facial-recognition security systems in the present state-of-the technology, due to external change such as exterior lighting, cosmetic, eyes, ornaments, stable recognition performance can not be shown, the Shape Recognition System of hand due to the position of required regular user's hand, the convenient poor performance of user, because the biological characteristic such as rheumatism, arthritis easily changes, cause the degraded performance of system.
Such as, individual identity identification system utilizes hand back vein sample.For improving the shortcoming of original physiology recognition system, individual identity identification system, on the design aspect of instrument, is guaranteed to use this mode in the position of regular user's hand; The side of algorithm is applied high bandwidth process and binary conversion treatment, extraction user vein blood vessel.In other words, utilize individual identity identification system use safety means, the extraction means of hand back vein sample, compare means, wherein security means is that user wears the clothes can holding inductive rod, hold inductive rod, the picture of input user the back of the hand, extraction means extract user's hand back vein sample, relatively means are the corrections to the rotation of user's hand and movement, use and supplement and revise take-off point, this take-off point is the take-off point of take-off point with right side lower floor according to extracting upper strata on the left of the portable the back of the hand blood vessel pattern got.
In addition, when extracting vein sample by extraction means from user's the back of the hand, the directivity of vein sample is not considered, with high bandwidth process and binary conversion treatment procedure extraction sample.According to this extraction means, according to the hand back vein sample individual identity identification system of technology before, by the directivity of sample, the proper vector (sample take-off point) of loss section of vein sample.
Produce system mistake because of the rotation of user's hand and movement to refuse, with when comparing the method for means resolution system False Rejects, the take-off point on upper strata and the take-off point of right side lower floor on the left of the hand back vein sample that utilization is extracted, as standard point, the take-off point of loss extracting section, wrong identification standard scores fulcrum is the main cause causing system performance low.
As mentioned above, in physiology recognition system, the recognition systems such as the Shape Recognition System of fingerprint recognition, face recognition, hand have such problem, due to the external contamination such as fingerprint picture, sweat, wound of breakage, cause degraded performance, and utilize the system of hand back vein sample by being positioned at subepidermal feature, such problem can be solved.But because the hand back vein sample individual identity identification system according to technology before has following problem, owing to losing part sample take-off point in the distortion of vein sample and vein sample, lead to errors identification and False Rejects, causes system performance low.
The hand back vein example extraction algorithms of technology is the original image being obtained user's the back of the hand by camera before, picture in using the vein sample of integrated distribution as area-of-interest (ROI:Region of Interest), obtain the picture of area-of-interest, vein sample is utilized to emphasize filtering (HVPEF:Hand Vein Pattern Emphasizing Filter) in ROI picture, only outstanding vein sample, the boundary member content between the vein sample that outstanding the back of the hand distributes and its background.
Vein sample emphasizes that filtering is with the form with logical (band-pass) filtering, emphasizes the boundary member between the vein sample that in ROI picture, the back of the hand distributes and background, eliminates the low frequency (low-frequency) meeting background parts.Such vein sample emphasizes that the characteristic of filtering is in transitional region broad between high frequency bandwidth sum low frequency bandwidth, the interests of passband (pass-band) can not be maintained, vein sample to be extracted in ROI picture, do not consider the thickness of vein sample, the key character of vein sample is vein sample connectivity (vein-pattern connectivity), loss section of vein sample connectivity.
Binary conversion treatment process is the result using vein sample to emphasize filtering, and gray scale picture converts binaryzation picture (binary image) to, only extracts the distribution situation of hand back vein sample.The result obtained by binary conversion treatment process is divided into black the vein sample that will extract, and all the other background parts are divided into white.
Normalization (normalization) processing procedure is applied to binaryzation picture, and binaryzation similarly is through binary conversion treatment, the vein sample that separately will extract and the picture of background.In other words, normalization process process according to the restriction internal memory of system, must use the database of user equably, in identifying, judges that whether object is as obtaining the normalization of permission as user.
Secondly, elimination between binary conversion treatment process and normalization process process is mixed shadow journey, it is by counter-jamming filtering (NRF:Noise Removal Filter), through binary conversion treatment and positive planning process, eliminate bent by the back of the hand, assorted shadow that fine hair, subcutaneous layer of fat etc. cause.Counter-jamming filtering, as the conventional method eliminating assorted shadow, can use the middle numerical filter of (Moving Average) filtering of the running mean of linear sphere filter and non-linear sphere filter.During the hand back vein sample extraction calculation of technology before uses, the elimination of numerical filter is mixed shadow process.
The hand back vein distribution of technology before extracts that calculation is through the outstanding sphere filter of vein sample, binaryzation and normalization process, shadow process is mixed in elimination, the final vein sample only extracting the back of the hand distribution.Calculation is extracted in this hand back vein distribution following problem, in the ROI picture obtained, contraction and the swelling of vein sample is caused by exterior temperature change, do not consider to shrink and swelling and extract vein sample, lose most of vein sample connectivity in vein sample distribution, stable calculation performance can not be provided.
Furtherly, the main cause of calculation degraded performance is extracted in hand back vein distribution as technology before, there is following problem: in the broad transitional region of passband, boundary information between loss section of vein sample and background and vein sample interior information, the key character of the vein sample extracted is vein sample connectivity, can lose its part connectivity.
Such problem etc., can cause the degraded performance phenomenon that system identification performance is serious.Calculation is extracted by hand back vein distribution, do not consider the thickness of the vein sample that will extract, extract vein sample with single filtering characteristic (single filter characteristics), the key character of the vein sample that extract is vein sample connectivity, can lose its part connectivity.
Summary of the invention
In view of above-mentioned the deficiencies in the prior art, the object of the present invention is to provide a kind of biological sign authentication method.
