CN104881634B - A kind of illumination face recognition method based on complete Local Convex diesinking - Google Patents

A kind of illumination face recognition method based on complete Local Convex diesinking Download PDF

Info

Publication number
CN104881634B
CN104881634B CN201510223240.2A CN201510223240A CN104881634B CN 104881634 B CN104881634 B CN 104881634B CN 201510223240 A CN201510223240 A CN 201510223240A CN 104881634 B CN104881634 B CN 104881634B
Authority
CN
China
Prior art keywords
mrow
msub
image
pixel
clccp
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.)
Expired - Fee Related
Application number
CN201510223240.2A
Other languages
Chinese (zh)
Other versions
CN104881634A (en
Inventor
陈熙
晋杰
潘晓露
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Kunming University of Science and Technology
Original Assignee
Kunming University of Science and Technology
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Kunming University of Science and Technology filed Critical Kunming University of Science and Technology
Priority to CN201510223240.2A priority Critical patent/CN104881634B/en
Publication of CN104881634A publication Critical patent/CN104881634A/en
Application granted granted Critical
Publication of CN104881634B publication Critical patent/CN104881634B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/16—Human faces, e.g. facial parts, sketches or expressions
    • G06V40/161—Detection; Localisation; Normalisation
    • G06V40/162—Detection; Localisation; Normalisation using pixel segmentation or colour matching

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Oral & Maxillofacial Surgery (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The present invention relates to a kind of illumination face recognition method based on complete Local Convex diesinking, belongs to area of pattern recognition.Piecemeal is carried out to image first；Bilinear interpolation is carried out to each block image；By being encoded to the symbolic feature of the local difference of each pixel and amplitude characteristic in each image block, symbolic feature matrix, the amplitude characteristic matrix of each image block are obtained.Then the pixel of each image block is encoded to obtain the center pixel eigenmatrix of each image block, then the histogram feature of this three eigenmatrixes is extracted, three characteristic vectors are obtained, this three characteristic vectors is sequentially connected and obtains the histogram feature vector of image block；The histogram feature vector for finally connecting each image block obtains the histogram feature vector of this original image, and this feature vector is sent into nearest neighbor classifier and classified, to identify the identity of original facial image.The present invention is the method for describing texture of image based on second-order differential, can effectively carry out photoenvironment human face identification.

