CN114780769B - Personnel checking method based on bloom filter - Google Patents
Personnel checking method based on bloom filter Download PDFInfo
- Publication number
- CN114780769B CN114780769B CN202210662860.6A CN202210662860A CN114780769B CN 114780769 B CN114780769 B CN 114780769B CN 202210662860 A CN202210662860 A CN 202210662860A CN 114780769 B CN114780769 B CN 114780769B
- Authority
- CN
- China
- Prior art keywords
- image
- iris
- fingerprint
- rectangular
- calculating
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/53—Querying
- G06F16/532—Query formulation, e.g. graphical querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/55—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Mathematical Physics (AREA)
- Collating Specific Patterns (AREA)
Abstract
The invention discloses a bloom filter-based personnel checking method, which is used for defining detail characteristic areas of fingerprint images, face images and iris images in a gridding mode and calculating、、、The method comprises the steps of representing detail feature differences, and dividing each person under the same crowd classification into corresponding data subsets according to the differences, wherein the divided data subsets are more scientific; calculating by using the corresponding relation between the bit array point position and the deviation degree range of the bloom filter、、、The deviation degree of each of the first to fourth image sets can be quickly matched with the image set corresponding to the corresponding point according to the deviation degree range in which the deviation degree falls, the acquisition speed of the image set serving as a basis for personnel checking and matching is improved, and the method is favorable for acquiring the image setsThe speed of personnel checking is further improved; the preset matching rules are used for checking and matching layer by layer, so that the matching accuracy is ensured, and meanwhile, the matching speed is considered.
Description
Technical Field
The invention relates to the technical field of personnel checking, in particular to a personnel checking method based on a bloom filter.
Background
At present, the information means of people verification which is relatively common is mainly fingerprint identification, face identification and iris identification, and the verification mode is generally that the fingerprint information or face information or iris information of people to be verified is compared with the fingerprint/face/iris information of people in a database, and when the comparison is successful, the comparison result is output. However, the database integrates the identity recognition features of all people groups of different ages, sexes and the like, and when identity verification is performed on specific personnel, particularly on specific batch personnel, the comparison process needs to take a long time. In order to solve the technical problem, in the prior art, people are generally classified by defining a data subset, and when people belonging to the same class are subjected to identity verification, identity characteristics of the people are firstly compared with identity characteristics of all people in the data subset, so that the comparison speed is increased. However, the simple incorporation of the identity characteristics of persons identified as the same class into the same data subset does not take into account the detail differences in identity characteristics between different persons, and this kind of determination is not scientific enough, and directly affects the comparison speed and the comparison accuracy.
Disclosure of Invention
The invention provides a bloom filter-based personnel checking method aiming at improving the identity comparison speed and the comparison accuracy of personnel checking.
In order to achieve the purpose, the invention adopts the following technical scheme:
the personnel checking method based on the bloom filter comprises the following steps:
s1, calculating eachDistance value and value of fingerprint imageOr accumulated value of quantity(ii) a And calculating the ratio and value of each face image(ii) a And calculating the sum ratio of each iris image;
S2, drawing each fingerprint image intoThe first image set corresponding to the falling sum value interval or each fingerprint image is marked intoThe second image set corresponding to the falling number interval; drawing each face image intoThe third image set corresponding to the falling sum value interval; drawing each iris image intoThe fourth image set corresponding to the falling ratio interval;
s3, based on a preset deviation degree range, mapping each of the first to fourth image sets as elements of a bloom filter to corresponding points of a bit array;
s4, calculating the corresponding fingerprint image in each first image setAverage value of (1), noteIs composed of(ii) a And calculating the corresponding fingerprint image in each second image setIs the average value of(ii) a And calculating the corresponding face image in each third image setIs the average value of(ii) a And calculating the corresponding of each iris image in each fourth image setIs the average value of;
S5, acquiring the fingerprint image, the face image and the iris image of the person to be checked, and then calculating the distance value and the value of the fingerprint image of the person to be checkedOr accumulated value of quantityAnd calculating the ratio and value of the face image of the person to be checkedAnd calculating the sum ratio of the iris images of the person to be checked;
s7, matching an image set corresponding to the point location according to the corresponding relation between the deviation degree range in which the deviation degree falls and the point location in the point location array;
and S8, according to a preset personnel checking method, performing personnel checking matching on the acquired fingerprint image, face image and iris image of the personnel to be checked and each image set matched in the step S7 according to a preset matching rule, and outputting a checking result.