For solving the problems of the technologies described above, the present invention program comprises:
A kind of biological sign authentication method, it comprises:
Two firm handles are found at the middle part of horizontal axis, middle finger is inserted in the middle of described handle, at this moment obtained middle finger, forefinger, the digital picture in finger and the back of the hand certification end and the server that send the digital picture of the back of the hand of the contact portion of little finger of toe and the back of the hand, all be connected with communication line, described hand back vein certification is exactly the back of the hand Verification System.
Described biological sign authentication method, wherein,
Described server:
Temporary storage
Middle finger inserts in the middle of firm handle by authenticator, with take the hand back vein branch point obtained in advance and be connected, middle finger is inserted the second reference edge line at the second bending range edge that the first reference edge line at the edge of the first bending range that middle finger that handle photographs in advance is connected with little finger of toe and middle finger are connected with forefinger by the data comprising polygonal optimal characteristic distributions and described authenticator, and the two is all as the database that the memory authentication information of finger edge reference information is special;
The digital picture in finger and the digital picture of the back of the hand of described the back of the hand certification end all can be preserved in described having in the shooting storehouse of temporary storage server;
Be saved in the digital picture of the back of the hand of memory at the server of described temporary storage, for other part of vein case that certification is very important, ROI picture is it can be used as to extract, and thick vein case and thin vein case are extracted synthesize, by binaryzation, removing interference vein synthesis case generates ROI picture and extracts, and setting position correction amendment part and vein case compatible portion;
Described biological sign authentication method, wherein, concrete also the comprising of above-mentioned steps:
Position correction amendment portion will arrange following step:
In the digital picture in described finger, the method extracted as finger edge picture of first bending range and the second bending range is for the first bending range of described finger edge picture, in addition transparent processing, obtain around middle finger with little finger of toe around first edge line at edge and the step of middle finger and forefinger edge second edge line:
And/or the focus comprised in the first described edge line combines with the focus comprised in the second edge line and obtains the first straight line, the navigate benchmark focus that comprises of the benchmark focus that first reference edge line comprises and the second reference edge line combines and obtains the second straight line, combined by one or two straight lines and ask for the first relative angle, Here it is obtains the step that angle is revised in the first correction;
The relative angle of the reference opening amount of the first reference edge line that the first edge line aperture comprised by described finger edge picture and finger edge reference information are comprised obtains the second relative angle, in like manner, the relative angle of the reference opening amount of the second edge line aperture that finger edge picture comprises and the second reference edge line that finger edge reference information comprises obtains third phase to angle, obtains the second correction amendment angle by average second relative angle and third phase to angle;
Described amendment in angle in the first correction carries out vein as correction, and carries out correct vein as correction in the second correction amendment angle;
And/or described vein is as matching part:
Obtain each branch point of the correction picture of described vein case, multiple branch point is bonded polygon, by this polygon compared with the suitableeest described distribution character data information, judge whether consistent;
It is described that can to show the back of the hand certification end announcer to what represent the feature of the method for result be exactly the back of the hand Verification System.
Described biological sign authentication method, wherein, concrete also the comprising of above-mentioned steps: the special shooting in described finger camera part and the back of the hand special shooting camera part are one of features of the back of the hand Verification System of the CCD responded to visible rays and infrared spectral range or the application Xiang Yizhong photographing feature with MOS.
Described biological sign authentication method, wherein, concrete also the comprising of above-mentioned steps:
The suitableeest described distribution character packet contains the difference angle of the hand back vein picture in branch point, branch point number, branch point pixel coordinate;
Described vein is as matching part
Obtain the pixel coordinate of described vein as the branch point of the vein picture of correction picture, branch point number, difference angle, judge whether consistent.
Described biological sign authentication method, wherein, concrete also the comprising of above-mentioned steps: the vein picture of the suitableeest described distribution character data comprises thick vein picture and thin vein picture.
Described biological sign authentication method, wherein, concrete also the comprising of above-mentioned steps: the suitableeest distribution character data are relevant to authentication password also can be saved.
Described biological sign authentication method, wherein, utilizes probability match algorithm to calculate the threshold value of ROI area occupied ratio in described the back of the hand preimage.
Described biological sign authentication method, wherein, first edge line of described finger edge picture, the aperture ak of the second edge line is positioned between two pixels to represent x-axis distance, bk is the range difference of the pixel of y-axis direction angulation, utilize following mathematical formulae to be merged by two the θ k tried to achieve, this is the feature of the back of the hand Verification System of application item one ~ 6, can be obtained by formula (1):
Θ k=tan ak/bk formula (1).
Described biological sign authentication method, wherein, concrete also the comprising of above-mentioned steps: the suitableeest distribution character packet of described authentication information private database contains the position on the described each summit of benchmark polygon, the difference angle of the straight line that each summit sends, length between summit, polygonal area;
Described vein is as matching part
For the position on each summit, the difference angle of the straight line sent from each summit, the length between summit, polygonal area, judges its degree of integration of understanding;
Described vein is as matching part
To the branch point application branch point weight of described vein picture of trying to achieve, to the position on each summit, the difference angle of the straight line that this each summit sends, the length between summit, difference number understanding and sound judgment degree of integration;
Described the back of the hand certification end
Prepare finger photographic camera part and the back of the hand photographic camera part, these cameras are connected with microswitch, take pictures in the lower part of this microswitch.
The biological sign authentication method of one provided by the invention, it is the vein picture extracting dorsal cutaneous upper/lower positions, confirms rear use via user; By incomplete fingerprint picture or because sweat, wound, the external environment conditions such as ornament change and cause system performance to be lowly tending towards minimizing, and can provide stable system performance, can ensure the person recognition method of the hand back vein picture of the convenience of user to greatest extent; System sensitivity subtle change according to external environment condition and grip picture judged and the false rejection rate caused because the hand of user is in rotary moving, erroneous authentication discrimination reduces, and in a planned way improves the stability of the certification identification of vein picture and system.
