CN101707401B - Electrical anti-misoperation locking system and anti-misoperation locking method based on iris identification - Google Patents

Abstract

The invention discloses an electrical anti-misoperation locking system and an anti-misoperation locking method based on iris identification. The system comprises a five-anti monitoring computer, a computer key and an iris identification unit, wherein the computer key is connected with the five-anti monitoring computer and carries out unlocking control on a code lock system, and the iris identification unit is connected with the five-anti monitoring computer and is formed by an iris information acquisition module and a main control module. The method comprises the following steps: 1. acquiring the iris image; 2. using the main control module to analyze the obtained iris image and realize automatic personal identity authentication, namely matching, specifically comprising image processing, characteristic extraction and characteristic matching; and 3. using the five-anti monitoring computer to correspondingly carry out unlocking control on the computer key according to the matching results transmitted by the main control module. The invention features reasonable design, simple and convenient use and operation, flexible operating mode, highest identification accuracy, fast speed and strongest anti-counterfeit capacity, and can effectively overcome various defects and shortcomings in the current electrical anti-misoperation locking systems.

Description

Electric power anti-error locking system and anti-misoperation locking method based on the iris identification

Technical field

The invention belongs to electric power anti-misoperation locking technical field, especially relate to a kind of electric power anti-error locking system and anti-misoperation locking method based on the iris identification.

Background technology

The electric misoperation of transformer station possibly cause large-area power-cuts, device damage, loss of life or personal injury; Even serious consequence such as cause that electrical network vibration is disintegrated; China's electric power system is in face of lesson written in blood many times; In conjunction with the practice of China and foreign countries' electrical operation, in order effectively to prevent to move the person and the substantial equipment accident that misoperation of electrical equipment causes, former Ministry of Water Resources and Electric Power will prevent that in 1980 electric fault-operation accident from classifying the key technical problem that electrical production is badly in need of solving as and issuing.And proposed the requirement of electric equipment " five is anti-", and electric anti-error management, operation, design and use principle have been stipulated with rules form ([1990] No. 1110 literary compositions of energy security) style of writing in nineteen ninety.Five anti-meaning: prevent that 1. the on-load mistake from drawing, mistake is closed isolating switch; 2. prevent that mistake from drawing, mistake is closed circuit breaker; 3. prevent to be with ground wire or ground connection plug-in strip to close a floodgate; 4. prevent charged hanging earthing conductor or splice grafting ground plug-in strip; 5. prevent mistakenly entering charged chamber.One of important measures that have been embodied as electric power safety production of electric " five is anti-" function.Along with the continuous development of electrical network, the continual renovation of technology, error-proof device is updated and is perfect.

From 1980; Former Ministry of Water Resources and Power Industry proposes in electric power system, to prevent five kinds of pernicious electric fault-operation accidents, advocates and has adopted since the five anti-technical measures, and the conventional anti-misoperation locking mode that occurs at home mainly contains 4 kinds: mechanical latching; Procedure lock, electrical interlock and electromagnetic lock.Wherein electrical interlock is a kind of anti-error function that is based upon on the secondary operation loop, and the exclusive circuit that auxiliary contact and the electric wiring through switch and disconnecting link forms, electromagnetic lock are then as the executive component of locking, so both are interdependent.These closedown modes have been brought into play positive role in anti-delaying work, use and move test through 20 years, and the pluses and minuses of various traditional closedown modes all fully show.

Above-mentioned traditional anti-misoperation locking mode has his own strong points, and respectively has it short, and certain range of application is all arranged, and anti-error function also all has certain limitation.1. the traditional anti-misoperation locking mode mainly auxiliary contact through relevant device connects and realizes locking, and locking is reliable; But need to insert a large amount of secondary cables, the mode of connection is comparatively complicated, and operation maintenance is difficulty comparatively; 2. traditional anti-misoperation locking mode generally can only prevent switch, at a distance from the misoperation of cutter and ground cutter, then powerless to the articulating of mistakenly entering charged chamber, earth connection (dismounting) etc.; 3. in the transformer station of large-scale and main junction complicacy, adopt blocking lock,, will make troubles to grid switching operation because there have the key of a greater number to gather to be diffusing; 4. at last also be topmost complete " five the is anti-" function that can not realize exactly.

Go out from the nineties in last century, microcomputer technology has just got into the anti-misoperation locking field, the domestic electric equipment factory shutting device to prevent mistakes in microcomputer that released one after another.Through the development over more than 10 years, these shutting device to prevent mistakes in microcomputer are ripe gradually, and in electric power system, extensively promoted.The microcomputer anti-error system becomes the five-defence block rule base in the computer through the software secondary locked loop that the scene is a large amount of; Realized the digitlization of anti-misoperation locking; And can realize can not realizing or the very difficult anti-error function that realizes, should be the state-of-the-art technology and the leap of electric equipment anti-misoperation locking technology. in the pastThough the computer anti-misoperation and lockout mode is the developing direction of anti-misoperation locking, the intrinsic weakness of its existence must cause our attention.

Except that the equipment dependability reason, from the process of misoperation, misoperation is made up of three parts, and operations staff, maintainer and other personnel's misoperation causes.Microcomputer anti-error just occurs in mainly that mistake is drawn, mistake is closed disconnecting link and switch, is strayed at interval etc. and causes serious accident design easily to operations staff's misoperation; It can only satisfy operations staff's operation requirement, can not solve maintainer and other personnel's misoperation problem fully.For the maintainer, misoperation mainly occurs in maintenance, the process of the test, not within the control of microcomputer anti-error system.For old electric substation, can in transformation process, solve by the suitable old closedown mode of reserve part; For the microcomputer anti-error device was directly installed by new electric substation, then this problem can only rely on management means to solve now.In addition,, be not comprised within the operation rules, promptly do not have locking, can only rely on operations staff's technical merit and sense of responsibility to guarantee that this part does not go wrong though operating in of electric substation's Secondary Part has embodiment in the anti-error system.

At present; Micro_computer five_proof operating system generally is the mode of carrying out cipher authentication when adopting operation for operating personnel's identity verification scheme, if password is correct, could allow to carry out the operation of corresponding authority; Otherwise quiescing; Because itself there is unmanageable in the cipher authentication mode, is prone to lose, use loaded down with trivial details shortcoming, often can not effectively control and record the illegal operation personnel, following problem occurs through regular meeting:

1＞, because password keeping is not good at, cause the personnel that do not have operating right illegally to login, operate in violation of rules and regulations;

2＞, since the user of service forget login password, cause carrying out anti-misoperation;

3＞, because Password Management is not good at, cause operating personnel to surpass the operation of self authority, like illegal modifications logical relation, illegal modifications system configuration etc.;

4＞, operator password and operating personnel are difficult to accomplish corresponding one by one, can't carry out effective record, can't everyone has responsibility after going wrong.