Preferably, in step S1, the distance value and the value of each fingerprint imageThe method is calculated by the following method steps:
a1, equally dividing each side into a plurality of sections in an equally-spaced mode for the width and the height of each fingerprint image classified by the same crowd framed in a rectangular frame selection mode;
a2, starting from each bisector, connecting lines to the opposite bisector of the opposite side in a mode of being perpendicular to the side where the starting point is located, so as to disperse the fingerprint image into a plurality of rectangular sub-blocks, taking the rectangular sub-block at the left top corner of the image as a starting sub-block of a standard sequence, and marking each rectangular sub-block according to a convolution sequence in a mode of circling the standard sequence anticlockwise inwards;
a3, filtering the rectangular subblocks which do not carry fingerprint information and are fully loaded with fingerprint information in the fingerprint image, wherein the fully loaded fingerprint information indicates that pixels representing the fingerprint information touch each side of the rectangular subblocks to which the pixels belong;
a4, filtering each of the remaining rectangular sub-blocks in the step A3, and searching boundary pixel points of fingerprint information in each rectangular sub-block;
a5, calculating the distance between each boundary pixel point and the left vertex of the rectangular sub-block, and summing the distances according to the following formula (1) to obtain the distance value corresponding to each rectangular sub-block:
In the formula (1), the first and second groups,representing the second in the fingerprint imageA distance value of each of the rectangular sub-blocks;
denotes the firstThe first of the rectangular sub-blocksThe distance between each boundary pixel point and the left vertex of the rectangular sub-block;
a6, calculating the distance value and the value of all the rectangular sub-blocks remaining after the filtering of step A3 by the following formula (2):
In the formula (2), the first and second groups of the compound,to representIn the calculation ofThe weight occupied by the hour;
Preferably, in step S1, each of the fingerprint imagesThe method is calculated by the following method steps:
b1 calculated for each of the fingerprint image and the standard fingerprint imageAnd the two rectangular sub-blocks having the same rank number calculate the difference in distance value by the following formula (3):
In the formula (3), the first and second groups of the compound,indicating participationA calculated number one of the fingerprint imagesA distance value of each of the rectangular sub-blocks;
indicating participationThe calculated second one of the standard fingerprint image and the fingerprint imageThe first sub-blocks of the rectangle have the same rank numberA distance value of each of the rectangular sub-blocks;
b2, pairThe rectangular subblocks in the fingerprint image which are smaller than the difference value threshold are listed as quantity accumulation objects, and each fingerprint image conforms to the requirementsThe quantity accumulation of the rectangular blocks under the quantity accumulation condition is carried out to obtain the quantity accumulation value associated with each fingerprint image。
Preferably, each rectangular sub-block in the standard fingerprint image is divided into a plurality of sub-blocksThe value is calculated by the following equation (4):
in the formula (4), the first and second groups,representing a first of said first or second set of imagesOpening the sum in the fingerprint imageThe corresponding rectangular sub-blocks have the distance values of the rectangular sub-blocks with the same row sequence number;
representing a number of the fingerprint images stored in the first image set or the second image set.
Preferably, in step S1, each of the face imagesThe method is calculated by the following method steps:
c1, shooting the face of each person under the same crowd classification at a fixed distance and a fixed angle to obtain the face image of each person with the same size;
c2, equally dividing each edge into a plurality of segments in an equally-spaced mode for the width and the height of each face image;
c3, starting from each bisector, connecting lines to the opposite bisectors of the opposite sides in a mode of being perpendicular to the side where the starting point is located, so as to disperse the face image into a plurality of rectangular blocks, using the rectangular block at the left top corner of the image as a starting block for marking, and marking each rectangular block according to a circling sequence in a mode of circling the rectangular blocks anticlockwise inwards to mark;
c4, filtering out the rectangular blocks which do not carry face information and are fully loaded with the face information in the face image, wherein the fully loaded face information means that pixels representing the face information touch each edge of the rectangular block to which the pixels belong;
c5, searching the face pixels in each rectangular block which is filtered by the step C4 and calculating the number of the searched face pixels and the number of the face pixels in the face imageThe ratio of the number of the pixel points in each rectangular block is recorded as;
C6, calculating the ratio and value of all the rectangular blocks remaining after the filtering of step C4 by the following equation (5):
In the formula (5), the first and second groups of the chemical reaction materials are selected from the group consisting of,to representIn the calculation ofThe weight occupied by (c);
Preferably, in step S1, each of the iris imagesThe method is calculated by the following method steps:
d1, shooting eye images at a fixed distance and a fixed angle for each person under the same crowd classification, and framing out an iris image from each eye image in a rectangular frame selection mode;
d2, halving each side of the width and the height of each iris image, and connecting the unequal points which are not opposite to each other to obtain a space quadrangle;
d3, calculating the area of the space quadrangle and the rectangular frame of the iris image, and respectively recording as、;
D4, halving each side of the space quadrangle, and then dividing each side from each halving pointConnecting lines to the iris boundaries of the iris images in a manner of being vertical to the edges, and marking the connected points as vertexes,、Respectively represent the first on the space quadrangleBisector point of the sides, and from the bisector pointVertex connecting to iris boundary;
d5, from vertexTo the space quadrilateral at firstConnecting two end points of the sides to obtain a triangle, and marking asTriangle shapeWill be provided withThe outer iris region is separated into two arc iris regions, which are respectively marked as、;
D7 in the form of the triangleThe two waists areAnd the edges are used for equally dividing each waist, connecting lines from the equal division points to the iris boundary of the arc iris region in a mode of being vertical to the waist to obtain connection vertexes, and connecting the connection vertexes to the triangleThe two end points of the waist are connected to obtain a triangle, and the triangle is marked as,q=1 or 2;
D9 in the form of the triangleThe two waists areAnd (3) connecting the lines to obtain triangles and calculating the areas of the triangles by the method described in the steps D7-D8 until reaching the preset number of times of triangle construction, and then calculating the area of the iris image by the following formula (6):
In the formula (6), the first and second groups of the compound,representing that the first division of any division block of the iris image is performed by taking the central point of the iris image as the origin of an XY axis coordinate systemNext toDividing the edges equally;
representing pairs of arc-shaped iris areasOr to arc-shaped iris areasTo carry outNumber of edge equal divisions;
representing pairs of arc-shaped iris areasOr to arc-shaped iris areasTo proceed withThe edges being equally spacedThe number of edges;
Preferably, in step S6, the calculation is performed by the following formula (7)、、、Degree of deviation from each of the first to fourth image sets:
in the formula (7), the first and second groups,to representThe degree of deviation from each first image set;
Preferably, in step S8, the human audit matching is performed by a matching rule expressed by the following method steps:
e1, judging whether the first image set or the second image set is matched as the checking basis for the person to be checked,
if yes, go to step E2;
if not, jumping to the step E3;
e2, fingerprint comparison is carried out on the fingerprint image of the person to be checked and each fingerprint image in the first image set or the second image set which is matched with the fingerprint image of the person to be checked,
if the comparison is successful, outputting a fingerprint comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E3;
e3, determining whether the third image set is matched as the checking basis for the person to be checked,
if yes, go to step E4;
if not, jumping to step E5;
e4, comparing the face image of the person to be checked with each face in the third image set,
if the comparison is successful, outputting a face comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E5;
e5, determining whether the fourth image set is matched as the checking basis for the person to be checked,
if yes, go to step E6;
if not, jumping to step E7;
e6, comparing the iris image of the person to be checked with each iris image in the fourth image set,
if the comparison is successful, outputting an iris comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E7;
e7, comparing the fingerprint image of the person to be checked with the fingerprints of all the first image sets or the second image sets, comparing the face image of the person to be checked with the faces of all the third image sets, comparing the iris image of the person to be checked with the irises of all the fourth image sets,
if any item is successfully compared, outputting a comparison result and terminating the personnel checking process, otherwise, outputting a comparison failure result.