Attached explanation
Fig. 1 is the flow process signal of biological sign authentication method in the present invention.
Embodiment
The invention provides a kind of biological sign authentication method, for making object of the present invention, technical scheme and effect clearly, clearly, the present invention is described in more detail below.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.
The invention provides a kind of biological sign authentication method, as shown in Figure 1, two firm handles are found at the middle part of horizontal axis, middle finger is inserted in the middle of described handle, at this moment obtained middle finger, forefinger, the digital picture in finger and the back of the hand certification end and the server that send the digital picture of the back of the hand of the contact portion of little finger of toe and the back of the hand, all be connected with communication line, described hand back vein certification is exactly the back of the hand Verification System.
Next, the back of the hand authentication method of the present invention is described in detail.
First, user inputs individual recognition code according to individual recognition code input medium.The individual recognition code be transfused to is transported to individual recognition code detection unit by communication adjustment portion.The individual recognition code detection unit being subject to individual recognition code, by database access portion, is retrieved the information remembered as the individual recognition code authentication information database that retrieval is crucial and is determined whether to log in.Result of determination, with representing that means export, when not logging in, shows mistake.Only have and confirm to log in, just can carry out the back of the hand shooting.
Confirm that middle finger is inserted in the finger insertion part of stable handle by the user logged in.Finger panoramic camera just can take the state picture of the finger being positioned at stable handle.（S103）
Then, it is motionless that hand is placed on the stable handle epimere time by user, keeps stationary state.In this state, the back of the hand panoramic camera takes the back of the hand picture, and this back of the hand picture is exactly preimage.（S105）
The finger picture photographed and the back of the hand preimage are sent to video recorder code translator and carry out biometrics process.The photogra storage unit of processor delivered to by picture after biometrics by communication adjustment portion, each digital finger picture and digital preimage are temporarily stored apparatus and preserve.（S106）
Next, in the digital preimage of the back of the hand, the vein picture of necessity when needing to understand fully certification, and only extract ROI picture part.（S0）
In order to extract the ROI picture of the vein in user's the back of the hand preimage as integrated distribution, correct ROI position is very important.So once the module position setup error of lens and camera, the position of ROI can present difference in each camera.So the ROI automatic detection algorithm according to following description just can determine correct ROI position.
First, in ROI picture (Hri) leaching process, the overall pixel forming the back of the hand numeral original image, transverse axis obtains the gray-scale value of bright level (0 to 5), the longitudinal axis obtains the pixel number that this gray-scale value has, make Nogata (S1) with this.
Secondly, use Ptile method, calculate according to Nogata threshold value (T) (S2) of the gray-scale value of ROI field and background border.Textural at the back of the hand certification end device, during the ROI field of the vein blood vessel sample that shooting is necessary, may there be deviation each camera site, but in the whole digital original image photographed, shared area ratio is certain.Therefore, in picture, know the area ratio shared by target, according to Nogata, the P% point of its area ratio regards threshold value (T) as, uses Ptile method.Except Ptile method, also Land use models (mode) method may detect Nogata paddy.
Extract the pixel of threshold value (T) above gray-scale value, determine ROI field (S3).For giving prominence to the edge in ROI field further, carrying out Sobel filtering process, extracting ROI picture (Hri) (S4).
Secondly, for the vein sample in the ROI picture (Hri) that reading has been extracted, carry out the vein example extraction algorithms (FBHVPEA:Filter-bank Hand Vein-pattern extraction Algorithm) process (S300) of bank of filters.According to this process, the change because of external environment condition produces contraction and the swelling of vein sample, and the vein sample information of loss part be made as far as possible to minimize, and connectivity, as vein sample key character, maintains connectivity to greatest extent
First, in the vein sample that the back of the hand distributes, with general mode (Normal mode) filtering of the vein example extraction algorithms of bank of filters, effectively extract the sample (SPi) of thick vein blood vessel.And then, with enhancement mode (Enhancement mode) filtering, effectively extract thin vein blood vessel sample (ZPi) (S302) of thin vein blood vessel and contraction state.The thick vein blood vessel sample (SPi) obtained in progress before synthesis and thin vein blood vessel sample (ZPi), make the vein sample synthesized image (GPi) of the back of the hand.General mode filtering can maintain low-pass information in vein sample and sample boundary information, and compared with thick vein blood vessel, enhancement mode filtering maintains that low-pass information is relatively less, and the frequency of sample boundary is low, effectively outstanding thin vein blood vessel sample.
Then, carry out the binary conversion treatment of vein sample synthesized image (GPi), convert to two-value picture (BPi) (S304).Now, for obtain two-value picture (BPi) use binary conversion treatment time, no matter B&W distinguishes the pixel met, in this, as the method for decision threshold, two kinds of methods can be considered.A kind of be local shade mean value use as threshold value, another kind is with 7 bit field numerical value definite thresholds.Attempt application these two kinds of methods, obtain as a result, with 7 bit field numerical value definite threshold time, more correct vein sample information can be obtained.Be designed to × kernel program, be expected to the size calculating the spatial mask using binary conversion treatment.
Finally, comprise with the two-value picture (BPi) that binary conversion treatment obtains the assorted shadow that the back of the hand fine hair of user, bending, subcutaneous layer of fat etc. produce.For eliminating this assorted shadow, as elimination impurity filtering process, using median (Median) filtering of 3 × 3 kernel processes programs, making vein sample picture (VPi) (S305) of the back of the hand.
As mentioned above, with original vein example extraction algorithms, do not consider the thickness of the vein blood vessel sample that will extract, by the characteristic of single filtering, extract the loss of vein sample and generating portion vein sample link information (vein-pattern connectivity information), the vein example extraction algorithms of bank of filters improves the loss of section of vein sample link information, for contraction and swelling that vein sample occurs because of external environment change, provide more stable algorithm performance.