Summary of the invention

Technical problem to be solved by this invention is to above-mentioned deficiency of the prior art the electric power that a kind of easy-to-connect, modern design and service behaviour are safe and reliable, error rate is low anti-error locking system to be provided.

For solving the problems of the technologies described above; The technical scheme that the present invention adopts is: a kind of electric power anti-error locking system based on the iris identification; Comprise " five anti-" monitoring host computer and join with " five is anti-" monitoring host computer and mutually reply coding lock system separate the computer key of lock control; It is characterized in that: also comprise the iris identity recognizer that joins with " five is anti-" monitoring host computer; Said iris identity recognizer comprises the iris information acquisition module and joins with the iris information acquisition module and institute's Information Monitoring is carried out analyzing and processing and realized the main control module of automatic personal identification that said main control module joins with " five is anti-" monitoring host computer respectively.

Said iris information acquisition module comprises the optical lens that is used to absorb iris image, be arranged on the anterior and optical filtering that be used with optical lens of optical lens, the secondary light source that is arranged on optical lens week side, the auxiliary light-operated module that the intensity of illumination of secondary light source is controlled and to the iris image that optical lens absorbed amplify, filtering and digitized processing also tackle the image-signal processor that fill-in light control module is controlled mutually; Said optical lens and image-signal processor join; Image-signal processor joins with main control module and auxiliary light-operated module respectively, and auxiliary light-operated module and secondary light source join.

Said main control module is the ARM9 microprocessor.

Said ARM9 microprocessor is microprocessor S3C2410.

Be connected through USB interface between said image-signal processor and main control module.

In addition, the present invention provides also that a kind of method step is simple, the fast and intelligent ability degree of reaction speed is high based on the method for the electric power anti-misoperation locking of iris identification, it is characterized in that this method may further comprise the steps:

Step 1, iris image obtain: adopt the iris image of iris information acquisition module acquisition operations personnel eye and with the iris image synchronized transmission of said collection to main control module;

Step 2, main control module are stored in the memory received iris image, and synchronously said iris image are carried out analyzing and processing and realizes automatic personal identification, and its analyzing and processing process is following:

201, image preliminary treatment, it mainly may further comprise the steps:

2011, iris image location: when said iris image is positioned; Confirm the inner and outer boundary line of iris earlier respectively through main control module; Said inner and outer boundary line is respectively the boundary line of iris and pupil and the boundary line of iris and sclera and the inner and outer boundary line of being determined and is ellipse; The iris portion image that will be between pupil and sclera based on determined inner and outer boundary line is separated from said iris image afterwards, obtains the iris portion image;

2012, gradation of image is demarcated and enhancement process: adopt main control module that the iris portion image of being separated is carried out gray scale and demarcate and enhancement process;

2013, image normalization is handled: adopt main control module to demarcate through gray scale and enhancement process after the iris portion image carry out normalization and handle, be the rectangular area with the annular region linear stretch of iris portion image, obtain rectangle iris portion image;

202, feature extraction: adopt in the said rectangle iris portion of the main control module image and extract one group of unique parameters representing and to distinguish this rectangle iris portion image, and this group unique parameters is encoded;

203, characteristic matching: adopt main control module that one group of unique parameters extracting in the step 202 and encode is mated with the known iris feature parameter of in said memory, storing in advance; When extracting and one group of unique parameters of coding and any parameter in the said known iris feature parameter when being complementary, matching result be " through "; Otherwise matching result is " refusal "; Said known iris feature parameter is stored in the iris feature parameter database corresponding in the said memory; Said main control module is uploaded to matching result " five is anti-" monitoring host computer synchronously;

Step 3, " five is anti-" monitoring host computer are tackled computer key mutually and are separated lock control according to the matching result that main control module transmitted.

During passing through main control module and confirm the inner and outer boundary line of iris described in the above-mentioned steps 2011, when confirming the inner edge boundary line of inner edge boundary line and definite iris of iris earlier, may further comprise the steps:

Step 1), utilize threshold method to confirm pupil region and determined pupil region is filled;

Step 2), utilize the canny operator to obtain the border of the pupil region of filling, the center of a plurality of somes calculating of picked at random inner edge boundary line correspondence ellipse and the length of major axis and minor axis in the border of acquisition pupil region;

Confirm to confirm again behind the inner edge boundary line of iris the boundary line, outside of iris; And during the boundary line, outside of definite iris; A plurality of boundary points along said inner edge boundary line are sought to both sides in the horizontal direction, find behind the gaussian filtering gray variance greater than a plurality of corresponding points of predetermined threshold value, as the boundary point in boundary line, outside; And utilize a plurality of corresponding points of having found out to calculate the corresponding oval center, boundary line, outside and the length of major axis and minor axis, said predetermined threshold value is 4～7.

Employing main control module described in the above-mentioned steps 2012 carries out that gray scale is demarcated and during enhancement process, adopt to carry out gray scale based on the demarcation Enhancement Method of quotient graph elephant and demarcate and enhancement process, and its demarcation and enhanced processes may further comprise the steps:

Step I, obtain the sample image collection: confirm the illumination of three kinds of varying strengths earlier, and under three kinds of varying strength illumination conditions, gather a plurality of iris image samples and use sample image, and be vector form with each sample picture inversion as demarcation; The sample collection of being gathered is A={A ₁, A ₂..., A _N, wherein N by the quantity of collection sample image, A _iBe [I _i ¹, I _i ², I _i ³] be generated data and the I of one of them sample image under said three kinds of illumination conditions _i ^jBe respectively the image vector of this sample image under said three kinds of illumination conditions, A is N the image set of different iris image samples under the light source of three kinds of linear independences, if order $\stackrel{\‾}{A}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{A}_i,$ A={I then ₁, I ₂, I ₃Wherein, i=1,2,3...N, j=1,2,3;

The quotient images of the iris portion image of separating in Step II, the solution procedure 2011, its solution procedure is following:

(a) utilize formula $v_i={\left(\underset{r=1}{\overset{N}{\mathrm{\Σ}}}{A_r}^T{A}_r\right)}^{-1}{A}_i^T{y}_s$ Find the solution N dimensional vector v _i, wherein i=1,2,3...N, y _s=ρ _yn ^TS is the phasor function of iris portion image under the uniform source of light s irradiation, ρ in the formula _yThe surface reflectivity of representing any point in the said iris portion image, n ^TThe surface normal of representing this point, s are uniform source of light;

(b) pass through equation group:

$\begin{array}{ccccc}{\mathrm{\α}}_1(v_1^T{A}_1^T{y}_s-y_S^T{y}_s)& +\·\·\·& +& {\mathrm{\α}}_N{v}_N^T{A}_1^T{y}_s& =0\\ {\mathrm{\α}}_1{v}_1^T{A}_2^T{y}_s& +\·\·\·& +& {\mathrm{\α}}_N{v}_N^T{A}_2^T{y}_s& =0\\ \·& \·& \·& \·& \·\\ \·& \·& \·& \·& \·\\ \·& \·& \·& \·& \·\\ {\mathrm{\α}}_1{v}_1^T{A}_N^T{y}_s& +\·\·\·& +& {\mathrm{\α}}_N(v_N^T{A}_N^T{y}_s-y_S^T{y}_s)& =0\end{array}$ Find the solution α _i, wherein i=1,2,3...N;

(c) pass through formula ${\mathrm{\α}}_i^{\′}=\frac{{\mathrm{\α}}_i}{\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\α}}_i}\×N$ To α _iChange and this moment $\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\α}}_i^{\′}=1,$ And work as $\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\α}}_i^{\′}=1$ The time, energy function $f\left(x\right)=\frac{1}{2}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{|A_{i}x-{\mathrm{\α}}_i{y}_s|}^2$ Functional value minimum; X is 3 dimensional vectors, i.e. { x in the energy function formula ₁, x ₂, x ₃;

(d) pass through formula $x=\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{\mathrm{\α}}_i^{\′}{v}_i$ Calculating energy function f (x) functional value is hour corresponding three-dimensional vector x;

(e) utilize formula $Q_y=\frac{y_s}{\stackrel{\‾}{A}x}$ After each corresponding pixel points is divided by, obtain quotient images;

Step II I, calculate that said iris portion image is demarcated and the calibration formula during enhancement process does $I_{\mathrm{New}}=\underset{i=1}{\overset{3}{\mathrm{\Σ}}}{z}_i{\stackrel{\‾}{I}}_i\⊗Q_y,$ Z in the formula _iBe control calibrated image irradiation value and $\underset{i=1}{\overset{3}{\mathrm{\Σ}}}{z}_i=1,$

For each corresponding pixel points in cartesian product and its presentation video multiplies each other.

When adopting main control module (4-2) to carry out feature extraction in the above-mentioned steps 202, its characteristic extraction procedure mainly may further comprise the steps:

(L) adopt the 2D-Gabor filter to handling through the preliminary treatment of image described in the step 201

I _Gabor，e(x，y)＝g ^e(x，y)*I(x，y)

After the iris portion image that obtains carry out Filtering Processing, obtain I _{Gabor, o}(x, y)=g ^o(x, y) * I (x, y), I (x, y) the iris portion image of expression before the filtering wherein; g ^e(x, y) and g ^o(x y) is respectively the real part and the imaginary part of 2D Gabor filter, and $g^e(x,y)=\frac{1}{2\mathrm{\π}{\mathrm{\σ}}_x{\mathrm{\σ}}_y}\mathrm{Exp}(-\frac{1}{2}\frac{x_1^2+y_1^2}{{\mathrm{\σ}}^2})\mathrm{Cos}\left(2\mathrm{\π\; f}{x}_1\right),$ $g^o(x,y)=\frac{1}{2\mathrm{\π}{\mathrm{\σ}}_x{\mathrm{\σ}}_y}\mathrm{Exp}(-\frac{1}{2}\frac{x_1^2+y_1^2}{{\mathrm{\σ}}^2})\mathrm{Sin}\left(2\mathrm{\π\; f}{x}_1\right),$

X wherein ₁=xcos (θ)+ysin (θ), y ₁=-xsin (θ)+ycos (θ); F is a sinusoidal plane wave frequency, $f=\frac{1}{e^k},$ Wherein K=1,2,3... ∞; * number expression convolution algorithm;

(M) adopt LBP algorithm and come the iris portion image behind the 2D-Gabor filter filtering is carried out the LBP coding through the number of times of measuring 0 and 1 saltus step, when carrying out the LBP coding, the employing formula:

${\mathrm{LBP}}_P=\left\{\begin{array}{cc}{\mathrm{\Σ}}_{j=0}^{P-1}s(g_j-g_c)& \mathrm{ifU}\left({\mathrm{LBP}}_P\right)\≤2\\ P+1& \mathrm{otherwise}\end{array}\right.$

Wherein, $U\left({\mathrm{LBP}}_P\right)=|s(g_{P-1}-g_c)-s(g_0-g_c)|+\underset{j=1}{\overset{P-1}{\mathrm{\Σ}}}|s(g_j-g_c)-s(g_{j-1}-g_c)|,$ P is the interior pixel count of institute's neighborhood of getting; g _iThe pixel in the zone, g are got in expression _cThe central point in the zone is got in expression;

(N), just obtain a width of cloth LBP characteristic pattern through each pixel in the said iris portion image is asked its LBP feature coding.

When the employing main control module described in the above-mentioned steps 203 carries out characteristic matching, mainly may further comprise the steps:

2031, common region is blocked and is obtained in rejecting:

Utilize formula code _Available=code ∩ mask _Object∩ mask _Sample, finding out the iris portion image after feature extraction is target image to be measured and the common effective coverage that is stored in the known class iris standard picture of the known iris feature parameter correspondence in the said iris feature parameter database in advance; In the formula, Code is a code pattern, mask _ObjectThe LBP characteristic pattern of target image to be measured, madk _SampleLBP characteristic pattern for known class iris standard picture;

2032, obtain characteristic vector: all convert the common region among many feature coding figure of target image to be measured or known class iris standard picture to column vector earlier earlier; Again column vector is spliced into a characteristic vector afterwards, just can obtains target image to be measured and known class iris standard picture characteristic of correspondence vector separately;

2033, dimensionality reduction: to the target image to be measured that obtained in the step (P) and known class iris standard picture separately the characteristic of correspondence vector carry out dimensionality reduction respectively; Promptly pairing characteristic vector is separately carried out the PCA conversion respectively, choose the part element that energy reaches 95% the vector that coefficient constituted and use characteristic as Classification and Identification;

2034, identification: adopt minimum distance classification that the characteristic vector X of target image to be measured is compared and corresponding the classification with the characteristic vector of all known class iris standard pictures of storage in advance; When comparing; Adopt that normalization is European to be measured the distance between characteristic vector X and i class known class iris standard picture characteristic vector, and if only if $d\left(i^{*}\right)=\underset{i}{\mathrm{Min}}d\left(i\right)=\underset{i}{\mathrm{Min}}\underset{k}{\mathrm{\Σ}}\frac{{(X_k-m_k^i)}^2}{{\mathrm{\σ}}_k^i}$ During establishment, said target image to be measured is sorted into i class known class iris standard picture; Otherwise said target image to be measured can not be sorted into i class known class iris standard picture; In the formula, m _k ⁱ, σ _k ⁱAverage and the standard variance of representing k characteristic of i class known class iris standard picture respectively.