The invention has the following beneficial effects:
1. defining the detail characteristic regions of fingerprint image, human face image and iris image in a gridding mode, and calculating、、、The detail feature differences are characterized, and each person under the same crowd classification is classified into a corresponding data subset (namely each image set) according to the differences, and the classified data subset is more scientific;
2. calculating by using the corresponding relation between the bit array point position and the deviation degree range of the bloom filter、、、The deviation degree of each of the first to fourth image sets can be matched with the image set corresponding to the corresponding point position quickly according to the deviation degree range in which the deviation degree falls, so that the acquisition speed of the image set serving as a personnel checking matching basis is increased, and the personnel checking speed is further increased;
3. and the preset matching rules are utilized, the identity matching is carried out on the personnel to be checked layer by layer, the matching result is output and the subsequent matching process is terminated as long as a certain layer is successfully matched, and the matching accuracy is ensured while the matching speed is considered.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required to be used in the embodiments of the present invention will be briefly described below. It is obvious that the drawings described below are only some embodiments of the invention, and that for a person skilled in the art, other drawings can be derived from them without inventive effort.
FIG. 1 is a diagram illustrating implementation steps of a bloom filter-based people checking method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a fingerprint image being equally divided into a number of rectangular sub-blocks and the rectangular sub-blocks being sorted;
FIG. 3 is a schematic diagram of calculating the distance between the boundary pixel point and the left vertex of the rectangular sub-block in the fingerprint image;
Detailed Description
The technical scheme of the invention is further explained by the specific implementation mode in combination with the attached drawings.
Wherein the showings are for the purpose of illustration only and are shown by way of illustration only and not in actual form, and are not to be construed as limiting the present patent; to better illustrate the embodiments of the present invention, some parts of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted.
The same or similar reference numerals in the drawings of the embodiments of the present invention correspond to the same or similar components; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right", "inner", "outer", etc. are used for indicating the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, it is only for convenience of description and simplification of description, but it is not indicated or implied that the referred device or element must have a specific orientation, be constructed in a specific orientation and be operated, and therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and are not to be construed as limitations of the present patent, and the specific meanings of the terms may be understood by those skilled in the art according to specific situations.
In the description of the present invention, unless otherwise explicitly specified or limited, the term "connected" or the like, if appearing to indicate a connection relationship between components, is to be understood broadly, for example, as being either fixedly connected, detachably connected, or integrated; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be connected through any combination of two or more members or structures. The specific meanings of the above terms in the present invention can be understood in specific cases to those skilled in the art.
The personnel checking method based on the bloom filter provided by the embodiment of the invention is described in figure 1, and comprises the following steps:
s1, calculating the distance value and the value of each fingerprint imageOr accumulated value of quantity(ii) a And calculating the ratio and value of each face image(ii) a And calculating the sum ratio of each iris image;
Distance value and value of each fingerprint imageThe method is calculated by the following method steps:
a1, equally dividing each edge into a plurality of segments at equal intervals for the width and the height of each fingerprint image under the same crowd classification (such as male teenagers with age groups of 13-17) framed in a rectangular frame selection mode, and referring to fig. 2 for an example of equal division;
a2, starting from each bisector, connecting lines to the opposite bisector of the opposite side in a manner of being perpendicular to the side where the starting point is located, so as to disperse the fingerprint image into a plurality of rectangular sub-blocks, taking the rectangular sub-block at the left top corner of the image as an initial sub-block of a standard sequence (the sub-block with the sequence of '1' in the figure 2 represents the initial sub-block), and marking each rectangular sub-block according to a convolution sequence in a manner of circling the standard sequence counterclockwise inwards;
a3, filtering out the fingerprint information (which is the pixels carrying the fingerprint information and not representing the fingerprint information in the sub-blocks, such as the rectangular sub-blocks with the sequence of "1", "2", "12", "18", "29" in FIG. 2) and the rectangular sub-blocks with full fingerprint information (which is the rectangular sub-blocks with the sequence of "8", "36", "22", "24" in FIG. 2) which indicate that the pixels with full fingerprint information touch each side of the sub-blocks);
a4, for each of the rectangular sub-blocks filtered in step A3, searching boundary pixels of the fingerprint information in each rectangular sub-block (as indicated by marks P1 and P2 in fig. 3), where there are many existing methods for searching boundary pixels, such as identifying a point where the fingerprint information inside the rectangular sub-block is interrupted as a boundary pixel;
a5, calculating the distance (as represented by L1 and L2 in figure 2) between each boundary pixel point and the left vertex (as represented by the mark P0 in figure 2) of the rectangular sub-block, and summing each distance according to the following formula (1) to obtain the distance value corresponding to each rectangular sub-block:
In the formula (1), the first and second groups,representing the second in a fingerprint imageDistance values of the individual rectangular sub-blocks;
is shown asThe first of the rectangular sub-blocksThe distance between each boundary pixel point and the left vertex of the rectangular sub-block;
a6, calculating the distance value and the value of all the rectangular sub-blocks remaining after the filtering of the step A3 by the following formula (2):
In the formula (2), the first and second groups,to representIn the calculation ofThe weight occupied by the hour;
It should be noted that, the degree of influence of the fingerprint information in each of the rectangular sub-blocks remaining after the filtering in step a3 on the accuracy of the fingerprint comparison result is not the same, for example, the rectangular sub-block with the "4" order in fig. 2 has richer fingerprint features than the rectangular sub-block with the "3" order, so the distance value and the value are the sameTime, can give the bigger weight of the rectangle subblock that the standard order is "4", carry out the image set to the fingerprint image and divide the in-process, through considering the different detail characteristic of fingerprint edge to the different influence degree of fingerprint identification rate of accuracy, be favorable to promoting the division meticulous degree of fingerprint image set, and then be favorable to promoting subsequent fingerprint and compare speed and the degree of accuracy.