It is the vein example extraction algorithms by bank of filters, the picture of outstanding vein sample process result, a () obtains vein sample with general mode filtering and enhancement mode filtering, and the vein sample synthesized image (GPi) of synthesis, b () is by binary conversion treatment, the two-value picture (BPi) of black and white picture, (c) is the deimpurity vein sample picture (VPi) that disappears.
To be compared with (b) by (a) and following result can be obtained: to the certain fields in ROI field, attempt comparing its vein sample extracted, filtering is given prominence to original vein sample, to the vein sample process result extracting thick vein blood vessel be, lose the internal information of vein sample, the sample connectivity of loss section of vein sample, but be by the result of general mode filtering of the present invention, effectively supplement the information revising low bandwidth, maintain the sample connectivity of vein sample.
But, can be found out by general mode filtering, the blood vessel that generation is shunk for external temperature is low and thin thin vein blood vessel, when extracting vein sample, the sample connectivity of part can be lost.For this loss of supplementary correction, use the enhancement mode filtering in the vein example extraction algorithms of bank of filters.By enhancement mode filtering, reduce the interests meeting the low bandwidth of background, the slight change of outstanding sample, supplements the loss of the part sample connectivity revising thin thin vein sample.
Secondly, when user's finger is positioned over stable handle, obtain numeral finger picture (Fdi) by video camera shooting finger picture (Fi), analyze numeral finger picture, carry out position correction, react this process (S0).
First, display position correction algorithm in detail below, according to position correction algorithm, calculates in numeral finger picture (Fdi) and rotates compensating value (R) and mobile compensating value (S) (S1).According to rotation compensating value (R) and mobile compensating value (S), the vein sample of vein sample picture (VPi) can be compared, with the vein sample canonical picture logged in advance, make vein sample correction picture (mVPi) (S2) after supplementary correction position.
For middle finger being put into the middle finger insertion part of stable handle, when raising hand, shooting forefinger, middle finger, the third finger.From the finger picture (Fi) photographed, utilize the intersection point of the left and right of stable handle, the intersection point of forefinger and middle finger, middle finger and the third finger, detect hand and be positioned over which direction.
By position correction algorithm, first, the boundary line (S51) between stable handle 6 and finger is extracted.
By the extraction process of boundary line, during rotation angle prediction, ask 6 pixel coordinates of needs, but the average filter of the low-pass filtering that the arranged type of 3 × 3 sizes is formed (Low pass Filter), and it is applied to original image.In low-pass filtering, average filter is the one of most representative.Low-pass filtering can not see overall picture brightly, but when picture frequency rate is passed through, compared with the surrounding pixel of high-frequency picture part, contributes to eliminating the sudden change of numerical value and the part of dressed to eh nines.Utilize the advantage of average filter to be in picture processing procedure, reduce the assorted shadow as clutter spot (Spot Noise) and impulsive noise (Salt & Pepper Noise).
The example application of display average filter, average filter is that arrangement value all becomes 1.Original image, by the average filter of the arrangement of 3 × 3 sizes, take center pixel as standard mobile 1 pixel at every turn, from left to right mobile from top to bottom again, the content then indicated by repeatable operation calculating.After the pixel value of superposition image is multiplied with the arrangement value of same position in average filter, ask total value and mean value, and on mean value centre pixel value opposition position.Mean value prevents on the position of the pixel value 3 of original picture.
The application process of above-mentioned explanation average filter is in embodiments of the invention, and utilizes being implemented as disposal route of suitable shade (mask).
On the picture obtained in application of low-pass, when extracting in test most suitable edge, application shade, extracts the boundary line between stable handle and finger.The shade applied for extracting edge has Sobel shade (Sobel Mask), Kirsh shade (Kirsh Mask), Edge shade (Edge Mask) etc., often uses Edge shade at large.But, when extracting edge, outside the edge of requirement, together try to achieve other object etc., because of the thickness that edge is not fixed, in leaching process, correctly can not ask for the pixel coordinate of needs.
According to position correction algorithm, because only extract correct location boundary line, becoming as processing procedure of next is more prone to, and in addition, reduces the process eliminating assorted shadow and impurity, can improve the arrangement travelling speed of algorithm.
Then, for extracting outer linea angulata (S52), extract the angle (S53) between finger, rotate correction, as center when rotating as rotary middle point, after asking this rotary middle point, by extracting boundary line and outer linea angulata, obtain coordinate and the rotary middle point of pixel etc., select the pixel coordinate of its integer form, save (S55) as positional information.
By position correction algorithm, the pixel of requirement is in whole 6 pixels are individual, and because whole pixel is positioned on position, boundary line, utilize the pixel left side, prediction rotates correction angle.In addition, in 6 pixels, specify 1 pixel, rotate supplement revise time, as hand back vein sample rotary middle point and used.
Boundary line between stable handle and finger is formed with thick line, for asking pixel coordinate, needs the manifolding of graph thinning process and outer linea angulata.Graph thinning process normally utilizes bone (Skeleton) as process skill and technique, and this only stays the central point of word and object, makes the process of a line.
For making the crooked position on the upside of the boundary line of extraction (Bay) consistent with finger diplomacy, compared with graph thinning process, with position correction algorithm, on the manifolding relatively less boundary line of calculated amount the exterior angle of crooked position, the outer linea angulata of boundary line is made carbon copies in application.In addition, for the boundary line of extracting, pixel coordinate can estimate the numerical value of the coordinate system of picture left side epimere, is applied to estimation coordinate figure.
In the outer linea angulata extracted, ask pixel (C) bottom and the pixel (A) of the top and pixel (B) about it.From in 3 pixels that 1 boundary line is extracted, (Fdi) is inner photographs 2 boundary lines for a numeral finger picture, i.e. 2 boundary lines of the intersection point of the intersection point of forefinger and middle finger, middle finger and the third finger, thus can in the hope of pixel be whole 6 pixels.For prediction rotates the rotation angle of correction, correction algorithm in position utilizes the angle between finger, and the angle between finger is 3 pixels obtained on every bar boundary line, utilizes 3 pixels and the angle asked for.