Description

A kind of illumination face recognition method based on complete Local Convex diesinking
Technical field
The present invention relates to a kind of illumination face recognition method based on complete Local Convex diesinking, belong to mode identification technology Field.
Background technology
Local binary patterns (Local binary pattern, LBP) [L.Wang and D.C.He, " Texture classification using texture spectrum”,Pattern Recognition,vol.23,pp.905-910, 1990.] it is a kind of important image characteristics extraction operator, there is the characteristics of amount of calculation is small and effective.Although LBP regards in computer Feel and area of pattern recognition has been obtained for very big success, but its working mechanism still there are worth improvements.Dominant part two Value pattern (Dominant local binary patterns, DLBP) [S.Liao, M.W.K.Law, and A.C.S.Chung, “Dominant local binary patterns for texture classification,”IEEE Trans.Image Process., vol.18, no.5, pp.1107-1118, May 2009.] on the basis of all patterns of LBP of statistical picture, sieve Select the pattern of upper frequency, and cumulative frequency reached 80% high frequency mode and form final characteristic vector.LBP is only examined Consider the symbolic information of center pixel and surrounding pixel difference, complete local binary patterns (Completed local binary pattern,CLBP)[Z.Guo,L.Zhang and D.Zhang,“A completed modeling of local binary pattern operator for texture classification,”IEEE Trans.Image Process., Vol.19, no.6, pp.1657-1663,2010.] not only allow for symbolic information, it is also contemplated that the amplitude information of difference and in The feature of imago vegetarian refreshments.LBP extraction be image first differential information, infinitesimal pattern (Local derivative pattern,LDP)[B.Zhang,Y.Gao,S.Zhao,and J.Liu,“Local derivative pattern versus local binary pattern:Face recognition with higher-order local pattern Descriptor, " IEEE Trans.Image Process., vol.19, no.2, pp.533-544, Feb.2010.] improve LBP algorithms, it is extracted the second-order differential information of image.In order to reduce the number of pattern in LBP algorithms, during researcher proposes The symmetrical infinitesimal pattern of the heart (Center-Symmetric Local derivative Pattern, CS-LDP) [G.Xue, L.Song, J.Sun, M.Wu, Hybrid Center-Symmetric Local Pattern for Dynamic Background Subtraction, ICME, Barcelona, Spain (2011), pp.1-6, July 2011.] and center pair Title local binary patterns algorithm (Center-symmetric local binary pattern, CS-LBP) [Marko H, Matti P,Cordelia S.Description of interest regions with center-symmetric local binary pattern[C]//Conference on Computer Vision Graphics and Image Processing.2006,4338:58-69].Local binary counts (Local binary count, LBC) [Zhao Y, Huang D S,Jia W,“Completed local binary count for rotation invariant texture Classification, " IEEE Trans.Image Process., vol.21, no.10, pp.4492-4497,2012.] only Consider that pattern is the number of " 1 " in binary pattern.Unified local binary patterns reduce model number, reduce amount of calculation [T.Ojala,M.T.“Gray scale and rotation invariant texture classification with local binary patterns,”in:D.Vernon(Ed.),Proceedings of the Sixth European Conference on Computer Vision(ECCV2000),Dublin,Ireland, pp.404–420,2000.].In order to strengthen the distinctive that LBP algorithms extract texture, LBP algorithms also with Gabor filter and Some Data Dimensionality Reduction Algorithms combine [Zhang W C, Shan S G, Gao W, et a1.Local Gabor Binary Pattern Histogram Sequence.(LGBPHS)：A Novel Non-Statistical Model for Face Representation and Recognition[C]Proc of the 10th IEEE Int’l Conf on Computer Vision, 2005：786—791.；B.Zhang,S.Shan,X.Chen,and W.Gao,“Histogram of Gabor Phase Patterns(HGPP):A novel object representation approach for face recognition,”IEEE Trans.Image Process.,vol.16,no.1,pp.57–68,2007.]。
LBP only considers the first differential information of image, and it is an object of the invention to provide one kind to be based on the complete office of image The facial image second order of portion's convexo-concave characteristic (Completed Local convex-and-concave Pattern, CLCCP) is micro- Textural characteristics are divided to describe method.
The content of the invention
The invention provides a kind of illumination face recognition method based on complete Local Convex diesinking, for solving illumination Environment human face identifies problem.The defects of being only capable of describing image first differential for local binary patterns, it is proposed by the present invention complete Standby Local Convex diesinking can effectively describe the second-order differential feature of image.Complete Local Convex diesinking not only allows for local difference Symbolic information, and consider the amplitude characteristic of difference, it is also contemplated that the distinctive of central pixel point.
What the illumination face recognition method of the invention based on complete Local Convex diesinking was realized in：Image is entered first Row piecemeal；Then bilinear interpolation being carried out to each block image so that each pixel can build 8 symmetry directions in image, Then each pixel is calculated in block image along 8 direction part difference；Then the symbolic feature of this local difference is encoded CLCCP_S and amplitude characteristic CLCCP_M；Each pixel of each image block is encoded, obtains the middle imago of each image block Plain feature CLCCP-C；Next histogram is extracted to CLCCP_S, CLCCP_M and the CLCCP_C eigenmatrix of each block image Characteristic vector, the histogram feature vector for being sequentially connected block image CLCCP_S, CLCCP_S and CLCCP_C feature obtain respectively The histogram feature vector of block image；The histogram feature vector for finally connecting each block image obtains the straight of this original image Square figure characteristic vector, this feature vector is sent into the nearest neighbor classifier based on chi amount and classified, it is original to identify The identity of facial image.
The illumination face recognition method based on complete Local Convex diesinking comprises the following steps that：
Step1, image is subjected to piecemeal first：Image I(l)Uniformly be divided into 4 × 4 non-overlapping square, 16 pieces altogether, It is expressed as
Step2, bilinear interpolation computing is carried out to each block image so that each pixel can be built on the pixel Symmetrical 8 directions, then calculate local difference of each pixel along different directions, and the local difference is decomposed into symbol portion Point and and amplitude portion；
As shown in Fig. 2 pixel P1And P2Between pixel Q can be increased by interpolation1.Interpolation method as shown in figure 4, Wherein P11, P12, P21, P22It is four neighbor pixels original in image, new pixel Q is inserted out by interpolation method0.Interpolation Formula is as follows：
WhereinWithR is represented respectively1, R2WithThe pixel value of opening position, x1, x and x2Pixel is represented respectively P11, R1And P21The abscissa at place, y1, y and y2Pixel P is represented respectively11, Q0And P12The ordinate at place.Fig. 3 represents original image Middle pixel X0P be present in surrounding0, P1, P2, P3, P4, P5, P6And P78 Neighbor Points, it is only capable of forming four on pixel X0It is symmetrical Direction.Fig. 2 represents pixel X after interpolation0Q be present in surrounding0, Q1, Q2, Q3, Q4, Q5, Q6And Q78 interpolation points, therefore pixel after interpolation X0Surrounding one co-exists in 16 Neighbor Points, can obtain 8 on pixel X0Symmetry direction.Due to adding interpolation point, to image Resolution ratio enhance；
Pixel X in image block0Local difference along 8 directions isWithWherein i=0,1,2,3 and j=0,1,2,3；
Step3, carry out corresponding Local Convex diesinking coding respectively to symbolic component and amplitude portion, obtain each block diagram The symbolic feature CLCCP-S and amplitude characteristic CLCCP-M of picture, wherein pixel X0Symbolic feature and amplitude characteristic coding it is public Formula is respectively：
Wherein, CLCCP-S1,8(X0)DRepresent pixel X0The local concavity symbolic feature at place, CLCCP-M1,8(X0)DRepresent Pixel X0The local concavity amplitude characteristic at place, Represent X in image0The pixel value at place,WithRepresent P in imageiAnd Pi+4The pixel value at place,WithRepresent Q in imagejAnd Qj+4The pixel value at place, abs () Signed magnitude arithmetic(al), CLCCP-S are asked in expression1,8(X0)DAnd CLCCP-M1,8(X0)DMiddle subscript " 1 " represents to calculate used in convex-concave characteristic Pixel away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions convex-concave it is special Sign, it is decimal system amount that subscript " D ", which represents, and threshold is the threshold value pre-set；
Step4, each pixel to each image block encode, and obtain the center pixel feature CLCCP- of each image block C, coding formula are：Here cIThe average value of entire image is represented,Represent figure The X as in0The pixel value at place,CLCCP-C1,8(X0)DMiddle subscript " 1 " represents to calculate convex-concave characteristic Pixel used is away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions convex-concave Feature, subscript " D " represent decimal system amount, and threshold is the threshold value pre-set；
Step5, by step Step2, Step3 and Step4, be extracted image blockComplete local convexo-concave characteristic, bag Include and meet feature, amplitude characteristic and center pixel feature, work as image blockIn pixel X0When traveling through whole image block, obtain To each block imageCLCCP-S, CLCCP-M, CLCCP-C eigenmatrix, be respectively
Step6, the histogram feature vector for next extracting each three eigenmatrixes of image block, image blockThree EigenmatrixHistogram feature vector be expressed as：This three histogram feature vectors are sequentially connected, obtain image blockNogata Figure characteristic vectorHereinSubscript CLCCP represent it is complete Local Convex diesinking feature, it includes symbolic feature CLCCP-S, amplitude characteristic CLCCP-M and center pixel point feature CLCCP- C；
Step7, each image block of connection histogram feature vector, obtain the complete Local Convex diesinking Nogata of original image Figure characteristic vector is：
Step8, this feature vector be sent into the nearest neighbor classifier based on chi amount classified, it is original to identify The identity of facial image；
In the step Step8, when the nearest neighbor classifier based on chi amount is classified, chi is first calculated Amount；Set two width facial image I(0)And I(1)Complete Local Convex diesinking histogram feature vector be respectively：WithThen this two vectors The distance between, i.e., chi span is calculated from using equation below：
Wherein I(0) CLCCPAnd I (i)(1) CLCCP(i) texture feature vector I is represented respectively(0) CLCCPAnd I(1) CLCCPI-th yuan Element, K' represent the length of texture, and eps is a fixed value, are positive number minimum in Matlab, represent that causes an I(0) CLCCP (i)+I(1) CLCCP(i) the very small normal number of+eps ≠ 0.
The beneficial effects of the invention are as follows：
1st, the complete Local Convex diesinking facial image Texture Segmentation Algorithm that the present invention constructs is that one kind is based on image The texture description operator of second-order differential feature, overcome the defects of LBP can only describe image first differential information；
2nd, the complete Local Convex diesinking facial image Texture Segmentation Algorithm that the present invention constructs not only describes image The symbolic information of local convexo-concave characteristic, also describes the amplitude characteristic of image local convexo-concave characteristic, and considers picture centre The distinguishing ability of pixel, merge the distinctive that this three improves texture；
3rd, this method not only allows for the concavity of facial image local grain, it is also contemplated that image local texture concavity Size.Experiment in terms of recognition of face shows that computation complexity is low during algorithm progress illumination recognition of face, and accuracy of identification is high, There is insensitivity to illumination；
4th, two lines are used as using chi amount (Chi square statistic) in match cognization stage, the present invention The distance between characteristic vector measurement is managed, is classified using nearest neighbor classifier, algorithm is simple, convenience of calculation, can accomplish Real-time images match identification.
Brief description of the drawings
Fig. 1 is the complete Local Convex diesinking characteristic extraction step schematic block diagram of image in the present invention；
Fig. 2 is present invention pixel point X08 symmetry direction schematic diagrames；
Fig. 3 is 4 symmetry direction schematic diagrames of pixel in image of the present invention；
Fig. 4 is bilinear interpolation schematic diagram in the present invention；
Fig. 5 is 64 of a people in the extended YaleB face database illumination subsets used in the embodiment of the present invention Open sample image；
Fig. 6 is local binary patterns in the present invention, unified local binary patterns (Uniform local binary Pattern, Uniform LBP), complete local binary patterns and this method be on the extended Yale B datas storehouse Cumulative matching characteristic curve；
Fig. 7 is that local binary patterns in the present invention, unification local binary patterns, complete local binary patterns and this method exist Correct recognition rata curve on the extended Yale B datas storehouse.
Embodiment
Embodiment 1：As shown in figs. 1-7, a kind of illumination face recognition method based on complete Local Convex diesinking, it is right first Image carries out piecemeal；Then bilinear interpolation is carried out to each block image so that each pixel can build 8 symmetrically in image Direction, each pixel is then calculated in block image along 8 direction part difference；Then the symbol for encoding this local difference is special Seek peace amplitude characteristic；Each pixel of each image block is encoded, obtains the center pixel feature of each image block；Next It is vectorial to the eigenmatrix extraction histogram feature of the symbolic feature of each block image, amplitude characteristic, center pixel feature, successively Connect the block image symbolic feature, amplitude characteristic, the histogram feature vector of center pixel feature and obtain each block image Histogram feature vector；Finally connect each block image histogram feature vector obtain the histogram feature of this original image to Amount, this feature vector is sent into the nearest neighbor classifier based on chi amount and classified, to identify original facial image Identity.
The illumination face recognition method based on complete Local Convex diesinking comprises the following steps that：
Step1, image is subjected to piecemeal first：Image I(l)Uniformly be divided into 4 × 4 non-overlapping square, 16 pieces altogether, It is expressed as
Step2, bilinear interpolation computing is carried out to each block image so that each pixel can be built on the pixel Symmetrical 8 directions, then calculate local difference of each pixel along different directions, and the local difference is decomposed into symbol portion Point and and amplitude portion；
Step3, carry out corresponding Local Convex diesinking coding respectively to symbolic component and amplitude portion, obtain each block diagram The symbolic feature CLCCP-S and amplitude characteristic CLCCP-M of picture, wherein pixel X0Symbolic feature and amplitude characteristic coding it is public Formula is respectively：
Wherein, CLCCP-S1,8(X0)DRepresent pixel X0The local concavity symbolic feature at place, CLCCP-M1,8(X0)DRepresent Pixel X0The local concavity amplitude characteristic at place, Represent X in image0The pixel value at place,WithRepresent P in imageiAnd Pi+4The pixel value at place,WithRepresent Q in imagejAnd Qj+4The pixel value at place, abs () Signed magnitude arithmetic(al), CLCCP-S are asked in expression1,8(X0)DAnd CLCCP-M1,8(X0)DMiddle subscript " 1 " represents to calculate used in convex-concave characteristic Pixel away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions convex-concave it is special Sign, it is decimal system amount that subscript " D ", which represents, and threshold is the threshold value pre-set；
Step4, each pixel to each image block encode, and obtain the center pixel feature CLCCP- of each image block C, coding formula are：Here cIThe average value of entire image is represented,Represent figure The X as in0The pixel value at place,CLCCP-C1,8(X0)DMiddle subscript " 1 " represents to calculate convex-concave spy Pixel used in property is away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions it is convex Recessed feature, subscript " D " represent decimal system amount, and threshold is the threshold value pre-set；
Step5, by step Step2, Step3 and Step4, be extracted image blockComplete local convexo-concave characteristic, bag Include and meet feature, amplitude characteristic and center pixel feature, work as image blockIn pixel X0When traveling through whole image block, obtain Each block imageCLCCP-S, CLCCP-M, CLCCP-C eigenmatrix, be respectively
Step6, the histogram feature vector for next extracting each three eigenmatrixes of image block, image blockThree EigenmatrixHistogram feature vector be expressed as：This three histogram feature vectors are sequentially connected, obtain image blockNogata Figure characteristic vectorHereinSubscript CLCCP represent it is complete Local Convex diesinking feature, it includes symbolic feature CLCCP-S, amplitude characteristic CLCCP-M and center pixel point feature CLCCP- C；
Step7, each image block of connection histogram feature vector, obtain the complete Local Convex diesinking Nogata of original image Figure characteristic vector is：
Step8, this feature vector be sent into the nearest neighbor classifier based on chi amount classified, it is original to identify The identity of facial image.
In the step Step8, when the nearest neighbor classifier based on chi amount is classified, chi is first calculated Amount；Set two width facial image I(0)And I(1)Complete Local Convex diesinking histogram feature vector be respectively：WithThen this two vectors The distance between, i.e., chi span is calculated from using equation below：
Wherein I(0) CLCCPAnd I (i)(1) CLCCP(i) texture feature vector I is represented respectively(0) CLCCPAnd I(1) CLCCPI-th yuan Element, K' represent the length of texture, and eps is a fixed value, are positive number minimum in Matlab.
Embodiment 2：As shown in figs. 1-7, a kind of illumination face recognition method based on complete Local Convex diesinking, it is right first Image carries out piecemeal；Then bilinear interpolation is carried out to each block image so that each pixel can build 8 symmetrically in image Direction, each pixel is then calculated in block image along 8 direction part difference；Then the symbol for encoding this local difference is special Seek peace amplitude characteristic；Each pixel of each image block is encoded, obtains the center pixel feature of each image block；Next It is vectorial to the eigenmatrix extraction histogram feature of the symbolic feature of each block image, amplitude characteristic, center pixel feature, successively Connect the block image symbolic feature, amplitude characteristic, the histogram feature vector of center pixel feature and obtain each block image Histogram feature vector；Finally connect each block image histogram feature vector obtain the histogram feature of this original image to Amount, this feature vector is sent into the nearest neighbor classifier based on chi amount and classified, to identify original facial image Identity.
The illumination face recognition method based on complete Local Convex diesinking comprises the following steps that：
Step1, image is subjected to piecemeal first：Image I(l)Uniformly be divided into 4 × 4 non-overlapping square, 16 pieces altogether, It is expressed as
Step2, bilinear interpolation computing is carried out to each block image so that each pixel can be built on the pixel Symmetrical 8 directions, then calculate local difference of each pixel along different directions, and the local difference is decomposed into symbol portion Point and and amplitude portion；
As shown in Fig. 2 pixel P1And P2Between pixel Q can be increased by interpolation1.Interpolation method as shown in figure 4, Wherein P11, P12, P21, P22It is four neighbor pixels original in image, new pixel Q is inserted out by interpolation method0.Interpolation Formula is as follows：
WhereinWithR is represented respectively1, R2WithThe pixel value of opening position, x1, x and x2Pixel is represented respectively P11, R1And P21The abscissa at place, y1, y and y2Pixel P is represented respectively11, Q0And P12The ordinate at place.Fig. 3 represents original image Middle pixel X0P be present in surrounding0, P1, P2, P3, P4, P5, P6And P78 Neighbor Points, it is only capable of forming four on pixel X0It is symmetrical Direction.Fig. 2 represents pixel X after interpolation0Q be present in surrounding0, Q1, Q2, Q3, Q4, Q5, Q6And Q78 interpolation points, therefore pixel after interpolation X0Surrounding one co-exists in 16 Neighbor Points, can obtain 8 on pixel X0Symmetry direction.Due to adding interpolation point, to image Resolution ratio enhance；
Pixel X in image block0Local difference along 8 directions isWithWherein i=0,1,2,3 and j=0,1,2,3；
Step3, carry out corresponding Local Convex diesinking coding respectively to symbolic component and amplitude portion, obtain each block diagram The symbolic feature CLCCP-S and amplitude characteristic CLCCP-M of picture, wherein pixel X0Symbolic feature and amplitude characteristic coding it is public Formula is respectively：
Wherein, CLCCP-S1,8(X0)DRepresent pixel X0The local concavity symbolic feature at place, CLCCP-M1,8(X0)DRepresent Pixel X0The local concavity amplitude characteristic at place, Represent X in image0The pixel value at place,WithRepresent P in imageiAnd Pi+4The pixel value at place,WithRepresent Q in imagejAnd Qj+4The pixel value at place, abs () Signed magnitude arithmetic(al), CLCCP-S are asked in expression1,8(X0)DAnd CLCCP-M1,8(X0)DMiddle subscript " 1 " represents to calculate used in convex-concave characteristic Pixel away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions convex-concave it is special Sign, it is decimal system amount that subscript " D ", which represents, and threshold is the threshold value pre-set；
Step4, each pixel to each image block encode, and obtain the center pixel feature CLCCP- of each image block C, coding formula are：Here cIThe average value of entire image is represented,Represent figure The X as in0The pixel value at place,CLCCP-C1,8(X0)DMiddle subscript " 1 " represents to calculate convex-concave spy Pixel used in property is away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions it is convex Recessed feature, subscript " D " represent decimal system amount, and threshold is the threshold value pre-set；
Step5, by step Step2, Step3 and Step4, be extracted image blockComplete local convexo-concave characteristic, bag Include and meet feature, amplitude characteristic and center pixel feature, work as image blockIn pixel X0When traveling through whole image block, obtain To each block imageCLCCP-S, CLCCP-M, CLCCP-C eigenmatrix, be respectively
Step6, the histogram feature vector for next extracting each three eigenmatrixes of image block, image blockThree EigenmatrixHistogram feature vector be expressed as：This three histogram feature vectors are sequentially connected, obtain image blockNogata Figure characteristic vectorHereinSubscript CLCCP represent it is complete Local Convex diesinking feature, it includes symbolic feature CLCCP-S, amplitude characteristic CLCCP-M and center pixel point feature CLCCP- C；
Step7, each image block of connection histogram feature vector, obtain the complete Local Convex diesinking Nogata of original image Figure characteristic vector is：
Step8, this feature vector be sent into the nearest neighbor classifier based on chi amount classified, it is original to identify The identity of facial image.
In the step Step8, when the nearest neighbor classifier based on chi amount is classified, chi is first calculated Amount；Set two width facial image I(0)And I(1)Complete Local Convex diesinking histogram feature vector be respectively：WithThen this two vectors The distance between, i.e., chi span is calculated from using equation below：
Wherein I(0) CLCCPAnd I (i)(1) CLCCP(i) texture feature vector I is represented respectively(0) CLCCPAnd I(1) CLCCPI-th yuan Element, K' represent the length of texture, and eps is a fixed value, are positive number minimum in Matlab.
In order to prove the beneficial effect of methods described, by counting this method with other related algorithms in illumination human face data Discrimination in storehouse, cumulative match cognization rate simultaneously are compared to prove with other algorithms；
Discrimination of this method in illumination face database is counted first, and compared with related algorithm, draws phase Answer recognition performance curve.The present embodiment uses MATLAB software environments, and threshold takes 0 in the present embodiment, institute in the present embodiment With the illumination subset that face picture is the extended YaleB face databases, the subset shares 38 individuals, everyone 64 photos are shot under different light conditions, 2432 photos, photo size are 64x64 altogether, are the data as shown in Figure 5 64 samples pictures of a people in storehouse.The database can be in the database website (http://vision.ucsd.edu/~ Leekc/ExtYaleDatabase/ExtYaleB.html all face pictures cut are downloaded on).In this embodiment, Calculate the correct knowledge of this method, four kinds of local binary patterns, unified local binary patterns and complete local binary patterns algorithms Not rate and cumulative matching properties (cumulative match characteristic) curve.Counted using nearest neighbor classifier Discrimination is calculated, when calculating discrimination, training sample set is distinguished 1,2,3,4,5 samples by everyone and formed, and remaining image is used Test.Test sample is compared with all training samples, if the body of the training sample minimum with test sample distance Part is consistent with test sample, then it is assumed that identification is correct.All sample numbers correctly identified divided by all test sample numbers are i.e. For correct recognition rata.
Also calculate in addition this method, local binary patterns, complete local binary patterns, unified local binary patterns it is tired Add matching properties curve.Gallery picture libraries collection and Probe picture library collection are needed when calculating cumulative matching properties curve.Gallery Picture library collection by the extended Yale B datas storehouse everyone a pictures be provided formed, everyone remaining other 63 Picture forms Probe picture library collection.Assuming that it is L that Gallery picture libraries, which concentrate number of pictures, P is the full null vector that a length is L.Figure Storehouse Probe concentrates a pictures I to concentrate all pictures to enter row distance with Gallery and matches, obtain a distance vector D={ d1, d2,…,dL, if it is d that Probe, which concentrates picture I and Gallery to concentrate the distance between common identity picture, then d must be vector A D element, if by D vectors are arranged from small to large, it is assumed that d is arranged in D positions l, then on vectorial P l positions Element value adds 1.So Probe picture libraries concentration picture is repeated once, then by vectorial P each element divided by vectorial P Length, then " discrimination of order (rank) 1 " is exactly vectorial P first element value, and " discrimination of order (rank) 2 " is exactly vectorial P Second element value, the like.In the present embodiment, when calculating cumulative matching properties curve, at random from everyone photograph In piece select one composition Gallery picture library collection, everyone it is remaining 63 composition Probe collection.Local binary patterns, unified office The cumulative match curve of portion's binary pattern, complete local binary patterns, this method under this Gallery and Probe picture library collection is such as Shown in Fig. 6；
From fig. 6 it can be seen that the performance comparision of local binary patterns and unified local binary patterns approaches, but all it is weaker than Complete local binary patterns, and complete local binary patterns are substantially weaker than this method.As increasing to for " order (rank) " connects When nearly Gallery picture libraries concentrate number of pictures, several algorithm performances are close, but are now not much engineering practice value .
The correct recognition rata of each algorithm in the case of different training sample numbers is also simulated in the present embodiment.We will imitate Really it is repeated 5 times, calculates average correct recognition rata and standard deviation, and result is drawn in the figure 7；It is it can be seen from figure 7 that our The performance of method is significantly better than other several algorithms, when number of training is 5, is classified by the arest neighbors based on chi amount Device is to calculate local binary patterns, uniformly the average recognition rate of local binary patterns, complete local binary patterns and this method is： 66.09%, 62.34%, 70.92%, 76.76%.Wherein this method will be higher by than local binary patterns algorithm discrimination 10.67%, 14.42% is higher by than unified local binary patterns, 5.84% is higher by than complete local binary patterns, this explanation we Method is a kind of very efficient illumination face recognition method.
Above in conjunction with accompanying drawing to the present invention embodiment be explained in detail, but the present invention be not limited to it is above-mentioned Embodiment, can also be before present inventive concept not be departed from those of ordinary skill in the art's possessed knowledge Put that various changes can be made.