The method for dividing the fingerprint image into the corresponding first image set comprises the steps of firstly judging the fingerprint imageAnd then according to the corresponding relation between the preset sum value interval and the corresponding first image set, dividing the fingerprint image into the corresponding first image set. For example, a predetermined sum value interval is 100-150, the sum value interval corresponds to the first image set 1, if the fingerprint image has120, then it falls exactly into the sum interval and the fingerprint image is then scored into the first image set 1. For another example, if a certain fingerprint imageIf the sum value interval is 180 and falls into the preset sum value interval of 150-200, and the sum value interval corresponds to the first image set 2, the fingerprint image is classified into the first image set 2.
This embodiment provides two different fingerprint image clustering methods, one methodIs divided according to anotherThe division is made on the basis. To be provided withThe method for dividing according to the method comprises the following steps:
b1 calculated for each fingerprint image and standard fingerprint imageAnd the distance value difference is calculated by the following formula (3) for two rectangular sub-blocks with the same row number:
In the formula (3), the first and second groups,indicating participationIn the calculated fingerprint imageDistance values of the individual rectangular sub-blocks;
indicating participationIn the calculated standard fingerprint image and in the fingerprint imageThe first rectangular sub-blocks having the same row numberDistance values of the individual rectangular sub-blocks;
it should be noted here that although each image is calculatedAndwhen the people belonging to each fingerprint image are defined as a group under the same classification, such as a group of male teenagers in the ages of 13-18, so as to reduce the influence of different people on the sorting of the rectangular sub-blocks due to different fingerprint sizes as much as possible, but the problem that the sorting of the two rectangular sub-blocks with corresponding positions in the fingerprint image and the standard fingerprint image is the same is difficult to achieve, for example, the rectangular sub-block with the sequence of 4 in fig. 2 has a corresponding relation with the rectangular sub-block with the sequence of 5 in the standard fingerprint image, but the sorting numbers of the two rectangular sub-blocks are different, one is 4, the other is 4, and the calculation is performed according to the step B1According to the rule, the two rectangular sub-blocks cannot be corresponded, in order to solve the problem, each fingerprint image is firstly amplified to be the same in size, then the fingerprint image is selected by using the rectangular frame with the same width and height and the central point of the fingerprint image as the frame-selected central point frame, and therefore it is ensured that each fingerprint image has the same number of rectangular sub-blocks and the same row serial number corresponding to the position.
And of each rectangular sub-block in the standard fingerprint imageThe value is calculated by the following formula (4):
in the formula (4), the first and second groups,representing the first image set or the second image setIn a fingerprint imageThe corresponding rectangular sub-blocks have the distance values of the rectangular sub-blocks with the same row sequence number;
representing the number of fingerprint images stored in the first image set or the second image set.
Due to each rectangular sub-block in the standard fingerprint imageThe value is the average of the distance values of the corresponding rectangular sub-blocks of all the fingerprint images in the first image set or the second image set, thereby realizing that each rectangular sub-block in the standard index image has a corresponding rectangular sub-block in each fingerprint image.
B2, pairThe rectangular sub-block columns in the fingerprint image which are less than the difference value threshold value are quantity accumulation objects (The smaller the size, the higher the similarity of two rectangular sub-blocks with corresponding position), and the fingerprint image is matched with the two rectangular sub-blocksNumber accumulation barThe number of each rectangular block of the fingerprint image is accumulated to obtain the accumulated number value associated with each fingerprint image。
c1, shooting the face of each person under the same crowd classification at a fixed distance and a fixed angle to obtain the face image of each person with the same size;
c2, equally dividing each side into a plurality of sections in an equally-spaced mode for the width and the height of each face image;
c3, starting from each bisector, connecting lines to the opposite bisectors of the opposite sides in a mode of being perpendicular to the side where the starting point is located, so as to disperse the face image into a plurality of rectangular blocks, using the rectangular block at the left top corner of the image as a starting block for marking, and marking each rectangular block according to a circling sequence in a mode of circling the rectangular blocks anticlockwise inwards to mark;
c4, filtering out rectangular blocks which do not carry face information and are fully loaded with the face information in the face image, wherein the fully loaded face information means that pixels representing the face information touch each edge of the rectangular block to which the pixels belong;
the human face image discretization method adopted in the steps C2-C4 is the same as the discretization method of the fingerprint image recorded in the steps A1-A3, and therefore, the description is omitted.