3 points (A, B, C) are the outer linea angulata of manifolding boundary line and 3 pixels obtaining.In these pixels, suppose to do perpendicular line centered by pixel (C) bottom, line segment is connected into pixel (A) topmost, pixel (B) and pixel bottom (C), form angle θ 1 and θ r respectively, bring the size that small tenon numerical value 1 asks angle into, by the distance that numerical value 1, ak is x-axis direction between 2 pixels, bk is the range difference of dihedral pixel on y-axis direction.According to formula (1), merge 2 angles of trying to achieve, finally ask the angle theta 1 between finger, θ 2.
Secondly, when asking two pixels bottom on boundary line, one is select rounded coordinate, asks rotary middle point.For maximally reducing the error that rotary middle point has, in the boundary line between the finger be separated a little from stable handle, the pixel of rotary middle point is the pixel bottom expecting to select it to extract.This is in the characteristic of hand, during contact object, considers that shape in one's hands is easily out of shape, crooked position between the finger particularly touching stable handle, can be different from the shape that self-assembling formation is bending.
Position correction algorithm is after the positional information of the data of preserving integer form, in finger field, ask data coordinates, rotates correction angle with the prediction of this data coordinates.Now, rotate correction angle and be divided into 1 time and 2 times, all solve 2 times.Utilize the rotation correction angle of so trying to achieve, carry out rotating supplementing and revise.
Be be applied in embodiments of the invention, show the forecasting process that 1 time rotates correction angle.1 time rotates correction angle is all rotate at the back of the hand that can supplement correction, asks the process of its correction angle; Rotate back to swing angle degree be from the finger picture logged in and the finger that compares as inner, utilize 4 pixel coordinates bottom, the angle of trying to achieve.
Pixel (A) and pixel (B) they are from the finger picture logged in, the pixel bottom of trying to achieve, pixel (A ') and pixel (B ') be from order to contrast, in the finger picture of shooting, the pixel bottom of trying to achieve.
The coordinate figure of this pixel (A, B, A ', B ') is for the data with integer form are saved, and is called data, the pixel coordinate value easily asked.In 4 pixels, every 2 pixels, are connected to the pixel that same finger is tried to achieve, and ask any line segment a, b.
At this, for compared with the finger picture logged in, 2 line segments of trying to achieve in the finger picture of shooting must intersect, angulation between 2 line segments.Now, intersection point is, the rotary middle point determined in the finger picture that logs in is standard, for contrast and the finger taken are as in line segment, the pixel being positioned at rotary middle point homonymy moved to rotary middle point.In, pixel (B) becomes the rotary middle point logged in picture, for contrast and the finger taken are as in line segment, becomes pixel (B ') with the pixel of pixel (B) same lateral position.
In the finger picture taken for contrast, for doing intersection point, by the pixel outside intersection point pixel, be applied to the distance length of movement, same distance.So, the same with θ 3, with 1 line segment for standard, ask the relative dip angle of all the other line segments.
At this, θ 3 with log in the back of the hand picture (official portrait) compared with, for contrast and take the back of the hand picture display relative dip angle.That is, θ 3 becomes rotation correction angle, and the back of the hand picture of rotation, this angle of inclination, correct relative angle of inclination.Rotate correction for 2 times to be used for supplementing correction, when between finger, crooked position touches stable handle 6 at full tilt, according to the rotation of the back of the hand self and the deliberate action of the back of the hand, between finger, the size of angle changes, even without this change, for make integral inclined, ask and rotate correction angle, the angle between finger can be utilized, carry out supplementing and revise.
Show the forecasting process that 2 times rotate correction angle.Extract boundary line and ask for pixel, according to this pixel, the angle between display finger.θ 1 logs in the angle in picture between the finger that obtains, and θ 1 ' is the angle in contrast and the finger picture taken between finger.The angle (θ 1, θ 1 ') of 2 angles is pictures different from each other, but be the angle between finger, identical position obtained, asking the difference of the angle in the angle in the picture of login between the finger that obtains and the finger picture taken to contrast between finger, asking the angle rotating corrections for 2 times.
Pixel (B) and pixel (B ') are pixels bottom on boundary line.In addition, the dotted line distinguishing angle between finger is the center line of angle between finger.
The straight line supposed by pixel (B) and pixel (B ') is standard, and this center line and pixel (B) and pixel (B ') intersect, ask the angle of inclination.This center line is with the straight line of hypothesis for standard, and if any field, left and right, then there is positive value in field, left side, and there is cloudy value in field, right side.
As above method tries to achieve the angle of the centerline dip of θ 1 and θ 1 ', use this angle, the angle θ 1 between finger is asked in the picture logged in, with the center line of angle θ 1 for standard, in the finger picture taken for contrast, with the angle of inclination of the center line of the angle θ 1 ' between finger, ask relative angle of inclination (θ 1-θ 1 ').
With said method, for the angle logging in picture between all the other fingers, ask the relative dip angle in shooting picture.Calculate the average of 2 relative dip angle so obtained, finally rotate correction angle as 2 times.
In last process, with correct probability match algorithm (PMA:Probability Matching Algorithm), compared with the authentication information logged in authentication information database, carry out certification (S0).
First, from authentication information database, read the data (S1) of the most suitable distribution characteristics of vein sample.The data of the most suitable distribution characteristics of vein sample comprise branching characteristics (VPBC:Vein Pattern Branch Characteristic) and the branch weight factor (BPEF:Branch Point Weighting Factor) of vein sample.