Claims (2)

  1. A kind of 1. illumination face recognition method based on complete Local Convex diesinking, it is characterised in that：Image is divided first Block；Then bilinear interpolation is carried out to each block image so that each pixel can build 8 symmetry directions in image, then Each pixel is calculated in block image along 8 direction part difference；Then the symbolic feature and amplitude of this local difference are encoded Feature；Each pixel of each image block is encoded, obtains the center pixel feature of each image block；Next to each piecemeal Symbolic feature, amplitude characteristic, the eigenmatrix extraction histogram feature vector of center pixel feature of image, are sequentially connected this point The histogram that block pictorial symbol feature, amplitude characteristic, the histogram feature vector of center pixel feature obtain each block image is special Sign vector；The histogram feature vector for finally connecting each block image obtains the histogram feature vector of this original image, this Characteristic vector is sent into the nearest neighbor classifier based on chi amount and classified, to identify the identity of original facial image；
    The illumination face recognition method based on complete Local Convex diesinking comprises the following steps that：
    Step1, image is subjected to piecemeal first：Image I(l)Uniformly it is divided into 4 × 4 non-overlapping square, 16 pieces altogether, represents For(i=0,1,2 ..., 15)；
    Step2, bilinear interpolation computing is carried out to each block image so that each pixel can be built on the pixel point symmetry 8 directions, then calculate local difference of each pixel along different directions, by the local difference be decomposed into symbolic component and And amplitude portion；
    Step3, carry out corresponding Local Convex diesinking coding respectively to symbolic component and amplitude portion, obtain each block image Symbolic feature CLCCP-S and amplitude characteristic CLCCP-M, wherein pixel X0Symbolic feature and amplitude characteristic coding formula point It is not：
    <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> <mo>-</mo> <msub> <mi>S</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>8</mn> </mrow> </msub> <msub> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>0</mn> </msub> <mo>)</mo> </mrow> <mi>D</mi> </msub> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>0</mn> </mrow> <mn>3</mn> </munderover> <mo>&amp;lsqb;</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>I</mi> <msub> <mi>X</mi> <mn>0</mn> </msub> </msub> <mo>-</mo> <mn>0.5</mn> <mo>*</mo> <mo>(</mo> <mrow> <msub> <mi>I</mi> <msub> <mi>P</mi> <mi>i</mi> </msub> </msub> <mo>+</mo> <msub> <mi>I</mi> <msub> <mi>P</mi> <mrow> <mi>i</mi> <mo>+</mo> <mn>4</mn> </mrow> </msub> </msub> </mrow> <mo>)</mo> <mo>)</mo> </mrow> <mo>*</mo> <msup> <mn>2</mn> <mrow> <mi>i</mi> <mo>*</mo> <mn>2</mn> </mrow> </msup> <mo>&amp;rsqb;</mo> <mo>+</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>0</mn> </mrow> <mn>3</mn> </munderover> <mo>&amp;lsqb;</mo> <mi>f</mi> <mrow> <mo>(</mo> <msub> <mi>I</mi> <msub> <mi>X</mi> <mn>0</mn> </msub> </msub> <mo>-</mo> <mn>0.5</mn> <mo>*</mo> <mo>(</mo> <mrow> <msub> <mi>I</mi> <msub> <mi>Q</mi> <mi>j</mi> </msub> </msub> <mo>+</mo> <msub> <mi>I</mi> <msub> <mi>Q</mi> <mrow> <mi>j</mi> <mo>+</mo> <mn>4</mn> </mrow> </msub> </msub> </mrow> <mo>)</mo> <mo>)</mo> </mrow> <mo>*</mo> <msup> <mn>2</mn> <mrow> <mi>j</mi> <mo>*</mo> <mn>2</mn> <mo>+</mo> <mn>1</mn> </mrow> </msup> <mo>&amp;rsqb;</mo> </mrow>
    <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> <mo>-</mo> <msub> <mi>M</mi> <mrow> <mn>1</mn> <mo>,</mo> <mn>8</mn> </mrow> </msub> <msub> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mn>0</mn> </msub> <mo>)</mo> </mrow> <mi>D</mi> </msub> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>0</mn> </mrow> <mn>3</mn> </munderover> <mo>&amp;lsqb;</mo> <mi>f</mi> <mrow> <mo>(</mo> <mi>a</mi> <mi>b</mi> <mi>s</mi> <mo>(</mo> <mrow> <msub> <mi>I</mi> <msub> <mi>X</mi> <mn>0</mn> </msub> </msub> <mo>-</mo> <mn>0.5</mn> <mo>*</mo> <mrow> <mo>(</mo> <mrow> <msub> <mi>I</mi> <mi>P</mi> </msub> <mo>+</mo> <msub> <mi>I</mi> <msub> <mi>P</mi> <mrow> <mi>i</mi> <mo>+</mo> <mn>4</mn> </mrow> </msub> </msub> </mrow> <mo>)</mo> </mrow> </mrow> <mo>)</mo> <mo>-</mo> <mi>c</mi> <mo>)</mo> </mrow> <mo>*</mo> <msup> <mn>2</mn> <mrow> <mi>i</mi> <mo>*</mo> <mn>2</mn> </mrow> </msup> <mo>&amp;rsqb;</mo> <mo>+</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>j</mi> <mo>=</mo> <mn>0</mn> </mrow> <mn>3</mn> </munderover> <mo>&amp;lsqb;</mo> <mi>f</mi> <mrow> <mo>(</mo> <mi>a</mi> <mi>b</mi> <mi>s</mi> <mo>(</mo> <mrow> <msub> <mi>I</mi> <msub> <mi>X</mi> <mn>0</mn> </msub> </msub> <mo>-</mo> <mn>0.5</mn> <mo>*</mo> <mrow> <mo>(</mo> <mrow> <msub> <mi>I</mi> <msub> <mi>Q</mi> <mi>j</mi> </msub> </msub> <mo>+</mo> <msub> <mi>I</mi> <msub> <mi>Q</mi> <mrow> <mi>j</mi> <mo>+</mo> <mn>4</mn> </mrow> </msub> </msub> </mrow> <mo>)</mo> </mrow> </mrow> <mo>)</mo> <mo>-</mo> <mi>c</mi> <mo>)</mo> </mrow> <mo>*</mo> <msup> <mn>2</mn> <mrow> <mi>j</mi> <mo>*</mo> <mn>2</mn> <mo>+</mo> <mn>1</mn> </mrow> </msup> <mo>&amp;rsqb;</mo> </mrow>