The present invention compared with prior art has the following advantages:

1, used based on the iris identification electric power anti-error locking system easy-to-connect, modern design and service behaviour is safe and reliable, error rate is low.Iris has advantages such as uniqueness, stability, collection property, non-infringement property as important identity diagnostic characteristics.The living things feature recognition of non-infringement property (or contactless) is the inexorable trend of identity authentication research and application development, compares with contactless authentication identifying methods such as face picture, sound, and iris has higher accuracy.The error rate of iris recognition is minimum in the various living things feature recognitions.

2, the iris information acquisition module that is adopted uses flexible operation mode, low cost, the clear picture that obtains; And, can in the 20-30cm distance range, collect qualified iris image through technology such as voice suggestion, active vision feedbacks through constantly updating.

3, adopt and to carry out based on the demarcation Enhancement Method of quotient graph elephant that gray scale is demarcated and enhancement process can suppress the even influence of uneven illumination, obtain the relatively more consistent iris image of illuminance, and have contrast preferably, be beneficial to the raising of discrimination.

4, it is the highest to have an accuracy of identification, and speed is fast, the advantage that anti-counterfeit capability is the strongest; Apply it in operating personnel's the authentication, can avoid occurring existing existing number of drawbacks of electric power anti-error locking system and deficiency from technological means, through the iris ID authentication device of distribution some in transformer station; In operating process, suitably operating personnel's identity is forced to confirm again, errorless like identity validation, just can proceed operation; Otherwise the quiescing personnel are proceeded operation, and the correctness at user interval of living in is pointed out and confirmed, correct at interval like the place; Then allow to proceed operation, otherwise prompting user interval mistake, and the quiescing personnel proceed operation; Thereby effectively operating personnel are controlled, and can further effectively prevent the generation of misoperation.

In sum, the present invention is reasonable in design, use is easy and simple to handle, flexible operation mode and accuracy of identification is the highest, speed is fast, anti-counterfeit capability is the strongest, can effectively solve existing existing number of drawbacks of electric power anti-error locking system and deficiency.

Through accompanying drawing and embodiment, technical scheme of the present invention is done further detailed description below.

Description of drawings

Fig. 1 is the theory diagram that the present invention is based on the electric power anti-error locking system of iris identification.

Fig. 2 is the flow chart that the present invention is based on the electric power anti-misoperation locking method of iris identification.Description of reference numerals:

1-" five is anti-" monitoring host computer; The 2-computer key; 3-coding lock system;

4-iris identity recognizer; The 4-1-iris information is gathered mould 4-11-optical lens;

Piece;

The 4-12-optical filtering; The 4-13-image-signal processor; The 4-14-secondary light source;

4-15-assists light-operated module; The 4-2-main control module.

Embodiment

A kind of electric power anti-error locking system as shown in Figure 1 based on the iris identification; Comprise " five anti-" monitoring host computer 1 and join with " five is anti-" monitoring host computer 1 and mutually reply coding lock system 3 separate the computer key 2 of lock control, also comprise the iris identity recognizer 4 that joins with " five is anti-" monitoring host computer 1.Said iris identity recognizer 4 comprises iris information acquisition module 4-1 and joins with iris information acquisition module 4-1 and institute's Information Monitoring is carried out analyzing and processing and realized the main control module 4-2 of automatic personal identification that said main control module 4-2 joins with " five is anti-" monitoring host computer 1.Said coding is locked system 3 and is comprised multiple codings locks such as intelligent lock, romote controlled locking relay, electricity coding lock, mechanical coding lock.

In the present embodiment; Said iris information acquisition module 4-1 comprises the optical lens 4-11 that is used to absorb iris image, be arranged on the anterior and optical filtering 4-12 that be used with optical lens 4-11 of optical lens 4-11, the secondary light source 4-14 that is arranged on optical lens 4-11 week side, the auxiliary light-operated module 4-15 that the intensity of illumination of secondary light source 4-14 is controlled and iris image that optical lens 4-11 is absorbed amplifies, filtering and digitized processing are also tackled the image-signal processor 4-13 that fill-in light control module 4-15 controls mutually; Said optical lens 4-11 and image-signal processor 4-13 join; Image-signal processor 4-13 joins with main control module 4-2 and auxiliary light-operated module 4-15 respectively, and auxiliary light-operated module 4-15 and secondary light source 4-14 join.

Said main control module 4-2 is the ARM9 microprocessor, and said ARM9 microprocessor is microprocessor S 3C2410.Be connected through USB interface between said image-signal processor 4-13 and main control module 4-2.

In conjunction with Fig. 2, the present invention is based on the electric power anti-misoperation locking method of iris identification, may further comprise the steps:

Step 1, iris image obtain: adopt the iris image of iris information acquisition module 4-1 acquisition operations personnel eye and with the iris image synchronized transmission of said collection to main control module 4-2.

Step 2, main control module 4-2 are stored in the memory received iris image, and synchronously said iris image are carried out analyzing and processing and realizes automatic personal identification, and its analyzing and processing process is following:

201, image preliminary treatment, it mainly may further comprise the steps:

2011, iris image location: when said iris image is positioned; Confirm the inner and outer boundary line of iris earlier respectively through main control module 4-2; Said inner and outer boundary line is respectively the boundary line of iris and pupil and the boundary line of iris and sclera and the inner and outer boundary line of being determined and is ellipse; The iris portion image that will be between pupil and sclera according to determined inner and outer boundary line is afterwards separated from said iris image, obtains the iris portion image.

In the present embodiment, when confirming the inner and outer boundary line of iris, when the inner edge boundary line of earlier definite iris and the inner edge boundary line of definite iris, may further comprise the steps through main control module 4-2:

Step 1), utilize threshold method to confirm pupil region and determined pupil region is filled;

Step 2), utilize the canny operator to obtain the border of the pupil region of filling, the center of a plurality of somes calculating of picked at random inner edge boundary line correspondence ellipse and the length of major axis and minor axis in the border of acquisition pupil region;

Confirm to confirm again behind the inner edge boundary line of iris the boundary line, outside of iris; And during the boundary line, outside of definite iris; A plurality of boundary points along said inner edge boundary line are sought to both sides in the horizontal direction, find behind the gaussian filtering gray variance greater than a plurality of corresponding points of predetermined threshold value, as the boundary point in boundary line, outside; And utilize a plurality of corresponding points of having found out to calculate the corresponding oval center, boundary line, outside and the length of major axis and minor axis, said predetermined threshold value is 4～7.

In the actual use, 5 points on the picked at random boundary line can be obtained the corresponding oval center, boundary line and the length of major axis and minor axis usually.When for example asking for the inner edge boundary line, 5 points on the picked at random inner boundary and require these 5 somes distance between any two greater than 10 pixels.

If the corresponding oval equation in inner edge boundary line is following: Ax ²+ 2Bxy+Cy ²+ Dx+Ey+1=0 brings the coordinate of choosing 5 points into above-mentioned elliptic equation, and the group of solving an equation can obtain coefficient A, B, C, D and E.

If the corresponding oval central coordinate of circle in inner edge boundary line be (u, v), can equation be write as:

A(x-u) ²+2B(x-u)(y-v)+C(y-v) ²+f＝0

With the contrast of general equation formula, have $-2\left[\begin{array}{cc}A& B\\ B& C\end{array}\right]\left[\begin{array}{c}u\\ v\end{array}\right]=\left[\begin{array}{c}D\\ E\end{array}\right]$

Can calculate thus the oval center of circle (u, v), elliptic equation is under situation without spin:

$\frac{{(x-u)}^2}{a^2}+\frac{{(y-v)}^2}{b^2}=1$

Order $L_{}$ ${}_1=\frac{1}{a^2},{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">L</figure-callout>}}_2=\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="b"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">b</figure-callout>}}^2},$ Simultaneously on inner boundary, look for 2 points again, establish its coordinate and be respectively (x ₁, y ₁), (x _y, y _y), through following formula calculate (L ₁, L ₂),

$\left[\begin{array}{cc}{(x_1-u)}^2& {(y_1-v)}^2\\ {(x_2-u)}^2& {(y_2-v)}^2\end{array}\right]\left[\begin{array}{c}{L}_1\\ {\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">L</figure-callout>}}_2\end{array}\right]=\left[\begin{array}{c}1\\ 1\end{array}\right]$

Obtain $\left[\begin{array}{c}{L}_1\\ {\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">L</figure-callout>}}_2\end{array}\right]={\left[\begin{array}{cc}{(x_1-u)}^2& {(y_1-v)}^2\\ {(x_2-u)}^2& {(y_2-v)}^2\end{array}\right]}^{-1}\left[\begin{array}{c}1\\ 1\end{array}\right],$ Correspondingly just obtain: $a=\sqrt{1/{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">L</figure-callout>}}_1},b=\sqrt{1/{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">L</figure-callout>}}_2},$ Obtain the corresponding oval center of circle, inner edge boundary line and the size of major axis and minor axis thus.

2012, gradation of image is demarcated and enhancement process: adopt main control module 4-2 that the iris portion image of being separated is carried out gray scale and demarcate and enhancement process.

Employing main control module 4-2 described in the step 2012 carries out that gray scale is demarcated and during enhancement process, adopt to carry out gray scale based on the demarcation Enhancement Method of quotient graph elephant and demarcate and enhancement process, and its demarcation and enhanced processes may further comprise the steps:

Step I, obtain the sample image collection: confirm the illumination of three kinds of varying strengths earlier, and under three kinds of varying strength illumination conditions, gather a plurality of iris image samples and use sample image, and be vector form with each sample picture inversion as demarcation; The sample collection of being gathered is A={A ₁, A ₂..., A _N, wherein N by the quantity of collection sample image, A _iBe [I _i ¹, I _i ², I _i ³Be generated data and the I of one of them sample image under said three kinds of illumination conditions _i ^jBe respectively the image vector of this sample image under said three kinds of illumination conditions, A is N the image set of different iris image samples under the light source of three kinds of linear independences, if order $\stackrel{\&OverBar;}{A}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{A}_i,$ A={I then ₁, I ₂, I ₃; Wherein, i=1,2,3...N, j=1,2,3;

The quotient images of the iris portion image of separating in Step II, the solution procedure 2011, its solution procedure is following:

(a) utilize formula $v_i={\left(\underset{r=1}{\overset{N}{\mathrm{\&Sigma;}}}{A_r}^T{A}_r\right)}^{-1}{A}_i^T{y}_s$ Find the solution N dimensional vector v _i, wherein i=1,2,3...N, y _s=ρ _yn ^TS is the phasor function of iris portion image under the uniform source of light s irradiation, ρ in the formula _yThe surface reflectivity of representing any point in the said iris portion image, n ^TThe surface normal of representing this point, s are uniform source of light;

(b) pass through equation group:

$\begin{array}{ccccc}{\mathrm{\&alpha;}}_1(v_1^T{A}_1^T{y}_s-y_S^T{y}_s)& +\&CenterDot;\&CenterDot;\&CenterDot;& +& {\mathrm{\&alpha;}}_N{v}_N^T{A}_1^T{y}_s& =0\\ {\mathrm{\&alpha;}}_1{v}_1^T{A}_2^T{y}_s& +\&CenterDot;\&CenterDot;\&CenterDot;& +& {\mathrm{\&alpha;}}_N{v}_N^T{A}_2^T{y}_s& =0\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;& \&CenterDot;\\ {\mathrm{\&alpha;}}_1{v}_1^T{A}_N^T{y}_s& +\&CenterDot;\&CenterDot;\&CenterDot;& +& {\mathrm{\&alpha;}}_N(v_N^T{A}_N^T{y}_s-y_S^T{y}_s)& =0\end{array}$ Find the solution α _i, wherein i=1,2,3...N;

(c) pass through formula ${\mathrm{\&alpha;}}_i^{\&prime;}=\frac{{\mathrm{\&alpha;}}_i}{\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i}\&times;N$ To α _iChange and this moment $\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i^{\&prime;}=1,$ And work as $\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i^{\&prime;}=1$ The time, energy function $f\left(x\right)=\frac{1}{2}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{|A_{i}x-{\mathrm{\&alpha;}}_i{y}_s|}^2$ Functional value minimum; X is 3 dimensional vectors, i.e. { x in the energy function formula ₁, x ₂, x ₃;

(d) pass through formula $x=\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{\mathrm{\&alpha;}}_i^{\&prime;}{v}_i$ Calculating energy function f (x) functional value is hour corresponding three-dimensional vector x;

(e) utilize formula $Q_y=\frac{y_s}{\stackrel{\&OverBar;}{A}x}$ After each corresponding pixel points is divided by, obtain quotient images;

Step II I, calculate that said iris portion image is demarcated and the calibration formula during enhancement process does $I_{\mathrm{New}}=\underset{i=1}{\overset{3}{\mathrm{\&Sigma;}}}{z}_i{\stackrel{\&OverBar;}{I}}_i\&CircleTimes;Q_y,$ Z in the formula _iBe control calibrated image irradiation value and $\underset{i=1}{\overset{3}{\mathrm{\&Sigma;}}}{z}_i=1,$ For each corresponding pixel points in cartesian product and its presentation video multiplies each other.