Then, in vein sample correction picture (mVPi), using each take-off point of vein sample as summit, make polygon.So, this polygon is overlapped with summit, compares and judge its degree of integration.
The result integrated is the fiducial value obtained, if unanimously, as the standard value of certification, can judge whether it meets standard value (S504), and not meeting is forbid certification (S505), satisfied then allow certification (S506).
Hand back vein specimen discerning algorithm
At this, explain that explanation utilizes hand back vein specimen discerning algorithm of the present invention in detail.
The branching characteristics of the back of the hand vascular venous comprises the sample length in take-off point, Branch Angle, take-off point.Utilize the branching characteristic of the vein sample so extracted, identify personal identification.In vein sample branching characteristics, the most basic branching characteristic is on the take-off point of vein sample, first calculates the part sample connectivity of losing in vein sample, during practice with original vein example extraction algorithms, to the change of external environment condition, the change etc. of input picture, reaction sensitivity very.In other words, in order to apply even weighting factor (Uniform Weighting Factor), identify the distribution character of hand back vein sample, as the physiology recognition system of essence, the shortcoming of vein example extraction algorithms is originally that stability declines significantly.
Such as, user's Reusability system, causes the input change of picture, the change etc. of external environment condition, and the vein sample take-off point that extract a loss part is the main cause causing algorithm performance low.Because the branching characteristic of the vein sample extracted is also few, in order to obtain certain identification, the loss of vein sample take-off point becomes the main cause of left and right recognition performance.
The branching characteristic of the same people's vein sample obtained.A () and (b) is in different time, input same picture result.Display 8 take-off points in (a), display 14 take-off points in (b).If this take-off point is employed as branching characteristic, at the cognitive phase of user, no matter whether be same people by mistake, others can be thought.
A () and (b) is in different time, input same picture result.Change in location is there is in user because rotating and move, supplementing and revising the method for its change in location is the position utilizing 2 take-off points (take-off point that the take-off point of top, left side, right side are following), supplement and revise its position, but the most basic branching characteristic is take-off point, due to loss element branches point, different results can be shown.
In embodiments of the invention, for the vein sample branching characteristics that everyone has, application confidence level, the most stable VPBP(Vein-Pattern Branch-Point to be extracted) for everyone, by the change of external environment condition, the slight change etc. of input picture, the sensitivity of generation system and the minimum change in location of user's the back of the hand, and rotate and move, hand back vein specimen discerning algorithm can improve the stability of system identification performance.
The recognizer of hand back vein sample was made up of two stages, experimental stage and Main Stage respectively, experimental stage is on the branching characteristics for everyone, the confidence level of application branching characteristics, Main Stage obtains confidence level when being applied in experimental stage, everyone log-on message of auto-changing, application characteristic essential factor comparison algorithm (FVMP:Feature Vector Matching Algorithm).
In experimental stage, according to the extraction algorithm of vein sample, use the vein sample picture extracted, by direct comparison algorithm (DMA:Direct Matching Algorithm), determine whether user.DMA is by branching characteristics, in authentication information database, judges the direct similarity (Correlation) of official portrait and the vein sample picture logged in, judges forbid or allow person identification.If it is determined that for being he or she, according to the confidence level (PMA) judging take-off point, user's Reusability system, the final branching characteristics detecting everyone the most stable vein sample, to various branching characteristics, determines the tap weights factor (BPWF).
In Main Stage, be applied in the branching characteristics that experimental stage obtains confidence level, automatically convert everyone log-on message to, application FVMA, identifies whether as user.In identifying, FVMA is by polygonal feature and branching characteristics probability, determines whether user.
Application confidence level obtains everyone branching characteristics, applies everyone branching characteristics, can show stable recognition performance, user is rotated and the correction of movement, with the most stable VPBP for standard carries out additional modifications, keep on top, improve the performance of system and the convenience of user.
But, during Reusability individual hand back vein sample distribution system, application branching characteristics, but input the change of picture because individual external environment condition and system repeatedly use, extract the branching characteristics that individual is most suitable, demonstrate more stable system performance.
The log-on message of user is the initial test stage in the system of use, from the user of input information, preserve the prototype of the vein sample extracted, the log-on message of user is familiar with according to DMA, but due to Reusability system, automatically convert probability match algorithm (Probability Matching Algorithm) to, meanwhile, extract the branching characteristics of the most suitable vein sample of user, use the most suitable BPWF(S61 ~ S64 extracted).Result is in the log-on message of user, only preserves the vein sample distribution characteristic (S65) extracted from user.
Extract take-off point, comprise the field of extracting take-off point by the method, application Thinning algorithm.
Apply in the process of above-mentioned Thinning algorithm, extract take-off point there is minimum change in location in position, show the field that it allows change, even if input the original image of same people, changed by the thickness that vein blood vessel sample is minimum and produce change in location, the take-off point extracted in permission field, is on the middle position in permission field, determines the position of take-off point.
A () is the most suitable branching characteristics be kept in database, the take-off point connected without line presents low reliability.In other words, because of the change of external environment condition and the Reusability of system, owing to inputting the little change of picture, take-off point performance losses probability is high, and such take-off point applies low weight factor (Low Weighting Factor).On the one hand, reliability degree is shown with the take-off point that line connects high.Namely to the change of external environment condition and input as minimum change, mean the VPBP that reliability is high, stable, apply higher weight factor (Higher Weighting Factor) and identify.
B () is by the rotation of user's hand and the mobile result obtained, the branching characteristic of database in (c) display (a), and (b) is rotated to the method for correction and mobile correction.By the recognizer of hand back vein sample, connect the most stable VPBP, form polygon, according to the position that polygon is the most consistent, rotation and mobile implementation are supplemented and revises.Result is, is obtained by the recognizer of hand back vein sample everyone branching characteristics, application fetches confidence level on respective branching characteristics, and at cognitive phase, application the most stable VPBP and optimal BPWP, can obtain stable recognition performance.
Be applied in example of the invention process, by the recognizer of hand back vein sample, use the most stable VPBP, carry out rotating and the correction process of movement.A VPBP that () and (b) extract, (c) is the most stable VPBP of the selection in (a) and (b), and it is overlapping, the result that display is overlapping.
Hand back vein specimen discerning algorithm uses probability match algorithm, carries out person identification, at this cognitive phase, finally judges to confirm user whether as me by FVMA.
At cognitive phase, first link the most stable VPBP in database and the polygon that makes and, at the polygon made as the VPBP obtained of input now, during coupling two polygonal summits, the length between the degree of integration on each summit of judgement identification, the degree of integration of polygon vertex angle, summit and the degree of integration of area of a polygon.
Secondly, at cognitive phase, by confidence level, apply the most suitable BPWF of each branching characteristics, judge to identify the degree of integration to branching characteristics (length between the Branch Angle between take-off point, take-off point, the branch's number between take-off point, take-off point).
In the database of vein blood vessel identification, authorize the login number that every user fixes, continue the data that upgrading is relevant to its vein sample branchedness, preserve the branching characteristics relevant to the vein sample of user and official portrait.
Certainly; more than illustrate and be only preferred embodiment of the present invention; the present invention is not limited to enumerate above-described embodiment; should be noted that; any those of ordinary skill in the art are under the instruction of this instructions; made all equivalently to substitute, obvious form of distortion, within the essential scope all dropping on this instructions, protection of the present invention ought to be subject to.
Claims (10)
1. a biological sign authentication method, it comprises:
Two firm handles are found at the middle part of horizontal axis, middle finger is inserted in the middle of described handle, at this moment obtained middle finger, forefinger, the digital picture in finger and the back of the hand certification end and the server that send the digital picture of the back of the hand of the contact portion of little finger of toe and the back of the hand, all be connected with communication line, described hand back vein certification is exactly the back of the hand Verification System.
2. biological sign authentication method according to claim 1, is characterized in that,
Described server:
Temporary storage
Middle finger inserts in the middle of firm handle by authenticator, with take the hand back vein branch point obtained in advance and be connected, middle finger is inserted the second reference edge line at the second bending range edge that the first reference edge line at the edge of the first bending range that middle finger that handle photographs in advance is connected with little finger of toe and middle finger are connected with forefinger by the data comprising polygonal optimal characteristic distributions and described authenticator, and the two is all as the database that the memory authentication information of finger edge reference information is special;
The digital picture in finger and the digital picture of the back of the hand of described the back of the hand certification end all can be preserved in described having in the shooting storehouse of temporary storage server;
Be saved in the digital picture of the back of the hand of memory at the server of described temporary storage, for other part of vein case that certification is very important, ROI picture is it can be used as to extract, and thick vein case and thin vein case are extracted synthesize, by binaryzation, removing interference vein synthesis case generates ROI picture and extracts, and setting position correction amendment part and vein case compatible portion.
3. biological sign authentication method according to claim 2, is characterized in that, concrete also the comprising of above-mentioned steps:
Position correction amendment portion will arrange following step:
In the digital picture in described finger, the method extracted as finger edge picture of first bending range and the second bending range is for the first bending range of described finger edge picture, in addition transparent processing, obtain around middle finger with little finger of toe around first edge line at edge and the step of middle finger and forefinger edge second edge line:
And/or the focus comprised in the first described edge line combines with the focus comprised in the second edge line and obtains the first straight line, the navigate benchmark focus that comprises of the benchmark focus that first reference edge line comprises and the second reference edge line combines and obtains the second straight line, combined by one or two straight lines and ask for the first relative angle, Here it is obtains the step that angle is revised in the first correction;
The relative angle of the reference opening amount of the first reference edge line that the first edge line aperture comprised by described finger edge picture and finger edge reference information are comprised obtains the second relative angle, in like manner, the relative angle of the reference opening amount of the second edge line aperture that finger edge picture comprises and the second reference edge line that finger edge reference information comprises obtains third phase to angle, obtains the second correction amendment angle by average second relative angle and third phase to angle;
Described amendment in angle in the first correction carries out vein as correction, and carries out correct vein as correction in the second correction amendment angle;
And/or described vein is as matching part:
Obtain each branch point of the correction picture of described vein case, multiple branch point is bonded polygon, by this polygon compared with the suitableeest described distribution character data information, judge whether consistent;
It is described that can to show the back of the hand certification end announcer to what represent the feature of the method for result be exactly the back of the hand Verification System.
4. biological sign authentication method according to claim 3, it is characterized in that, concrete also the comprising of above-mentioned steps: the special shooting in described finger camera part and the back of the hand special shooting camera part are one of features of the back of the hand Verification System of the CCD responded to visible rays and infrared spectral range or the application Xiang Yizhong photographing feature with MOS.
5. biological sign authentication method according to claim 3, is characterized in that, concrete also the comprising of above-mentioned steps:
The suitableeest described distribution character packet contains the difference angle of the hand back vein picture in branch point, branch point number, branch point pixel coordinate;
Described vein is as matching part
Obtain the pixel coordinate of described vein as the branch point of the vein picture of correction picture, branch point number, difference angle, judge whether consistent.
6. biological sign authentication method according to claim 3, is characterized in that, concrete also the comprising of above-mentioned steps: the vein picture of the suitableeest described distribution character data comprises thick vein picture and thin vein picture.
7. biological sign authentication method according to claim 3, is characterized in that, concrete also the comprising of above-mentioned steps: the suitableeest distribution character data are relevant to authentication password also can be saved.
8. biological sign authentication method according to claim 3, is characterized in that, utilizes probability match algorithm to calculate the threshold value of ROI area occupied ratio in described the back of the hand preimage.
9. biological sign authentication method according to claim 3, it is characterized in that, first edge line of described finger edge picture, the aperture a k of the second edge line is positioned between two pixels to represent x-axis distance, b k is the range difference of the pixel of y-axis direction angulation, utilize following mathematical formulae to be merged by two the θ k tried to achieve, this is the feature of the back of the hand Verification System of application item one ~ 6, can be obtained by formula (1):
Θ k=tan a k/b k formula (1).
10. biological sign authentication method according to claim 3, it is characterized in that, concrete also the comprising of above-mentioned steps: the suitableeest distribution character packet of described authentication information private database contains the position on the described each summit of benchmark polygon, the difference angle of the straight line that each summit sends, length between summit, polygonal area;
Described vein is as matching part
For the position on each summit, the difference angle of the straight line sent from each summit, the length between summit, polygonal area, judges its degree of integration of understanding;
Described vein is as matching part
To the branch point application branch point weight of described vein picture of trying to achieve, to the position on each summit, the difference angle of the straight line that this each summit sends, the length between summit, difference number understanding and sound judgment degree of integration;
Described the back of the hand certification end
Prepare finger photographic camera part and the back of the hand photographic camera part, these cameras are connected with microswitch, take pictures in the lower part of this microswitch.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510136324.2A CN104809428A (en) | 2015-03-27 | 2015-03-27 | Biont sign authentication method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510136324.2A CN104809428A (en) | 2015-03-27 | 2015-03-27 | Biont sign authentication method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104809428A true CN104809428A (en) | 2015-07-29 |
Family
ID=53694239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510136324.2A Pending CN104809428A (en) | 2015-03-27 | 2015-03-27 | Biont sign authentication method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104809428A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056054A (en) * | 2016-05-24 | 2016-10-26 | 青岛海信移动通信技术股份有限公司 | Fingerprint identification method and terminal |
| CN109478322A (en) * | 2016-07-12 | 2019-03-15 | 株式会社生物电子 | Person authentication device |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008040619A (en) * | 2006-08-02 | 2008-02-21 | Tekusupiaa:Kk | Back of hand authentication system, and back of hand authentication method |
-
2015
- 2015-03-27 CN CN201510136324.2A patent/CN104809428A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008040619A (en) * | 2006-08-02 | 2008-02-21 | Tekusupiaa:Kk | Back of hand authentication system, and back of hand authentication method |
Non-Patent Citations (4)
| Title |
|---|
| CHIH-LUNG LIN等: "Biometric verification using thermal images of palm-dorsa vein patterns", 《IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY》 * |
| 张会林等: "人体手背静脉血管图像增强处理算法研究", 《仪器仪表学报》 * |
| 潘晓苹等: "手背静脉图像ROI提取算法研究", 《信息通信》 * |
| 许明: "基于手背静脉血管的身份识别技术研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106056054A (en) * | 2016-05-24 | 2016-10-26 | 青岛海信移动通信技术股份有限公司 | Fingerprint identification method and terminal |
| CN109478322A (en) * | 2016-07-12 | 2019-03-15 | 株式会社生物电子 | Person authentication device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6774580B2 (en) | Biometric template security and key generation | |
| CA2976049C (en) | Systems and methods for performing fingerprint based user authentication using imagery captured using mobile devices | |
| US9922238B2 (en) | Apparatuses, systems, and methods for confirming identity | |
| US11263432B2 (en) | Systems and methods for performing fingerprint based user authentication using imagery captured using mobile devices | |
| EP1423821B1 (en) | Method and apparatus for checking a person's identity, where a system of coordinates, constant to the fingerprint, is the reference | |
| CN110326001B (en) | System and method for performing fingerprint-based user authentication using images captured with a mobile device | |
| CN102073843B (en) | Non-contact rapid hand multimodal information fusion identification method | |
| EP2528018B1 (en) | Biometric authentication device and biometric authentication method | |
| US20180300571A1 (en) | Finger vein identification method and device | |
| US20060023921A1 (en) | Authentication apparatus, verification method and verification apparatus | |
| KR20110127270A (en) | A method for fingerprint template synthesis and fingerprint mosaicing using a point matching algorithm | |
| KR101025666B1 (en) | Method and apparatus of extracting finger vein feature point | |
| JP4809155B2 (en) | Back of hand authentication system and back of hand authentication method | |
| JP2005222419A (en) | Biometrics information collation system | |
| JP2008287433A (en) | Vein pattern management system, vein pattern registering device, vein pattern authentication device, vein pattern registering method, vein pattern authentication method, program, and vein data structure | |
| JP5480532B2 (en) | Image processing apparatus, image processing method, and program for causing computer to execute the method | |
| KR101774151B1 (en) | Method and apparatus of enhancing iris recognition security using distributed iris template storage and matching | |
| JP5377580B2 (en) | Authentication device for back of hand and authentication method for back of hand | |
| EP3214575B1 (en) | Biological image processing device, biological image processing method, and biological image processing program | |
| KR102138660B1 (en) | Combined authentication system using fingerprint and branched veins | |
| CN111460435A (en) | User registration method, verification method and registration device | |
| CN104809428A (en) | Biont sign authentication method | |
| US20040175023A1 (en) | Method and apparatus for checking a person's identity, where a system of coordinates, constant to the fingerprint, is the reference | |
| KR20140089277A (en) | Method for authenticating based on finger vein using SIFT keypoint | |
| Jagadeesh et al. | An conceptual view of an iris-biometric identification system canny edge detection techniques |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20160413 Address after: 130000, A6, building 4, North Lake Science Park, 3333 North Sheng street, hi tech North Zone, Jilin, Changchun Applicant after: Jilin lost biometric technology Co. Ltd. Address before: Xikang Road 130000 in Jilin Province, Chaoyang District City, No. 26 Changchun Applicant before: Bai Yu |
|
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150729 |