    Wherein, CLCCP-S1,8(X0)DRepresent pixel X0The local concavity symbolic feature at place, CLCCP-M1,8(X0)DRepresent pixel X0The local concavity amplitude characteristic at place, Represent X in image0The pixel value at place,WithRepresent P in imageiAnd Pi+4The pixel value at place,WithRepresent Q in imagejAnd Qj+4The pixel value at place, abs () represent to ask Take absolute value computing, CLCCP-S1,8(X0)DAnd CLCCP-M1,8(X0)DMiddle subscript " 1 " represents to calculate the pixel used in convex-concave characteristic Point is away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions convexo-concave characteristic, subscript It is decimal system amount that " D ", which is represented, and threshold is the threshold value pre-set；
    Step4, each pixel to each image block encode, and obtain the center pixel feature CLCCP-C of each image block, compile Code formula be：Here cIThe average value of entire image is represented,Represent X in image0 The pixel value at place,CLCCP-C1,8(X0)DMiddle subscript " 1 " represents to calculate used in convex-concave characteristic Pixel away from X0Distance be 1, i.e., yardstick be 1, subscript " 8 " represent calculate passes through pixel X08 directions convex-concave it is special Sign, subscript " D " represent decimal system amount, and threshold is the threshold value pre-set；
    Step5, by step Step2, Step3 and Step4, be extracted image blockComplete local convexo-concave characteristic, including symbol Feature, amplitude characteristic and center pixel feature are closed, works as image blockIn pixel X0When traveling through whole image block, each point is obtained Block imageCLCCP-S, CLCCP-M, CLCCP-C eigenmatrix, be respectively
    Step6, the histogram feature vector for next extracting each three eigenmatrixes of image block, image blockThree feature squares Battle arrayHistogram feature vector be expressed as：This three histogram feature vectors are sequentially connected, obtain image blockNogata Figure characteristic vectorHereinSubscript CLCCP represent it is complete Local Convex diesinking feature, it includes symbolic feature CLCCP-S, amplitude characteristic CLCCP-M and center pixel point feature CLCCP- C；
    Step7, each image block of connection histogram feature vector, the complete Local Convex diesinking histogram for obtaining original image are special Levying vector is：
    Step8, this feature vector be sent into the nearest neighbor classifier based on chi amount classified, to identify original face The identity of image.
  2. 2. the illumination face recognition method according to claim 1 based on complete Local Convex diesinking, it is characterised in that：Institute State in step Step8, when the nearest neighbor classifier based on chi amount is classified, first calculate chi amount；Setting two Width facial image I(0)And I(1)Complete Local Convex diesinking histogram feature vector be respectively：WithThen this two vectors The distance between, i.e., chi span is calculated from using equation below：
    <mrow> <msup> <mi>&amp;chi;</mi> <mn>2</mn> </msup> <mrow> <mo>(</mo> <msub> <msup> <mi>I</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msup> <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> </mrow> </msub> <mo>,</mo> <msub> <msup> <mi>I</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msup> <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> </mrow> </msub> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mo>&amp;Sigma;</mo> <mrow> <mi>i</mi> <mo>=</mo> <mn>1</mn> </mrow> <mi>K</mi> </munderover> <mfrac> <msup> <mrow> <mo>(</mo> <msub> <msup> <mi>I</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msup> <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> </mrow> </msub> <mo>(</mo> <mi>i</mi> <mo>)</mo> <mo>-</mo> <msub> <msup> <mi>I</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msup> <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> </mrow> </msub> <mo>(</mo> <mi>i</mi> <mo>)</mo> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mrow> <mo>(</mo> <msub> <msup> <mi>I</mi> <mrow> <mo>(</mo> <mn>0</mn> <mo>)</mo> </mrow> </msup> <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> </mrow> </msub> <mo>(</mo> <mi>i</mi> <mo>)</mo> <mo>+</mo> <msub> <msup> <mi>I</mi> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </msup> <mrow> <mi>C</mi> <mi>L</mi> <mi>C</mi> <mi>C</mi> <mi>P</mi> </mrow> </msub> <mo>(</mo> <mi>i</mi> <mo>)</mo> <mo>+</mo> <mi>e</mi> <mi>p</mi> <mi>s</mi> <mo>)</mo> </mrow> </mfrac> </mrow>
    Wherein I(0) CLCCPAnd I (i)(1) CLCCP(i) texture feature vector I is represented respectively(0) CLCCPAnd I(1) CLCCPI-th of element, K' The length of texture is represented, eps is a fixed value, is positive number minimum in Matlab.
CN201510223240.2A 2015-05-05 2015-05-05 A kind of illumination face recognition method based on complete Local Convex diesinking Expired - Fee Related CN104881634B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510223240.2A CN104881634B (en) 2015-05-05 2015-05-05 A kind of illumination face recognition method based on complete Local Convex diesinking

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510223240.2A CN104881634B (en) 2015-05-05 2015-05-05 A kind of illumination face recognition method based on complete Local Convex diesinking

Publications (2)

Publication Number Publication Date
CN104881634A CN104881634A (en) 2015-09-02
CN104881634B true CN104881634B (en) 2018-02-09

Family

ID=53949122

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510223240.2A Expired - Fee Related CN104881634B (en) 2015-05-05 2015-05-05 A kind of illumination face recognition method based on complete Local Convex diesinking

Country Status (1)

Country Link
CN (1) CN104881634B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107292230B (en) * 2017-05-09 2020-07-28 华南理工大学 Embedded finger vein identification method based on convolutional neural network and having counterfeit detection capability
CN109410258B (en) * 2018-09-26 2021-12-10 重庆邮电大学 Texture image feature extraction method based on non-local binary pattern
CN110059606A (en) * 2019-04-11 2019-07-26 新疆大学 A kind of improved increment Non-negative Matrix Factorization face recognition algorithms

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101317183A (en) * 2006-01-11 2008-12-03 三菱电机株式会社 Method for localizing pixels representing an iris in an image acquired of an eye
CN101551858A (en) * 2009-05-13 2009-10-07 北京航空航天大学 Target recognition method based on differential code and differential code mode
CN101835037A (en) * 2009-03-12 2010-09-15 索尼株式会社 Method and system for carrying out reliability classification on motion vector in video
CN102722699A (en) * 2012-05-22 2012-10-10 湖南大学 Face identification method based on multiscale weber local descriptor and kernel group sparse representation
CN103246880A (en) * 2013-05-15 2013-08-14 中国科学院自动化研究所 Human face recognizing method based on multi-level local obvious mode characteristic counting

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
TWI454248B (en) * 2008-09-23 2014-10-01 Ind Tech Res Inst Method of multi-dimensional empirical mode decomposition for image morphology

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101317183A (en) * 2006-01-11 2008-12-03 三菱电机株式会社 Method for localizing pixels representing an iris in an image acquired of an eye
CN101835037A (en) * 2009-03-12 2010-09-15 索尼株式会社 Method and system for carrying out reliability classification on motion vector in video
CN101551858A (en) * 2009-05-13 2009-10-07 北京航空航天大学 Target recognition method based on differential code and differential code mode
CN102722699A (en) * 2012-05-22 2012-10-10 湖南大学 Face identification method based on multiscale weber local descriptor and kernel group sparse representation
CN103246880A (en) * 2013-05-15 2013-08-14 中国科学院自动化研究所 Human face recognizing method based on multi-level local obvious mode characteristic counting

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
《基于凹凸局部二值模式的纹理图像分类》;毋小省 等;《光电子·激光》;20140831;第25卷(第8期);全文 *
《基于纹理与特征选择的前视红外目标识别》;毋小省 等;《光电子·激光》;20141130;第25卷(第11期);全文 *

Also Published As

Publication number Publication date
CN104881634A (en) 2015-09-02

Similar Documents

Publication Publication Date Title
CN108537743B (en) Face image enhancement method based on generation countermeasure network
CN109543606B (en) Human face recognition method with attention mechanism
Gilani et al. Learning from millions of 3D scans for large-scale 3D face recognition
Cherian et al. Riemannian dictionary learning and sparse coding for positive definite matrices
CN110728209B (en) Gesture recognition method and device, electronic equipment and storage medium
CN106529468B (en) A kind of finger vein identification method and system based on convolutional neural networks
Bayraktar et al. Analysis of feature detector and descriptor combinations with a localization experiment for various performance metrics
CN106778796B (en) Human body action recognition method and system based on hybrid cooperative training
CN107194341A (en) The many convolution neural network fusion face identification methods of Maxout and system
Zheng et al. Attention-based spatial-temporal multi-scale network for face anti-spoofing
CN102663400B (en) LBP (length between perpendiculars) characteristic extraction method combined with preprocessing
CN104915673B (en) A kind of objective classification method and system of view-based access control model bag of words
CN104268593A (en) Multiple-sparse-representation face recognition method for solving small sample size problem
CN105956560A (en) Vehicle model identification method based on pooling multi-scale depth convolution characteristics
CN107729993A (en) Utilize training sample and the 3D convolutional neural networks construction methods of compromise measurement
CN110097029B (en) Identity authentication method based on high way network multi-view gait recognition
Zhou et al. Pose-robust face recognition with Huffman-LBP enhanced by divide-and-rule strategy
CN104881676B (en) A kind of facial image convex-concave pattern texture feature extraction and recognition methods
Zhang et al. Gait-based age estimation with deep convolutional neural network
CN104778472B (en) Human face expression feature extracting method
CN106529586A (en) Image classification method based on supplemented text characteristic
CN104881634B (en) A kind of illumination face recognition method based on complete Local Convex diesinking
CN110458178A (en) The multi-modal RGB-D conspicuousness object detection method spliced more
CN104200233A (en) Clothes classification and identification method based on Weber local descriptor
CN106778714A (en) LDA face identification methods based on nonlinear characteristic and model combination

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
CB03 Change of inventor or designer information
CB03 Change of inventor or designer information

Inventor after: Chen Xi

Inventor after: Jin Jie

Inventor after: Pan Xiaolu

Inventor before: Chen Xi

Inventor before: Jin Jie

GR01 Patent grant
GR01 Patent grant
CF01 Termination of patent right due to non-payment of annual fee
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20180209