C5, searching the face pixels in each rectangular block filtered by the step C4, and calculating the number of the searched face pixels and the number of the face pixels in the face imageThe ratio of the number of pixels in each rectangular block is recorded as(ii) a For example, the number of pixels characterizing a face in a rectangular block is 100, soThe first of genusA total of 200 pixels in each rectangular block, then;
C6, calculating the ratio and value of all the rectangular blocks remaining after the filtering of step C4 by the following formula (5):
In the formula (5), the first and second groups,to representIn the calculation ofThe weight occupied in (c);
Similarly, in calculatingIntroduction ofThe influence degree of different edge areas of the human face on the human face recognition result is considered to be different,for example, the cheekbone position and chin position in the edge area of the human face generally have a larger influence on the result of face recognition than the face position and forehead position.
d1, shooting the eye images of each person under the same crowd classification at a fixed distance and a fixed angle, and selecting an iris image from each eye image in a rectangular frame selection mode, wherein the iris image selected by the frame is shown in figure 4, the circle in figure 4 is an iris, and the external rectangle is a rectangular frame of the frame selected iris;
d2, halving each side of the width and the height of each iris image, and connecting the unequal points which are not opposite to each other to obtain a space quadrangle (indicated by the reference sign Q1 in figure 4);
d3, calculating the area of the space quadrangle and the rectangular frame of the iris image, and respectively recording as、;
D4, bisecting each side of the space quadrilateral (such as bisecting the side labeled "q 1" shown in FIG. 4), and then from each bisecting pointStarting from a line perpendicular to the edge, the line is connected to the iris boundary (e.g. the iris boundary indicated by "R1" in FIG. 4) of the iris image, and the point of connection is denoted as the vertex,、Respectively representing the first on a spatial quadrilateralBisector point of the sides, and from the bisector pointVertex connecting to iris boundary;
d5, from vertexTo the spatial quadrangleThe two ends of the edge are connected to obtain a triangle (such as the triangle denoted by the reference number "U1" in FIG. 4), which is marked asTriangle shapeWill be provided withThe outer iris region is separated into two arc iris regions, which are respectively marked as(e.g., as represented by the reference numeral area1 in FIG. 4),(indicated by reference character area2 in FIG. 4);
D7 in the form of triangleTwo waists of (A) areDividing each waist equally, connecting lines from the equal division points to the iris boundary of the arc iris region in a mode of being vertical to the waist to obtain connection vertexes, and connecting the connection vertexes to the triangleThe two end points of the waist are connected to obtain a triangle (such as the triangle marked by the reference number "U11" in FIG. 4), which is marked asSince there are two discrete arc iris areas per triangle, the iris shaping method is not limited to the above-mentioned methodq=1 or 2;
D9 in the form of triangleThe two waists areAnd (3) connecting the lines to obtain a triangle by the method of the steps D7-D8, calculating the area of the triangle until reaching the preset number of times of triangle construction, and then calculating the area of the iris image by the following formula (6):
In the formula (6), the first and second groups,an arbitrary bisecting block (such as the region indicated by the reference numeral "V1" and selected by the bold solid line frame in FIG. 4) for bisecting the iris image with the center point of the iris image as the origin of the XY axis coordinate system is shown as the second bisecting blockNext toDividing the edges equally;
representing pairs of arc-shaped iris areasOr to arc-shaped iris areasTo carry outNumber of edge equal divisions;
representing pairs of arc-shaped iris areasOr to arc-shaped iris areasTo proceed withThe edges being equally dividedThe number of edges;
representing the first obtained by equally dividing the iris imageDividing the blocks into equal parts;
Referring to fig. 1, the distance value and the value of each fingerprint image are calculatedOr accumulated value of quantityAnd calculating the ratio and value of each human face imageCalculating to obtain the sum ratio of each iris imageThen, the staff verification method based on the bloom filter provided by this embodiment proceeds to the following steps:
s2, drawing each fingerprint imageThe first image set corresponding to the falling sum value interval or each fingerprint image is marked intoThe second image set corresponding to the falling number interval; drawing each face image intoThe third image set corresponding to the falling sum value interval; draw each iris image intoThe fourth image set corresponding to the falling ratio interval is obtained, namely the application passesOrThe correspondence of the section in which the correspondence falls to the corresponding first image set or second image set, andthe correspondence of the section in which the correspondence falls and the corresponding third image set, andthe image set division of the fingerprint image, the face image and the iris image is realized by corresponding relation between the falling interval and the corresponding fourth image set, the provided brand new division mode considers the edge detail characteristics of each fingerprint image, face image and iris image, the core identification characteristics are not taken into the consideration range of the image set division basis, and the division speed of the data set is improved while the division accuracy is considered.