2013, image normalization is handled: adopt main control module 4-2 to demarcate through gray scale and enhancement process after the iris portion image carry out normalization and handle, be the rectangular area with the annular region linear stretch of iris portion image, obtain rectangle iris portion image.

In the present embodiment, when carrying out the normalization processing, adopt following formula to characterize:

I(x(ρ，θ)，y(ρ，θ))→I(ρ，θ)

$\left\{\begin{array}{c}x(\mathrm{\&rho;},\mathrm{\&theta;})=(1-\mathrm{\&rho;})\&times;x_p\left(\mathrm{\&theta;}\right)+\mathrm{\&rho;}\&times;x_i\left(\mathrm{\&theta;}\right)\\ y(\mathrm{\&rho;},\mathrm{\&theta;})=(1-\mathrm{\&rho;})\&times;y_p\left(\mathrm{\&theta;}\right)+\mathrm{\&rho;}\&times;y_i\left(\mathrm{\&theta;}\right)\end{array}\right.$

Wherein, (x _p(θ), y _p(θ)) and (x _i(θ), y _i(θ)) be respectively iris inward flange and outer peripheral point on the θ direction, θ ∈ [0,2 π], ρ ∈ [0,1].

The size of image array is by θ after the normalization, and the step-length of ρ confirms that it is 0.005-0.01 that the present invention gets angle step Δ θ, and radially step delta ρ is 0.001-0.005, and the resolution that concrete value can be when gathering is definite.

202, feature extraction: adopt in the said rectangle iris portion of the main control module 4-2 image and extract one group of unique parameters representing and to distinguish this rectangle iris portion image, and this group unique parameters is encoded.

In the present embodiment, when adopting main control module 4-2 to carry out feature extraction, its characteristic extraction procedure mainly may further comprise the steps:

(L) adopt the 2D-Gabor filter to handling through the preliminary treatment of image described in the step 201

I _Gabor，e(x，y)＝g ^e(x，y)*I(x，y)

After the iris portion image that obtains carry out Filtering Processing, obtain I _{Gabor, o}(x, y)=g ^o(x, y) * I (x, y), I (x, y) the iris portion image of expression before the filtering wherein; g ^e(x, y) and g ^o(x y) is respectively the real part and the imaginary part of 2D Gabor filter, and $g^e(x,y)=\frac{1}{2\mathrm{\&pi;}{\mathrm{\&sigma;}}_x{\mathrm{\&sigma;}}_y}\mathrm{Exp}(-\frac{1}{2}\frac{x_1^2+{\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">y</figure-callout>}}_1^2}{{\mathrm{\&sigma;}}^2})\mathrm{Cos}\left(2\mathrm{\&pi;\; f}{x}_1\right),$ $g^o(x,y)=\frac{1}{2\mathrm{\&pi;}{\mathrm{\&sigma;}}_x{\mathrm{\&sigma;}}_y}\mathrm{Exp}(-\frac{1}{2}\frac{x_1^2+{\mathrm{<figure-callout\; id="1"\; label="y"\; filenames="H2009102192081E00011.png"\; state="\{\{state\}\}">y</figure-callout>}}_1^2}{{\mathrm{\&sigma;}}^2})\mathrm{Sin}\left(2\mathrm{\&pi;\; f}{x}_1\right),$

X wherein ₁=xcos (θ)+ysin (θ), y ₁=-xsin (θ)+ycos (θ); F is a sinusoidal plane wave frequency, $f=\frac{1}{e^k},$ Wherein K=1,2,3... ∞; * number expression convolution algorithm;

(M) adopt LBP algorithm and come the iris portion image behind the 2D-Gabor filter filtering is carried out the LBP coding through the number of times of measuring 0 and 1 saltus step, when carrying out the LBP coding, the employing formula:

${\mathrm{LBP}}_P=\left\{\begin{array}{cc}{\mathrm{\&Sigma;}}_{j=0}^{P-1}s(g_j-g_c)& \mathrm{ifU}\left({\mathrm{LBP}}_P\right)\&le;2\\ P+1& \mathrm{otherwise}\end{array}\right.$

Wherein, $U\left({\mathrm{LBP}}_P\right)=|s(g_{P-1}-g_c)-s(g_0-g_c)|+\underset{j=1}{\overset{P-1}{\mathrm{\&Sigma;}}}|s(g_j-g_c)-s(g_{j-1}-g_c)|,$ P is the interior pixel count of institute's neighborhood of getting; g _iThe pixel in the zone, g are got in expression _cThe central point in the zone is got in expression;

(N), just obtain a width of cloth LBP characteristic pattern through each pixel in the said iris portion image is asked its LBP feature coding.

203, characteristic matching: adopt main control module 4-2 that one group of unique parameters extracting in the step 202 and encode is mated with the known iris feature parameter of in said memory, storing in advance; When extracting and one group of unique parameters of coding and any parameter in the said known iris feature parameter when being complementary, matching result be " through "; Otherwise matching result is " refusal "; Said known iris feature parameter is stored in the iris feature parameter database corresponding in the said memory; Said main control module 4-2 is uploaded to matching result " five is anti-" monitoring host computer 1 synchronously.

In the present embodiment, when adopting main control module 4-2 to carry out characteristic matching, mainly may further comprise the steps:

2031, common region is blocked and is obtained in rejecting:

Utilize formula code _Available=code ∩ mask _Object∩ mask _Sample, finding out the iris portion image after feature extraction is target image to be measured and the common effective coverage that is stored in the known class iris standard picture of the known iris feature parameter correspondence in the said iris feature parameter database in advance; In the formula, Code is a code pattern, mask _ObjectThe LBP characteristic pattern of target image to be measured, mask _SampleLBP characteristic pattern for known class iris standard picture;

2032, obtain characteristic vector: all convert the common region among many feature coding figure of target image to be measured or known class iris standard picture to column vector earlier earlier; Again column vector is spliced into a characteristic vector afterwards, just can obtains target image to be measured and known class iris standard picture characteristic of correspondence vector separately;

2033, dimensionality reduction: to the target image to be measured that obtained in the step (P) and known class iris standard picture separately the characteristic of correspondence vector carry out dimensionality reduction respectively; Promptly pairing characteristic vector is separately carried out the PCA conversion respectively, choose the part element that energy reaches 95% the vector that coefficient constituted and use characteristic as Classification and Identification;

2034, identification: adopt minimum distance classification that the characteristic vector X of target image to be measured is compared and corresponding the classification with the characteristic vector of all known class iris standard pictures of storage in advance; When comparing; Adopt that normalization is European to be measured the distance between characteristic vector X and i class known class iris standard picture characteristic vector, and if only if $d\left(i^{*}\right)=\underset{i}{\mathrm{Min}}d\left(i\right)=\underset{i}{\mathrm{Min}}\underset{k}{\mathrm{\&Sigma;}}\frac{{(X_k-m_k^i)}^2}{{\mathrm{\&sigma;}}_k^i}$ During establishment, said target image to be measured is sorted into i class known class iris standard picture; Otherwise said target image to be measured can not be sorted into i class known class iris standard picture; In the formula, m _k ⁱ, σ _k ⁱAverage and the standard variance of representing k characteristic of i class known class iris standard picture respectively.

Step 3, " five is anti-" matching result that monitoring host computer 1 is transmitted according to main control module 4-2 are tackled computer key 2 mutually and are separated lock control.

The above; It only is preferred embodiment of the present invention; Be not that the present invention is done any restriction, every technical spirit changes any simple modification, change and the equivalent structure that above embodiment did according to the present invention, all still belongs in the protection range of technical scheme of the present invention.

Claims (9)

1. the method for anti-misoperation locking is carried out in a utilization based on the electric power anti-error locking system of iris identification; The electric power anti-error locking system that is adopted based on the iris identification comprise " five anti-" monitoring host computer (1) and join with " five is anti-" monitoring host computer (1) and mutually reply coding lock system (3) separate the computer key (2) of lock control; Also comprise the iris identity recognizer (4) that joins with " five is anti-" monitoring host computer (1); Said iris identity recognizer (4) comprises iris information acquisition module (4-1) and joins with iris information acquisition module (4-1) and institute's Information Monitoring is carried out analyzing and processing and realized the main control module (4-2) of automatic personal identification; Said main control module (4-2) joins, it is characterized in that with " five is anti-" monitoring host computer (1) respectively: this method may further comprise the steps:

Step 1, iris image obtain: adopt the iris image of iris information acquisition module (4-1) acquisition operations personnel eye and with iris image synchronized transmission to the main control module (4-2) of said collection;

Step 2, main control module (4-2) are stored in the memory received iris image, and synchronously said iris image are carried out analyzing and processing and realizes automatic personal identification, and its analyzing and processing process is following:

201, image preliminary treatment, it mainly may further comprise the steps:

2011, iris image location: when said iris image is positioned; Confirm the inner and outer boundary line of iris earlier respectively through main control module (4-2); Said inner and outer boundary line is respectively the boundary line of iris and pupil and the boundary line of iris and sclera and the inner and outer boundary line of being determined and is ellipse; The iris portion image that will be between pupil and sclera according to determined inner and outer boundary line is afterwards separated from said iris image, obtains the iris portion image;

2012, gradation of image is demarcated and enhancement process: adopt main control module (4-2) that the iris portion image of being separated is carried out gray scale and demarcate and enhancement process;

2013, image normalization is handled: adopt main control module (4-2) to demarcate through gray scale and enhancement process after the iris portion image carry out normalization and handle, be the rectangular area with the annular region linear stretch of iris portion image, obtain rectangle iris portion image;

202, feature extraction: adopt main control module (4-2) from said rectangle iris portion image, to extract one group of unique parameters representing and to distinguish this rectangle iris portion image, and this group unique parameters is encoded;

203, characteristic matching: adopt main control module (4-2) that one group of unique parameters extracting in the step 202 and encode is mated with the known iris feature parameter of in said memory, storing in advance; When extracting and one group of unique parameters of coding and any parameter in the said known iris feature parameter when being complementary, matching result be " through "; Otherwise matching result is " refusal "; Said known iris feature parameter is stored in the iris feature parameter database corresponding in the said memory; Said main control module (4-2) is uploaded to matching result " five is anti-" monitoring host computer (1) synchronously;

Step 3, " five is anti-" monitoring host computer (1) are tackled computer key (2) mutually and are separated lock control according to the matching result that main control module (4-2) is transmitted.

2. according to the described method of claim 1; It is characterized in that: said iris information acquisition module (4-1) comprises the optical lens (4-11) that is used to absorb iris image, be arranged on the anterior and optical filtering (4-12) that be used with optical lens (4-11) of optical lens (4-11), the secondary light source (4-14) that is arranged on all sides of optical lens (4-11), the auxiliary light-operated module (4-15) that the intensity of illumination of secondary light source (4-14) is controlled and iris image that optical lens (4-11) is absorbed amplifies, filtering and digitized processing are also tackled the image-signal processor (4-13) that fill-in light control module (4-15) is controlled mutually; Said optical lens (4-11) joins with image-signal processor (4-13); Image-signal processor (4-13) joins with main control module (4-2) and auxiliary light-operated module (4-15) respectively, and auxiliary light-operated module (4-15) is joined with secondary light source (4-14).

3. according to claim 1 or 2 described methods, it is characterized in that: said main control module (4-2) is the ARM9 microprocessor.

4. according to the described method of claim 3, it is characterized in that: said ARM9 microprocessor is microprocessor S3C2410.

5. according to the described method of claim 2, it is characterized in that: be connected through USB interface between said image-signal processor (4-13) and main control module (4-2).

6. according to the described method of claim 1, it is characterized in that: during passing through main control module (4-2) and confirm the inner and outer boundary line of iris described in the step 2011, confirm the inner edge boundary line of iris earlier; And the inner edge boundary line of definite iris may further comprise the steps:

Step 1), utilize threshold method to confirm pupil region and determined pupil region is filled;

Step 2), utilize the canny operator to obtain the border of the pupil region of filling, the center of a plurality of somes calculating of picked at random inner edge boundary line correspondence ellipse and the length of major axis and minor axis in the border of acquisition pupil region;

Confirm to confirm again behind the inner edge boundary line of iris the boundary line, outside of iris; And during the boundary line, outside of definite iris; A plurality of boundary points along said inner edge boundary line are sought to both sides in the horizontal direction, find behind the gaussian filtering gray variance greater than a plurality of corresponding points of predetermined threshold value, as the boundary point in boundary line, outside; And utilize a plurality of corresponding points of having found out to calculate the corresponding oval center, boundary line, outside and the length of major axis and minor axis, said predetermined threshold value is 4～7.

7. according to claim 1 or 6 described methods; It is characterized in that: when the employing main control module (4-2) described in the step 2012 carries out gray scale demarcation and enhancement process; Employing is carried out gray scale demarcation and enhancement process based on the demarcation Enhancement Method of quotient images, and its gray scale is demarcated and enhanced processes may further comprise the steps:

Step I, obtain the sample image collection: confirm the illumination of three kinds of varying strengths earlier, and under three kinds of varying strength illumination conditions, gather a plurality of iris image samples and use sample image, and convert each sample image into vector form as demarcation; The sample image collection of being gathered is A={A ₁, A ₂..., A _N, wherein N is the quantity of institute's capturing sample image, A _iFor For one of them sample image under said three kinds of illumination conditions generated data and

Be respectively the image vector of this sample image under said three kinds of illumination conditions, A is N the image set of different iris image samples under the light source of three kinds of linear independences, if order

Then Wherein, i=1,2,3...N, j=1,2,3;

The quotient images of the iris portion image of separating in Step II, the solution procedure 2011, its solution procedure is following:

(a) utilize formula

Find the solution N dimensional vector v _i, wherein i=1,2,3...N, y _s=ρ _yn ^TS is the phasor function of iris portion image under the uniform source of light s irradiation, ρ in the formula _yThe surface reflectivity of representing any point in the said iris portion image, n ^TThe surface normal of representing this point, s are uniform source of light;

(b) pass through equation group:

$\begin{array}{ccc}{\mathrm{\&alpha;}}_1(v_1^T{A}_1^T{y}_s-y_S^T{y}_s)& +...+{\mathrm{\&alpha;}}_N{v}_N^T{A}_1^T{y}_s& =0\\ {\mathrm{\&alpha;}}_1{v}_1^T{A}_2^T{y}_s& +...+{\mathrm{\&alpha;}}_N{v}_N^T{A}_2^T{y}_s& =0\\ \&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\\ \&CenterDot;& \&CenterDot;\&CenterDot;\&CenterDot;& \&CenterDot;\\ {\mathrm{\&alpha;}}_1{v}_1^T{A}_N^T{y}_s& +...+{\mathrm{\&alpha;}}_N(v_N^T{A}_N^T{y}_s-y_S^T{y}_s)& =0\end{array}$ Find the solution α _i, wherein i=1,2,3...N;

(c) pass through formula

To α _iChange and this moment And work as

The time, energy function

Functional value minimum; X is 3 dimensional vectors, i.e. { x in the energy function formula ₁, x ₂, x ₃;

(d) through formula

calculating energy function f (x) functional value hour corresponding three-dimensional vector x;

(e) utilize formula

that each corresponding pixel points is divided by after, obtain quotient images;

Step II I, calculate that said iris portion image carries out that gray scale is demarcated and the calibration formula during enhancement process does

Z in the formula _iBe control calibrated image irradiation value and

For each corresponding pixel points in cartesian product and its presentation video multiplies each other.

8. according to claim 1 or 6 described methods, it is characterized in that: when adopting main control module (4-2) to carry out feature extraction in the step 202, its characteristic extraction procedure mainly may further comprise the steps:

(L) adopt the 2D-Gabor filter that the iris portion image that obtains after handling through the preliminary treatment of image described in the step 201 is carried out Filtering Processing, obtain $\begin{array}{c}{I}_{\mathrm{Gabor},e}(x,y)=g^e(x,y)*I(x,y)\\ I_{\mathrm{Gabor},o}(x,y)=g^o(x,y)*I(x,y)\end{array},$ Wherein (x y) representes the iris portion image that filtering is preceding to I; g ^e(x, y) and g ^o(x y) is respectively the real part and the imaginary part of 2D Gabor filter, and $g^e(x,y)=\frac{1}{2\mathrm{\&pi;}{\mathrm{\&sigma;}}_x{\mathrm{\&sigma;}}_y}\mathrm{Exp}(-\frac{1}{2}\frac{x_1^2+y_1^2}{{\mathrm{\&sigma;}}^2})\mathrm{Cos}\left(2\mathrm{\&pi;\; f}{x}_1\right),$ $g^o(x,y)=\frac{1}{2\mathrm{\&pi;}{\mathrm{\&sigma;}}_d{\mathrm{\&sigma;}}_y}\mathrm{Exp}(-\frac{1}{2}\frac{x_1^2+y_1^2}{{\mathrm{\&sigma;}}^2})\mathrm{Sin}\left(2\mathrm{\&pi;\; f}{x}_1\right),$

X wherein ₁=xcos (θ)+ysin (θ), y ₁=-xsin (θ)+ycos (θ); F is a sinusoidal plane wave frequency,

be K=1,2,3... ∞ wherein; * number expression convolution algorithm;

(M) adopt LBP algorithm and come the iris portion image behind the 2D-Gabor filter filtering is carried out the LBP coding through the number of times of measuring 0 and 1 saltus step, when carrying out the LBP coding, the employing formula:

Wherein, $U\left({\mathrm{LBP}}_P\right)=|s(g_{P-1}-g_c)-s(g_0-g_c)|+\underset{j=1}{\overset{P-1}{\mathrm{\&Sigma;}}}|s(g_j-g_c)-s(g_{j-1}-g_c)|,$ P is the interior pixel count of institute's neighborhood of getting; g _jThe pixel in the zone, g are got in expression _cThe central point in the zone is got in expression;

(N), just obtain a width of cloth LBP characteristic pattern through each pixel in the said iris portion image is asked its LBP feature coding.

9. according to the described method of claim 8, it is characterized in that: when the employing main control module (4-2) described in the step 203 carries out characteristic matching, mainly may further comprise the steps:

2031, common region is blocked and is obtained in rejecting:

Utilize formula code _Available=code ∩ mask _Object∩ mask _Sample, finding out the iris portion image after feature extraction is target image to be measured and the common effective coverage that is stored in the known class iris standard picture of the known iris feature parameter correspondence in the said iris feature parameter database in advance; In the formula, Code is a code pattern, mask _ObjectBe the LBP characteristic pattern of target image to be measured, mask _SampleLBP characteristic pattern for known class iris standard picture;

2032, obtain characteristic vector: all convert the common region among many feature coding figure of target image to be measured or known class iris standard picture to column vector earlier earlier; Again column vector is spliced into a characteristic vector afterwards, just can obtains target image to be measured and known class iris standard picture characteristic of correspondence vector separately;

2033, dimensionality reduction: to the target image to be measured that obtained in the step 2032 and known class iris standard picture separately the characteristic of correspondence vector carry out dimensionality reduction respectively; Promptly pairing characteristic vector is separately carried out the PCA conversion respectively, choose the part element that energy reaches 95% the vector that coefficient constituted and use characteristic as Classification and Identification;

2034, identification: adopt minimum distance classification that the characteristic vector X of target image to be measured is compared and corresponding the classification with the characteristic vector of all known class iris standard pictures of storage in advance; When comparing; Adopt that normalization is European to be measured the distance between characteristic vector X and i class known class iris standard picture characteristic vector; During establishment that and if only if

, said target image to be measured is sorted into i class known class iris standard picture; Otherwise said target image to be measured can not be sorted into i class known class iris standard picture; In the formula,

representes the average and the standard variance of k characteristic of i class known class iris standard picture respectively.

Images