CN104252614A - SIFT algorithm-based two-generation identity card face comparison method - Google Patents
SIFT algorithm-based two-generation identity card face comparison method Download PDFInfo
- Publication number
- CN104252614A CN104252614A CN201310261876.7A CN201310261876A CN104252614A CN 104252614 A CN104252614 A CN 104252614A CN 201310261876 A CN201310261876 A CN 201310261876A CN 104252614 A CN104252614 A CN 104252614A
- Authority
- CN
- China
- Prior art keywords
- generation identity
- identity card
- sift algorithm
- key point
- comparison method
- 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)
Abstract
The invention belongs to the identity verification field and relates to an SIFT algorithm-based two-generation identity card face comparison method. The SIFT algorithm-based two-generation identity card face comparison method includes the following steps of: face information is obtained through a document identification instrument and an image pickup device, and the face information is processed so as to be converted into a PGM gray-scale image to be processed by an SIFT algorithm, and the PGM gray-scale image is saved; feature point search and matching are performed through utilizing the SIFT algorithm; and whether the holder of a two-generation identity card is the owner of the two-generation identity card can be judged according to matching degree, and so that a recognition effect can be realized. According to the SIFT algorithm-based two-generation identity card face comparison method of the invention, a network database and training are not required, and only two-generation identity cards and portrait information are required so as to be compared, and therefore, timeliness can be ensured. The SIFT algorithm-based two-generation identity card face comparison method has high tolerance to light, noise, micro perspective variation, and has high detection rates of partial object obstruction, can calculate positions and orientations only with more than three SIFT object features; and the SIFT algorithm-based two-generation identity card face comparison method does not need to be based on Internet data, and only needs two-generation identity cards and the holders of the two-generation identity cards.
Description
Technical field
The invention belongs to authentication field, relate more specifically to a kind of second generation identity card face comparison method based on scale invariant feature matching algorithm.
Background technology
In these informationalized epoch, the personal information of citizen of the People's Republic of China is converted into data and is stored in second generation identity card and facilitates our daily life to use, the highly integrated usable range of a slight I.D. that makes of data becomes increasing, exactly because this feature, once the loss of I.D. is to may by other people illegal use while the life inconvenience of oneself, this brings harm to a certain degree to our living safety.
Except storing our information such as name, birthdate, registered permanent residence location, identification card number, face information is also had in second generation identity card, this has just had the possibility of automatic identification and comparison to the people holding I.D., now use in many instances second generation identity card only just identify I.D. and and nonrecognition possessor, so we think that the picture that should obtain end user while storing I.D. use information is compared and stores simultaneously, this makes the full-automation of some business also bring possibility.
Scale invariant feature conversion (Scale-invariant feature transform or SIFT) is that a kind of algorithm of computer vision is used for detecting and the locality characteristic described in image, it finds extreme point in space scale, and extract its position, yardstick, rotational invariants, this algorithm delivered in 1999 by David Lowe, within 2004, improves and sums up.
The research by recognition of face image procossing possessor has been there is in prior art, such as CN 201110070277.8 A and CN 201120461613.7 U, but because these need through a large amount of comparings, process slowly, cannot fast verification, the requirement of above technology to image is higher in addition, and the scene of authentication is often by the impact of environment.
Summary of the invention
1, object of the present invention.
The object of the invention is to prevent other people from falsely using others' I.D., and the method for a kind of possessor proposed and ID card verification, solve data processing in prior art loaded down with trivial details, the problem that verifying speed is slow, simultaneously can the data of effective recognition image, the accuracy of raising authentication.
2, the technical solution adopted in the present invention.
China second-generation identity card face comparison method based on SIFT algorithm of the present invention, carry out according to following steps:
Step one camera shooting face information, the face picture in certificate identifier identification chip;
Step 2 is by sharpening enhancement after the image denoising of acquisition;
Image after denoising in described step one and the face picture in chip are changed into PGM picture by step 3;
Step 4 uses SIFT algorithm to carry out unique point and searches and mate substantially, is divided into following steps:
A. metric space extremum extracting: search for the picture position on all yardsticks, identifies the potential point of interest for yardstick and invariable rotary by gaussian derivative function;
B. key point location: on the position of each candidate, the model meticulous by matching determines position and yardstick, and the selection gist of key point is in their degree of stability;
C. direction is determined: based on the gradient direction of image local, distribute to one or more direction, each key point position, all operations to view data below all convert relative to the direction of key point, yardstick and position, thus provide the unchangeability for these conversion;
D. key point describes: in the neighborhood around each key point, the gradient of measurement image local on the yardstick selected, and these gradients are transformed into a kind of expression, this distortion and the illumination variation representing the local shape that permission is larger;
E. key point coupling: the key point of carrying out two figure searching key point is mated, and obtains corresponding matching degree information.
3. beneficial effect of the present invention.
(1) do not need network data base and training, only need China second-generation identity card and personage's head image information to compare;
(2) substantially can think real-time, the SIFT efficiency of algorithm be optimized reaches real-time effect;
(3) tolerance for light, the change of noise, slightly visual angle is also quite high, makes noise to a certain degree not affect matching result with distortion;
(4) the detecting rate of covering for fractional object is higher, even only needs the SIFT object features of more than 3 to be just enough to calculate position and orientation;
(5) do not need based on internet data, only need China second-generation identity card and holder.
Accompanying drawing explanation
Fig. 1 is China second-generation identity card face alignment system construction drawing.
Fig. 2 is the China second-generation identity card face alignment process flow diagram based on SIFT algorithm.
Fig. 3 is metric space---gaussian pyramid.
Fig. 4 is difference of Gaussian pyramid.
Embodiment
Embodiment
Certificate identifier adopts IDIWAY credential reading instrument IDR-A2, and for obtaining face picture information in China second-generation identity card chip, this equipment carries scanner functions, a compatible generation ID face picture acquisition of information; Head image information Real-time Obtaining equipment adopts the convenient acquisition such as general mobile phone, camera, camera head image information more clearly; Treatment facility adopts computing machine.
System construction drawing as shown in Figure 1, is made up of certificate identifier, picture pick-up device and computing machine, and certificate identifier is in order to obtain head portrait picture information in China second-generation identity card chip, and the PGM gray-scale map be converted to after acquisition in order to SIFT algorithm process is preserved, picture pick-up device get convenient carry out mutual equipment with computing machine can (providing API to carry out programming makes whole system integrated), this equipment is in order to obtain the real-time face image information of holder, gray-scale map conversion is carried out in image transmitting after acquisition to computing machine, and eliminate the salt-pepper noise because insufficient light causes with medium filtering, picture is carried out sharpening enhancement (SIFT algorithm can inquire more key point in sharp-featured situation) again, finally the picture handled well is converted into the PGM gray-scale map being convenient to SIFT algorithm process, wait for that computing machine processes pictorial information.
The operational scheme of entire system as shown in Figure 2, after obtaining the PGM gray-scale map from certificate identifier and picture pick-up device, both processing modes are identical, all use SIFT algorithm to carry out searching of key point and the keypoints that two figure obtain is mated, to obtain the reliable judgement that two figure are same people.Be described for SIFT algorithm process process below:
SIFT algorithm searches key point on different metric spaces, and the acquisition of metric space needs to use Gaussian Blur to realize, Gaussian Blur is a kind of image filter, it uses normal distribution (Gaussian function) to calculate Fuzzy Template, and use this template and original image to do convolution algorithm, reach the object of blurred picture.N dimension space normal distribution equation is:
Metric space uses gaussian pyramid to represent when realizing, the structure of gaussian pyramid is divided into two parts:
1. pair image does the Gaussian Blur of different scale;
2. pair image does down-sampled (dot interlace sampling).
As shown in Figure 3, the pyramid model of image refers to the structure of gaussian pyramid, by continuous for original image depression of order sampling, obtains a series of image not of uniform size, descending, the tower-like model formed from top to bottom.Original image is the ground floor of gold tower, and down-sampled obtained new images is pyramidal one deck (every layer of image) at every turn, and each pyramid is n layer altogether.The pyramidal number of plies determines jointly according to the size of the original size of image and tower top image, and its computing formula is as follows:
Obtain the pyramidal generation of difference of Gaussian (as shown in Figure 4) according to gaussian pyramid, when actual computation, use gaussian pyramid often adjacent upper and lower two-layer image subtraction in group, obtain difference of Gaussian image, wait pending extremum extracting.
Key point is made up of the Local Extremum in DOG space, tentatively detecting by having compared between each DoG adjacent two layers image in same group of key point.In order to find DoG Function Extreme Value point, the consecutive point that each pixel will be all with it compare, and see that it is whether large or little than the consecutive point of its image area and scale domain.The removing of each metric space topmost with nethermost two-layer outside, the check point of intermediate image and it with 8 consecutive point of yardstick and 9 × 2 points corresponding to neighbouring yardstick totally 26 points compare, to guarantee all extreme point to be detected at metric space and two dimensional image space.
The extreme point that above method detects is the extreme point of discrete space, position and the yardstick of key point is accurately determined again by the three-dimensional quadratic function of matching, remove key point and the unstable skirt response point (because DoG operator can produce stronger skirt response) of low contrast, to strengthen coupling stability, to improve noise resisting ability simultaneously.
In order to make descriptor have rotational invariance, need to utilize the local feature of image for distribute a reference direction to each key point.The method of image gradient is used to ask for the stabilising direction of partial structurtes.For the key point point detected in DOG pyramid, gather gradient and the directional spreding feature of pixel in the gaussian pyramid image 3 σ neighborhood window of its place.Modulus value and the direction of gradient are as follows:
By above step, for each key point, have three information: position, yardstick and direction.Next set up a descriptor for each key point, with one group of vector, this key point is described out, make it not change, such as illumination variation, visual angle change etc. with various change.This descriptor not only comprises key point, also comprise to its contributive pixel around key point, and descriptor should have higher uniqueness, to improve the probability that unique point is correctly mated.Finally the key that two figure obtain is carried out matching ratio pair one by one, can judge that whether holder is the owner of I.D. according to the matching degree obtained.
The maximum feature of the present invention is that the unique point obtained in SIFT algorithm is the local feature of image, maintains the invariance for rotation, scaling, brightness change, to the stability that visual angle change, affined transformation, various noise also keep to a certain degree; Its uniqueness is good, informative, can mate fast and accurately in magnanimity property data base; Even if the volume of SIFT algorithm makes to only have several objects of minority also can produce a large amount of SIFT feature vectors; SIFT algorithm through optimizing can reach the requirement of real-time to a certain extent; And the extensibility of SIFT algorithm can be combined with other forms of proper vector easily.
Above-described embodiment does not limit the present invention in any way, and the technical scheme that the mode that every employing is equal to replacement or equivalent transformation obtains all drops in protection scope of the present invention.
Claims (3)
1., based on the China second-generation identity card face comparison method of SIFT algorithm, it is characterized in that carrying out according to following steps:
Step one camera shooting face information, the face picture in certificate identifier identification chip;
Step 2 is by sharpening enhancement after the image denoising of acquisition;
Image after denoising in described step one and the face picture in chip are changed into PGM picture by step 3;
Step 4 uses SIFT algorithm to carry out unique point and searches and mate substantially, is divided into following steps:
A. metric space extremum extracting: search for the picture position on all yardsticks, identifies the potential point of interest for yardstick and invariable rotary by gaussian derivative function;
B. key point location: on the position of each candidate, the model meticulous by matching determines position and yardstick, and the selection gist of key point is in their degree of stability;
C. direction is determined: based on the gradient direction of image local, distribute to one or more direction, each key point position, all operations to view data below all convert relative to the direction of key point, yardstick and position, thus provide the unchangeability for these conversion;
D. key point describes: in the neighborhood around each key point, the gradient of measurement image local on the yardstick selected, and these gradients are transformed into a kind of expression, this distortion and the illumination variation representing the local shape that permission is larger;
E. key point coupling: the key point of carrying out two figure searching key point is mated, and obtains corresponding matching degree information.
2. the China second-generation identity card face comparison method based on SIFT algorithm according to claim 1, is characterized in that: in described step one, certificate identifier adopts IDR-A2.
3. the China second-generation identity card face comparison method based on SIFT algorithm according to claim 1, is characterized in that: in described step 3, image adopts medium filtering denoising.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310261876.7A CN104252614A (en) | 2013-06-27 | 2013-06-27 | SIFT algorithm-based two-generation identity card face comparison method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310261876.7A CN104252614A (en) | 2013-06-27 | 2013-06-27 | SIFT algorithm-based two-generation identity card face comparison method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104252614A true CN104252614A (en) | 2014-12-31 |
Family
ID=52187495
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310261876.7A Pending CN104252614A (en) | 2013-06-27 | 2013-06-27 | SIFT algorithm-based two-generation identity card face comparison method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104252614A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107360371A (en) * | 2017-08-04 | 2017-11-17 | 上海斐讯数据通信技术有限公司 | A kind of automatic photographing method, server and Automatic camera |
CN110210341A (en) * | 2019-05-20 | 2019-09-06 | 深圳供电局有限公司 | Authentication ids method and its system, readable storage medium storing program for executing based on recognition of face |
CN110287883A (en) * | 2019-06-26 | 2019-09-27 | 山东浪潮人工智能研究院有限公司 | A method of recognition of face is carried out based on nearest neighbor distance ratio method is improved |
CN111985500A (en) * | 2020-07-28 | 2020-11-24 | 国网山东省电力公司禹城市供电公司 | Method, system and device for checking relay protection constant value input |
CN113516597A (en) * | 2021-05-19 | 2021-10-19 | 中国工商银行股份有限公司 | Image correction method and device and server |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6301370B1 (en) * | 1998-04-13 | 2001-10-09 | Eyematic Interfaces, Inc. | Face recognition from video images |
CN101510257A (en) * | 2009-03-31 | 2009-08-19 | 华为技术有限公司 | Human face similarity degree matching method and device |
CN102004911A (en) * | 2010-12-31 | 2011-04-06 | 上海全景数字技术有限公司 | Method for improving accuracy of face identification |
CN102622590A (en) * | 2012-03-13 | 2012-08-01 | 上海交通大学 | Identity recognition method based on face-fingerprint cooperation |
-
2013
- 2013-06-27 CN CN201310261876.7A patent/CN104252614A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6301370B1 (en) * | 1998-04-13 | 2001-10-09 | Eyematic Interfaces, Inc. | Face recognition from video images |
CN101510257A (en) * | 2009-03-31 | 2009-08-19 | 华为技术有限公司 | Human face similarity degree matching method and device |
CN102004911A (en) * | 2010-12-31 | 2011-04-06 | 上海全景数字技术有限公司 | Method for improving accuracy of face identification |
CN102622590A (en) * | 2012-03-13 | 2012-08-01 | 上海交通大学 | Identity recognition method based on face-fingerprint cooperation |
Non-Patent Citations (1)
Title |
---|
罗佳: "基于SIFT的人脸识别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107360371A (en) * | 2017-08-04 | 2017-11-17 | 上海斐讯数据通信技术有限公司 | A kind of automatic photographing method, server and Automatic camera |
CN110210341A (en) * | 2019-05-20 | 2019-09-06 | 深圳供电局有限公司 | Authentication ids method and its system, readable storage medium storing program for executing based on recognition of face |
CN110210341B (en) * | 2019-05-20 | 2022-12-06 | 深圳供电局有限公司 | Identity card authentication method based on face recognition, system thereof and readable storage medium |
CN110287883A (en) * | 2019-06-26 | 2019-09-27 | 山东浪潮人工智能研究院有限公司 | A method of recognition of face is carried out based on nearest neighbor distance ratio method is improved |
CN111985500A (en) * | 2020-07-28 | 2020-11-24 | 国网山东省电力公司禹城市供电公司 | Method, system and device for checking relay protection constant value input |
CN111985500B (en) * | 2020-07-28 | 2024-03-29 | 国网山东省电力公司禹城市供电公司 | Verification method, system and device for relay protection fixed value input |
CN113516597A (en) * | 2021-05-19 | 2021-10-19 | 中国工商银行股份有限公司 | Image correction method and device and server |
CN113516597B (en) * | 2021-05-19 | 2024-05-28 | 中国工商银行股份有限公司 | Image correction method, device and server |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Fei et al. | Feature extraction methods for palmprint recognition: A survey and evaluation | |
Zhao et al. | Unsupervised salience learning for person re-identification | |
Pflug et al. | Ear biometrics: a survey of detection, feature extraction and recognition methods | |
Vijayan et al. | Twins 3D face recognition challenge | |
Winston et al. | A comprehensive review on iris image-based biometric system | |
Fei et al. | Learning compact multifeature codes for palmprint recognition from a single training image per palm | |
Yap et al. | A comparative study of mobile-based landmark recognition techniques | |
CN104680131A (en) | Identity authentication method based on identity certificate information and human face multi-feature recognition | |
Bakshi et al. | A novel phase-intensive local pattern for periocular recognition under visible spectrum | |
CN104252614A (en) | SIFT algorithm-based two-generation identity card face comparison method | |
AU2011207120B2 (en) | Identifying matching images | |
CN110363179B (en) | Map acquisition method, map acquisition device, electronic equipment and storage medium | |
Soltanpour et al. | Multimodal 2D–3D face recognition using local descriptors: pyramidal shape map and structural context | |
CN111783805A (en) | Image retrieval method and device, electronic equipment and readable storage medium | |
Aoyama et al. | A contactless palmprint recognition algorithm for mobile phones | |
Kurz et al. | Representative feature descriptor sets for robust handheld camera localization | |
Abdullah et al. | Smart card with iris recognition for high security access environment | |
Varytimidis et al. | W α SH: weighted α-shapes for local feature detection | |
Zeng et al. | Ear recognition based on 3D keypoint matching | |
CN103336964A (en) | SIFT image matching method based on module value difference mirror image invariant property | |
Narang et al. | Robust face recognition method based on SIFT features using Levenberg-Marquardt Backpropagation neural networks | |
Kim et al. | Improving ancient roman coin recognition with alignment and spatial encoding | |
Berretti et al. | 3D partial face matching using local shape descriptors | |
Mane et al. | Novel multiple impression based multimodal fingerprint recognition system | |
Yi et al. | Registration of infrared and visible images based on the correlation of the edges |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20141231 |
|
WD01 | Invention patent application deemed withdrawn after publication |