S3, based on the preset deviation degree range, mapping each of the first to fourth image sets as the pair of the bloom filter element to the bit arrayThe application point is up; the technical core of the bloom filter is as follows: the element is mapped to a point of a bit array, and it is known whether the element is in the data set by looking at this point to see if it is a "1" or a "0". The method utilizes the characteristic of the bloom filter to form corresponding relation between each bit array point position and the deviation degree range, and then calculates、、、The deviation degree of each of the first to fourth image sets can be matched with the image set corresponding to the corresponding point position quickly according to the deviation degree range in which the deviation degree falls, so that the acquisition speed of the image set serving as a personnel checking matching basis is increased, and the personnel checking speed is further increased;
in the staff checking method based on bloom filter provided by the present application, in steps S4-S8, a calculation method of staff deviation degree and a method for quickly matching out an image set corresponding to a corresponding point location based on a corresponding relationship between a deviation degree range and a point location array point location according to a deviation degree range in which the deviation degree falls are described in detail, and specifically referring to fig. 1, the method includes:
s4, calculating the corresponding fingerprint image in each first image setIs the average value of(ii) a And calculating the corresponding of each fingerprint image in each second image setAverage value of (2) is(ii) a And calculating the corresponding of each face image in each third image setIs the average value of(ii) a And calculating the corresponding relation of each iris image in each fourth image setIs the average value of;
S5, acquiring the fingerprint image, face image and iris image of the person to be checked, and then calculating the distance value and value of the fingerprint image of the person to be checkedOr accumulated value of quantityAnd calculating the ratio and value of the face image of the person to be checkedAnd calculating the sum ratio of the iris images of the person to be checked;
S6, calculated by the following formula (7)、、、Degree of deviation from each of the first to fourth image sets:
in the formula (7), the first and second groups,to representThe degree of deviation from each first image set;
S7, matching an image set corresponding to the point in the position array according to the corresponding relation between the deviation degree range in which the deviation degree falls and the point;
s8, according to a preset personnel checking method, the acquired fingerprint image, face image and iris image of the personnel to be checked are matched with each image set matched in the step S7 according to a preset matching rule, and a checking result is output, wherein the specific checking method comprises the following steps:
e1, judging whether the first image set or the second image set is matched as the checking basis of the person to be checked,
if yes, go to step E2;
if not, jumping to step E3;
e2, fingerprint comparison is carried out on the fingerprint image of the person to be checked and each fingerprint image in the first image set or the second image set,
if the comparison is successful, outputting a fingerprint comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E3;
e3, judging whether a third image set serving as the checking basis of the person to be checked is matched,
if yes, go to step E4;
if not, jumping to step E5;
e4, comparing the face image of the person to be checked with each face in the matched third image set,
if the comparison is successful, outputting a face comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E5;
e5, judging whether the fourth image set serving as the checking basis of the person to be checked is matched,
if yes, go to step E6;
if not, jumping to step E7;
e6, comparing the iris image of the person to be checked with each iris image in the fourth image set,
if the comparison is successful, outputting an iris comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E7;
e7, comparing the fingerprint image of the person to be checked with the fingerprints of all the first image sets or the second image sets, comparing the face image of the person to be checked with the faces of all the third image sets, comparing the iris image of the person to be checked with the irises of all the fourth image sets,
if any item is successfully compared, outputting a comparison result and terminating the personnel checking process, otherwise, outputting a comparison failure result.
It should be understood that the above-described embodiments are merely preferred embodiments of the invention and the technical principles applied thereto. It will be understood by those skilled in the art that various modifications, equivalents, changes, and the like can be made to the present invention. However, such variations are within the scope of the invention as long as they do not depart from the spirit of the invention. In addition, certain terminology used in the description and claims of the present application is not limiting, but is used for convenience only.
Claims (4)
1. A personnel checking method based on a bloom filter is characterized by comprising the following steps:
s1, calculating the distance value and the value of each fingerprint imageOr accumulated value of quantity(ii) a And calculating the ratio and value of each face image(ii) a And calculating the sum ratio of each iris image;
S2, drawing each fingerprint image intoThe first image set corresponding to the falling sum value interval or each fingerprint image is marked intoThe second image set corresponding to the falling number interval; drawing each face image intoThe third image set corresponding to the falling sum value interval; drawing each iris image intoThe fourth image set corresponding to the falling ratio interval;
s3, based on a preset deviation degree range, mapping each of the first to fourth image sets as elements of a bloom filter to corresponding points of a bit array;
s4, calculating the corresponding fingerprint image in each first image setIs the average value of(ii) a And calculating the corresponding fingerprint image in each second image setAverage value of (2) is(ii) a And computing each of said third imagesCorresponding to each face image in the setIs the average value of(ii) a And calculating the corresponding of each iris image in each fourth image setIs the average value of;
S5, acquiring the fingerprint image, the face image and the iris image of the person to be checked, and then calculating the distance value and the value of the fingerprint image of the person to be checkedOr accumulated value of quantityAnd calculating the ratio and value of the face image of the person to be checkedAnd calculating the sum ratio of the iris images of the person to be checked;
s7, matching an image set corresponding to the point location according to the corresponding relation between the deviation degree range in which the deviation degree falls and the point location in the point location array;
s8, according to a preset personnel checking method, carrying out personnel checking matching on the acquired fingerprint image, face image and iris image of the personnel to be checked and each image set matched in the step S7 according to a preset matching rule, and outputting a checking result;
in step S1, distance value and value of each fingerprint imageThe method is calculated by the following method steps:
a1, equally dividing each side into a plurality of sections in an equally-spaced mode for the width and the height of each fingerprint image classified by the same crowd framed in a rectangular frame selection mode;
a2, starting from each bisector, connecting lines to the opposite bisector of the opposite side in a mode of being perpendicular to the side where the starting point is located, so as to disperse the fingerprint image into a plurality of rectangular sub-blocks, taking the rectangular sub-block at the left top corner of the image as a starting sub-block of a standard sequence, and marking each rectangular sub-block according to a convolution sequence in a mode of circling the standard sequence anticlockwise inwards;
a3, filtering the rectangular subblocks which do not carry fingerprint information and are fully loaded with fingerprint information in the fingerprint image, wherein the fully loaded fingerprint information indicates that the pixel representing the fingerprint information touches each side of the rectangular subblock to which the pixel belongs;
a4, filtering each of the remaining rectangular sub-blocks in the step A3, and searching boundary pixel points of fingerprint information in each rectangular sub-block;
a5, calculating the rectangle in which each boundary pixel point is locatedThe distances of the left vertexes of the sub-blocks are summed according to the following formula (1) to obtain the distance value corresponding to each rectangular sub-block:
In the formula (1), the first and second groups,representing the second in the fingerprint imageA distance value of each of the rectangular sub-blocks;
is shown asThe first of the rectangular sub-blocksThe distance between each boundary pixel point and the left vertex of the rectangular sub-block;
a6, calculating the distance value and the value of all the rectangular sub-blocks remaining after the filtering of step A3 by the following formula (2):
In the formula (2), the first and second groups,to representIn the calculation ofThe weight occupied by the hour;
b1 calculated for each of the fingerprint image and the standard fingerprint imageAnd the two rectangular sub-blocks having the same rank number calculate the difference in distance value by the following formula (3):
In the formula (3), the first and second groups,indicating participationA calculated number one of the fingerprint imagesA distance value of each of the rectangular sub-blocks;
indicating participationThe calculated second one of the standard fingerprint image and the fingerprint imageThe first sub-blocks of the rectangle have the same rank numberA distance value of each of the rectangular sub-blocks;
b2, pairThe rectangular subblocks in the fingerprint image which are smaller than the difference value threshold are listed as quantity accumulation objects, and each fingerprint image conforms to the requirementsThe quantity accumulation of the rectangular blocks under the quantity accumulation condition is carried out to obtain the quantity accumulation value associated with each fingerprint image;
c1, shooting the face of each person under the same crowd classification at a fixed distance and a fixed angle to obtain the face image of each person with the same size;
c2, equally dividing each edge into a plurality of segments in an equally-spaced mode for the width and the height of each face image;
c3, starting from each bisector, connecting lines to the opposite bisectors of the opposite sides in a mode of being perpendicular to the side where the starting point is located, so as to disperse the face image into a plurality of rectangular blocks, using the rectangular block at the left top corner of the image as a starting block for marking, and marking each rectangular block according to a circling sequence in a mode of circling the rectangular blocks anticlockwise inwards to mark;
c4, filtering out the rectangular blocks which do not carry face information and are fully loaded with the face information in the face image, wherein the fully loaded face information means that pixels representing the face information touch each edge of the rectangular block to which the pixels belong;
c5, searching the face pixels in each rectangular block which is filtered by the step C4 and calculating the number of the searched face pixels and the number of the face pixels in the face imageThe ratio of the number of the pixel points in each rectangular block is recorded as;
C6, calculating the ratio and value of all the rectangular blocks remaining after the filtering of step C4 by the following formula (5):
In the formula (5), the first and second groups of the chemical reaction materials are selected from the group consisting of,to representIn the calculation ofThe weight occupied by (c);
d1, shooting eye images at a fixed distance and a fixed angle for each person under the same crowd classification, and framing out an iris image from each eye image in a rectangular frame selection mode;
d2, halving each side of the width and the height of each iris image, and connecting the unequal points which are not opposite to each other to obtain a space quadrangle;
d3, calculating the area of the space quadrangle and the rectangular frame of the iris image, and respectively recording the area as、;
D4, halving each side of the space quadrilateral, and then dividing each side from each halving pointStarting from a line connecting to the iris boundary of the iris image in a mode of being vertical to the edge, and marking the connected point as a vertex,、Respectively represent the first on the space quadrangleBisector point of the sides, and from the bisector pointVertex connecting to iris boundary;
d5, from vertexTo the space quadrangleConnecting two end points of the sides to obtain a triangle, and marking asTriangle shapeWill be provided withThe outer iris region is separated into two arc iris regions, which are respectively marked as、;
D7 in the form of the triangleThe two waists areAnd the edges are used for equally dividing each waist, connecting lines from the equal division points to the iris boundary of the arc iris region in a mode of being vertical to the waist to obtain connection vertexes, and connecting the connection vertexes to the triangleTwo end points of the waist are connected to obtain a triangle, and the triangle is marked as,;
D9 in the form of the triangleThe two waists areAnd (3) connecting the lines to obtain triangles and calculating the areas of the triangles by the method described in the steps D7-D8 until reaching the preset number of times of triangle construction, and then calculating the area of the iris image by the following formula (6):
In the formula (6), the first and second groups,representing that the first division of any division block of the iris image is performed by taking the central point of the iris image as the origin of an XY axis coordinate systemNext toDividing the edges equally;
representing pairs of arc-shaped iris areasOr to arc-shaped iris areasTo carry outNumber of edge equal divisions;
representing pairs of arc-shaped iris areasOr to arc-shaped iris areasTo proceed withThe edges being equally dividedThe number of edges;
2. The bloom filter-based people inspection method of claim 1, wherein each rectangular sub-block in the standard fingerprint image is a block of the fingerprint imageThe value is calculated by the following equation (4):
in the formula (4), the first and second groups,representing a first of said first or second set of imagesOpening the sum in the fingerprint imageThe corresponding rectangular sub-blocks have the distance values of the rectangular sub-blocks with the same row sequence number;
3. The bloom filter-based people-check method according to claim 1, wherein the calculation is performed by the following formula (8) in step S6、、、Degree of deviation from each of the first to fourth image sets:
in the formula (8), the first and second groups,to representThe degree of deviation from each first image set;
4. The bloom filter-based people-check method according to claim 1, wherein in step S8, people-check matching is performed by a matching rule expressed by the following method steps:
e1, judging whether the first image set or the second image set is matched as the checking basis for the person to be checked,
if yes, go to step E2;
if not, jumping to step E3;
e2, fingerprint comparison is carried out on the fingerprint image of the person to be checked and each fingerprint image in the first image set or the second image set,
if the comparison is successful, outputting a fingerprint comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E3;
e3, judging whether the third image set serving as the checking basis for the person to be checked is matched or not,
if yes, go to step E4;
if not, jumping to step E5;
e4, comparing the face image of the person to be checked with each face in the third image set,
if the comparison is successful, outputting a face comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E5;
e5, judging whether the fourth image set serving as the checking basis for the person to be checked is matched or not,
if yes, go to step E6;
if not, jumping to step E7;
e6, comparing the iris image of the person to be checked with each iris image in the fourth image set,
if the comparison is successful, outputting an iris comparison result and terminating the personnel checking process;
if the comparison fails, jumping to step E7;
e7, performing fingerprint comparison on the fingerprint image of the person to be checked and all the first image sets or the second image sets, performing face comparison on the face image of the person to be checked and all the third image sets, performing iris comparison on the iris image of the person to be checked and all the fourth image sets,
if any comparison is successful, outputting a comparison result and terminating the personnel checking process, otherwise, outputting a comparison failure result.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210662860.6A CN114780769B (en) | 2022-06-13 | 2022-06-13 | Personnel checking method based on bloom filter |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210662860.6A CN114780769B (en) | 2022-06-13 | 2022-06-13 | Personnel checking method based on bloom filter |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114780769A CN114780769A (en) | 2022-07-22 |
CN114780769B true CN114780769B (en) | 2022-09-13 |
Family
ID=82420940
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210662860.6A Active CN114780769B (en) | 2022-06-13 | 2022-06-13 | Personnel checking method based on bloom filter |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114780769B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109376725A (en) * | 2018-12-21 | 2019-02-22 | 北京无线电计量测试研究所 | A kind of identification check method and apparatus based on iris recognition |
WO2021027540A1 (en) * | 2019-08-13 | 2021-02-18 | 深圳壹账通智能科技有限公司 | Lost person information matching method and apparatus, computer device, and storage medium |
CN112562151A (en) * | 2020-12-03 | 2021-03-26 | 浪潮云信息技术股份公司 | Access control system based on bloom filter |
CN113705455A (en) * | 2021-08-30 | 2021-11-26 | 平安银行股份有限公司 | Identity verification method and device, electronic equipment and readable storage medium |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102722696B (en) * | 2012-05-16 | 2014-04-16 | 西安电子科技大学 | Identity authentication method of identity card and holder based on multi-biological characteristics |
CN103632132B (en) * | 2012-12-11 | 2017-02-15 | 广西科技大学 | Face detection and recognition method based on skin color segmentation and template matching |
US9224042B2 (en) * | 2013-04-17 | 2015-12-29 | Honeywell International Inc. | Cross-sensor iris matching |
CN108768966B (en) * | 2018-05-14 | 2019-05-03 | 北京邮电大学 | Block platform chain and member node and node identities authentication method |
US20200036708A1 (en) * | 2018-06-15 | 2020-01-30 | Proxy, Inc. | Biometric credential improvement methods and apparatus |
CN108932775B (en) * | 2018-07-10 | 2020-08-07 | 蒋钱 | Fingerprint lock identity identification system |
CN111435558A (en) * | 2018-12-26 | 2020-07-21 | 杭州萤石软件有限公司 | Identity authentication method and device based on biological characteristic multi-mode image |
CN111339885B (en) * | 2020-02-19 | 2024-05-28 | 平安科技(深圳)有限公司 | User identity determining method and related device based on iris recognition |
CN112329519B (en) * | 2020-09-21 | 2024-01-02 | 中国人民武装警察部队工程大学 | Safe online fingerprint matching method |
-
2022
- 2022-06-13 CN CN202210662860.6A patent/CN114780769B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109376725A (en) * | 2018-12-21 | 2019-02-22 | 北京无线电计量测试研究所 | A kind of identification check method and apparatus based on iris recognition |
WO2021027540A1 (en) * | 2019-08-13 | 2021-02-18 | 深圳壹账通智能科技有限公司 | Lost person information matching method and apparatus, computer device, and storage medium |
CN112562151A (en) * | 2020-12-03 | 2021-03-26 | 浪潮云信息技术股份公司 | Access control system based on bloom filter |
CN113705455A (en) * | 2021-08-30 | 2021-11-26 | 平安银行股份有限公司 | Identity verification method and device, electronic equipment and readable storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN114780769A (en) | 2022-07-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104866829B (en) | A kind of across age face verification method based on feature learning | |
Chen et al. | 3D free-form object recognition in range images using local surface patches | |
US7512255B2 (en) | Multi-modal face recognition | |
WO2017059591A1 (en) | Finger vein identification method and device | |
CN109934195A (en) | A kind of anti-spoofing three-dimensional face identification method based on information fusion | |
CN106709450A (en) | Recognition method and system for fingerprint images | |
CN106339707B (en) | A kind of gauge pointer image-recognizing method based on symmetric characteristics | |
CN110688901A (en) | Face recognition method and device | |
CN107145727A (en) | The medical image processing devices and method of a kind of utilization convolutional neural networks | |
CN102254165B (en) | Hand back vein identification method based on fusion of structural coding features and texture coding features | |
CN107066961B (en) | Fingerprint method for registering and device | |
CN108764041A (en) | The face identification method of facial image is blocked for lower part | |
CN111460950B (en) | Cognitive distraction method based on head-eye evidence fusion in natural driving conversation behavior | |
CN111339930A (en) | Face recognition method combining mask attribute loss function | |
CN106355139B (en) | Face method for anti-counterfeit and device | |
CN107958443A (en) | A kind of fingerprint image joining method based on crestal line feature and TPS deformation models | |
CN111126240A (en) | Three-channel feature fusion face recognition method | |
CN106934377A (en) | A kind of improved face detection system | |
CN104298995A (en) | Three-dimensional face identification device and method based on three-dimensional point cloud | |
CN110991258A (en) | Face fusion feature extraction method and system | |
CN110287847A (en) | Vehicle grading search method based on Alexnet-CLbpSurf multiple features fusion | |
Lin et al. | Local feature tensor based deep learning for 3d face recognition | |
CN114780769B (en) | Personnel checking method based on bloom filter | |
CN106980845B (en) | Face key point positioning method based on structured modeling | |
CN109815772A (en) | Fingerprint enhancement, recognition methods, device and Fingerprint enhancement identifying system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |