Four-finger fingerprint position recognition method, device and equipment

Abstract

The present application discloses a four-finger fingerprint position recognition method, which includes: obtaining a four-finger fingerprint image F; performing image preprocessing on the four-finger image to obtain a binary image; performing image segmentation on the binary image based on the region growing method to obtain a plurality of connected regions G1, G2,.., G n , calculating the minimum rectangular box corresponding to each connected region, and obtaining a set of minimum rectangular boxes of the plurality of connected regions R = {R1, R2,.., R n}; obtaining the connected regions of the four fingerprints from the plurality of connected regions based on vertical correction and prior knowledge; performing position sorting on the four fingerprints and determining whether the four-finger fingerprint belongs to the left hand or the right hand. The present application also provides a four-finger fingerprint position recognition device and equipment.

Description

Technical Field

[0001] The present application relates to the field of image processing technology, and more specifically, to a four-finger fingerprint position recognition method, device and equipment. Background Art

[0002] With the advancement of science and technology, fingerprint recognition technology has begun to slowly enter the computer world. The automatic fingerprint recognition system is a computer application system that uses computer technology to process, store, compare and assist in identification of fingerprint image information. The core algorithms of the system include automatic fingerprint classification, positioning, morphological feature extraction, detail feature extraction and fingerprint comparison.

[0003] Using a 4-finger fingerprint collector to collect 4 fingerprints at a time can effectively fix the finger positions of the collector, avoid various fingerprint deformations and other unqualified fingerprints, and greatly improve the accuracy and efficiency of collection.

[0004] The existing fingerprint recognition technology includes multi-finger fingerprint segmentation and extraction, and most of them do not determine the positions of multiple fingerprints. When the same person collects four fingerprints of the same finger, it is not determined which palm the four fingerprints belong to, nor is it determined which finger of the index finger, middle finger, ring finger and little finger each fingerprint belongs to. In this way, when performing fingerprint recognition and comparison, each fingerprint needs to be identified and compared with the fingerprints of the index finger, middle finger, ring finger and little finger in the fingerprint library, which easily increases the number of fingerprint comparisons, thereby affecting the efficiency of fingerprint recognition. Summary of the invention

[0005] In view of the prior art, the technical problem solved by the present application is to provide a four-finger fingerprint position recognition method, device and equipment that can effectively determine the left and right hand ownership of four-finger fingerprints and effectively identify the position of each fingerprint of the four-finger fingerprints.

[0006] In order to solve the above technical problems, the present application provides a four-finger fingerprint position recognition method, comprising:

[0007] Obtain four-finger fingerprint image F;

[0008] Performing image preprocessing on the four-finger image to obtain a binary image;

[0009] The binary image is segmented based on the region growing method to obtain multiple connected regions G 1 ,G 2 ,..,G n , calculate the minimum rectangular frame corresponding to each connected area, and obtain the minimum rectangular frame set R of multiple connected areas = {R 1 ,R 2 ,..,R n};

[0010] Acquire connected regions about four fingerprints from the plurality of connected regions based on vertical correction and prior knowledge;

[0011] Sort the positions of the four fingerprints and determine whether the four fingerprints belong to the left hand or the right hand;

[0012] Among them, the connected region set G = {G 1 ,G 2 ,..,G n}, n≥4, multiple connected regions in the connected region set G correspond one-to-one to multiple minimum rectangular frames in the minimum rectangular frame set R.

[0013] In one embodiment of the application, when the binary image is a binary image with a white foreground and a black background, the step of marking a plurality of connected regions in the binary image based on a region growing method includes:

[0014] Starting from the upper left corner of the binary image, the binary image is traversed row by row and column by column, and the first pixel point found during the traversal that is not marked as a region and has a pixel value of 255 is set as P 1 , with pixel P 1 is the center point O of the region growth;

[0015] Find all pixel points P in the binary image that belong to the 4-region pixels of the center point O and whose pixel values ​​are 255, and push all the pixel points P into a stack;

[0016] Take a pixel from the stack, update the pixel to the center point O of the region growth, find all the pixel points P in the binary image that belong to the four-domain pixels of the center point O and whose pixel value is 255, and push all the pixel points P into the stack; repeat this step until the stack is empty; then mark all the pixel points in and out of the stack as belonging to the same connected region;

[0017] Continue to traverse the binary image, and repeat the above steps until all the pixels of the binary image with a pixel value of 255 have completed the marking of the connected region attribution.

[0018] In one embodiment of the application, a connected region G is calculated i The steps of the minimum rectangular frame include:

[0019] The binary image is rotated with an angle of 0 as the starting angle, and the binary image is rotated multiple times with an interval of a degrees and in the same direction until the binary image is rotated 90 degrees to obtain the connected region G. i The rotation set G iR ;

[0020] Calculate the connected area G iThe coordinates of all pixels after each rotation;

[0021] Traverse the rotation set G iR , calculate the rotation set G iR The bounding rectangle of all connected regions in the , the area of ​​the bounding rectangle and the four vertices of the bounding rectangle;

[0022] Select the circumscribed rectangular box with the smallest area and record it as rectangular box B im , with the rectangular frame B im The corresponding connected area is rotated relative to the binary image by an angle that the rectangular frame B im The four vertices of are rotated back to the original binary image to obtain the connected region G i The minimum rectangular box R in the binary image i ; where 1≤i≤n.

[0023] In one embodiment of the application, the rotation set G is calculated iR The steps of determining the circumscribed rectangular frame and the area of ​​the circumscribed rectangular frame of each connected region in the example include:

[0024] Traverse the rotation set G iR For all pixels in each connected area, compare the horizontal coordinates of all pixels to obtain the minimum horizontal coordinate value x min and the maximum value of the horizontal axis x max , compare the coordinate tables of all pixels to obtain the left minimum value y of the ordinate min and the maximum value of the ordinate y max ;

[0025] Take the minimum value x of the horizontal axis min , the maximum value of the horizontal axis x max , the left minimum value of the vertical coordinate y min and the maximum value of the ordinate y max The circumscribed rectangular frame that constitutes the connected region;

[0026] Calculate the area of ​​the bounding rectangle as follows: max -y min) ×(x max -x min ).

[0027] In one embodiment of the application, the degree a is 1 degree, and the connected area G i The rotation set G iR There are 91 connected regions after rotation.

[0028] In one embodiment of the application, the step of obtaining the vertically corrected connected regions about four fingerprints from the multiple connected regions based on the vertical correction and prior knowledge includes:

[0029] Traverse the minimum rectangular frame set R = {R 1 ,R 2 ,..,R n}, calculate the lengths of the four sides of each minimum rectangular frame, and rotate the minimum rectangular frame to a vertical position based on the inclination angle of the longest side of the four sides of each minimum rectangular frame to obtain the rotated minimum rectangular frame set R′={R′ 1 ,R′ 2 ,..,R′ n}, R 1 ,R 2 ,..,R n The rotation angles are denoted as α 1 , α 2 ,…,α n ;

[0030] Find all the minimum rectangular boxes R' in the minimum rectangular box set R' 1 ,R′ 2 ,..,R′ n The four vertices of

[0031] Get the minimum value of the ordinates of the four vertices of each minimum rectangular frame in the minimum rectangular frame set R′, and sort all the rectangular frames in the minimum rectangular frame set R′ in ascending order according to the minimum ordinates of the four vertices of the rectangular frame to obtain a rectangular frame set

[0032] Set the comparison mark value k = 1, select R s The second smallest rectangle in Calculate the center C of the second minimum rectangular box 2 The coordinates of C 2 Whether the horizontal coordinate falls into the rectangular box The minimum and maximum values ​​of the horizontal coordinates of the four vertices are within the range of the horizontal coordinates; if so, delete the rectangle Otherwise, keep the rectangle

[0033] Update the comparison mark value k = k + 1, select the original rectangular frame set R s The k+1th rectangular box in Calculate the minimum rectangle Center C k+1 The coordinates of C k+1 Whether the horizontal coordinate falls into the original rectangular frame set R s The minimum value of the horizontal coordinates of the four vertices of any rectangular box from the first to the kth and not deleted is within the range of the maximum value of the horizontal coordinates; if so, delete the rectangular box If you do not keep the rectangular frame Repeat this step until k is greater than n-1;

[0034] Statistical rectangular frame set R s The number of rectangular boxes retained in the image. If the number of rectangular boxes retained is 4, then the four connected areas corresponding to the four rectangular boxes are the connected areas of the four fingerprints f 1 、f 2 、f 3 and f 4 .

[0035] In one embodiment of the application, the connected area of ​​the four fingerprints is denoted as f 1 、f 2 、f 3 and f 4 The steps of determining the left and right hands of the four-finger fingerprints and sorting the positions of the four fingerprints include:

[0036] For the connected region f of the four fingerprints 1 、f 2 、f 3 and f 4 The corresponding four minimum rectangular boxes, the vertical coordinates of the four vertices of the minimum rectangular box of each fingerprint connected area are compared and sorted to obtain the minimum vertical coordinate value, and the connected area f is recorded 1 、f 2 、f 3 and f 4 The minimum values ​​of the ordinates of the four vertices of the respective minimum rectangular boxes are A, B, C and D respectively;

[0037] Sort A, B, C and D by size, and mark the connected areas of the four fingerprints with symbols 4, 3, 2 and 1 in order from the largest to the smallest value of the size sorting results;

[0038] According to the connected area f 1 、f 2 、f 3 and f 4 Output the symbol marking results from left to right in the horizontal coordinate direction of the image, and the symbol marking results are sorted;

[0039] If sort = {4,2,1,3}, it means that the four-finger images collected are the four fingerprints of the left hand and are the little finger, ring finger, middle finger and index finger from left to right in the horizontal coordinate direction of the image;

[0040] If sort = {3, 1, 2, 4}, it means that the collected four-finger images are the four fingerprints of the right hand and are the index finger, middle finger, ring finger and little finger from left to right in the horizontal coordinate direction of the image.

[0041] In one embodiment of the application, the step of performing image preprocessing on the four-finger image to obtain a binary image includes:

[0042] Performing mean smoothing filtering on the four-finger images;

[0043] Perform threshold binarization processing on the result after mean smoothing filtering;

[0044] A closing operation is performed on the result of the threshold binarization processing to obtain the binarized image.

[0045] The present application also provides a four-finger fingerprint position recognition device, comprising:

[0046] An image acquisition unit, used to acquire four-finger fingerprint images F;

[0047] An image preprocessing unit, used for performing binarization processing on the four-finger fingerprint image F to obtain a binarized image;

[0048] A plurality of connected region acquisition units, used for performing image segmentation on the binary image based on a region growing method to obtain a plurality of connected regions, and for calculating a minimum rectangular frame corresponding to each connected region;

[0049] a four-fingerprint connected region acquisition unit, configured to acquire connected regions about four fingerprints from the plurality of connected regions based on vertical correction and prior knowledge; and

[0050] The four-fingerprint position determination unit is used to sort the positions of the four fingerprints and determine whether the four fingerprints belong to the left hand or the right hand.

[0051] The present application also provides a device, comprising at least one memory for storing program instructions and at least one processor for executing the program instructions; when the program instructions are executed by the at least one processor, the steps of the above-mentioned four-finger fingerprint position recognition method are implemented.

[0052] The beneficial effects of the four-finger fingerprint position recognition method provided by the present application are: pre-processing the four-finger fingerprint image F to obtain a binary image, segmenting the binary image based on the region growing method to obtain multiple connected regions, obtaining the connected regions of the four fingerprints from the multiple connected regions through vertical correction and prior knowledge, and finally determining whether the connected regions of the four fingerprints after vertical correction belong to the left hand or the right hand, and identifying the positions of the four fingerprints belonging to the index finger, middle finger, ring finger and little finger respectively. The four-finger fingerprint recognition method can effectively determine whether the four-finger fingerprint image acquisition is the fingerprint of the left hand or the right hand, and can effectively identify which finger of the index finger, middle finger, ring finger and little finger each fingerprint region in the four-finger fingerprint image F belongs to; this is beneficial to reduce the number of fingerprint comparisons in the subsequent four-finger fingerprint recognition, improve fingerprint recognition efficiency, save user time and thus improve user experience.

[0053] The beneficial effects of the four-finger fingerprint position recognition device provided by the present application are: it can effectively determine whether the four-finger fingerprint image collection is the fingerprint of the left hand or the right hand, and can effectively identify to which finger of the index finger, middle finger, ring finger and little finger each fingerprint area in the four-finger fingerprint image F belongs; this is beneficial to reducing the number of fingerprint comparisons during subsequent four-finger fingerprint recognition, improving fingerprint recognition efficiency, saving user time and thus improving user experience.

[0054] The beneficial effect of the device provided by the present application is that the processor executes the program instructions to determine whether the four-finger fingerprint image collection is a fingerprint of the left hand or the right hand, and to identify to which finger of the index finger, middle finger, ring finger and little finger each fingerprint area in the four-finger fingerprint image F belongs, thereby reducing the number of fingerprint comparisons in subsequent four-finger fingerprint recognition, improving fingerprint recognition efficiency, saving user time and thus improving user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0056] Figure 1 This is a flow chart of a four-finger fingerprint position recognition method according to Embodiment 1 of the present application;

[0057] Figure 2 This is a result diagram of the four-finger fingerprint image F obtained in Example 1 of the present application;

[0058] Figure 3 For Figure 2The result of threshold binarization processing;

[0059] Figure 4 For Figure 3 The result graph of the closed operation;

[0060] Figure 5 This is a flow chart of marking multiple connected regions in the binary image based on the region growing method according to the first embodiment of the present application;

[0061] Figure 6 For calculating any connected region G in the first embodiment of the present application i Flowchart of the smallest rectangular box;

[0062] Figure 7 This is a flowchart of obtaining vertically corrected connected regions about four fingerprints from the multiple connected regions based on vertical correction and prior knowledge according to Embodiment 1 of the present application;

[0063] Figure 8 This is a flow chart of left-right hand determination of four-finger fingerprints and position sorting of four fingerprints according to the first embodiment of the present application;

[0064] Fig. 9 This is a diagram showing the position determination results of the four fingerprints of the left hand in Example 1 of the present application;

[0065] Fig.10 This is a structural block diagram of a four-finger fingerprint position recognition device according to Embodiment 2 of the present application. DETAILED DESCRIPTION

[0066] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0067] It should be noted that when an element is referred to as being "fixed to" or "disposed on" another element, it can be directly on the other element or indirectly on the other element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or indirectly connected to the other element.

[0068] It should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present application.

[0069] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0070] The four-finger fingerprint position recognition method, four-finger fingerprint position recognition device and equipment provided in the embodiments of the present application are now described in conjunction with the accompanying drawings.

[0071] Embodiment 1

[0072] Reference Figure 1 The steps of the four-finger fingerprint position recognition method provided in the embodiment of the present application include:

[0073] Step S100: Acquire four-finger fingerprint images F.

[0074] Step S200: performing image preprocessing on the four-finger image to obtain a binary image.

[0075] Step S300: Segment the binary image based on the region growing method to obtain multiple connected regions G 1 ,G 2 ,..,G n , let the connected region set G = {G 1 ,G 2 ,..,G n},n≥4.

[0076] Step S400: Calculate the minimum rectangular frame corresponding to each connected area of ​​the connected area set G, and obtain the minimum rectangular frame set R of multiple connected areas = {R 1 ,R 2 ,..,R n},n≥4.

[0077] Step S500: Obtaining the connected regions f of the four fingerprints from the multiple connected regions based on vertical correction and prior knowledge 1 、f 2 、f 3 and f 4 ;

[0078] Step S600: sorting the positions of the four fingerprints and determining whether the four fingerprints belong to the left hand or the right hand;

[0079] In the above steps, the upper left corner of the binary image is recorded as the image coordinate origin. It can be understood that in this embodiment, the origin of the image coordinate system of the image involved is at the upper left corner of the image. The multiple connected regions in the connected region set G correspond to the multiple minimum rectangular frames in the minimum rectangular frame set R one by one.

[0080] The acquired four-finger fingerprint image F is obtained by collecting the four-finger fingerprint image. When collecting the four-finger fingerprint image, the index finger, middle finger, ring finger and little finger are placed on the fingerprint instrument panel at the same time, and the images are collected at the same time. Figure 2 The four-finger fingerprint image shown. Figure 2 As shown, the origin of the image coordinate system is the point Op in the upper left corner of the image, and the mutually perpendicular horizontal axis X and vertical axis Y of the image coordinate system are as shown in Figure 2 As shown; the coordinates mentioned in the embodiments of the present application are all obtained according to this image coordinate system.

[0081] Image acquisition is affected by many factors (optical system distortion, system noise, underexposure or overexposure, etc.) during the generation and transmission process, which deteriorates the image processing results. Therefore, suppressing or eliminating these noises to improve image quality is an important step in the image processing process.

[0082] In step S200, the step of performing image preprocessing on the four-finger image to obtain a binary image includes:

[0083] Step S201: performing mean smoothing filtering on the four-finger images.

[0084] Step S202: performing threshold binarization processing on the result after the mean smoothing filtering.

[0085] Step S203: performing a closing operation on the result of the threshold binarization processing to obtain the binarized image.

[0086] In step S201, mean smoothing filtering is also called linear filtering, and the main method used is the neighborhood averaging method.

[0087] In step S202, the binarized image is recorded as image bin. The binarization of the image is beneficial to the further processing of the image, making the image simpler, reducing the amount of data, and highlighting the outline of the target of interest. In this embodiment, the target area of ​​interest is the fingerprint texture area. During the threshold binarization process, a fixed threshold threshold is collected. The threshold binarization process is: traverse all pixels of the image. If F(x,y)>threshold, bin(x,y)=0. Otherwise bin(x,y)=255. Among them, F(x,y) represents the pixel value of any pixel point (x,y) in the four-finger fingerprint image F, and bin(x,y) is the result of the threshold binarization of the pixel point (x,y). Figure 3The figure shows the result image bin after the four-finger fingerprint is binarized.

[0088] In step S203, in mathematical morphology, the closing operation is defined as dilation followed by erosion, such as Figure 4 The following figure shows the effect of the closing operation of the four-finger fingerprint. The fingerprint image is perfectly closed.

[0089] It can be understood that after the image F is processed by threshold binarization, the foreground fingerprint texture area is a white area with a pixel value of 255, and the background is a black area with a pixel value of 0. It can be understood that a finger generally includes three knuckles, and the fingerprint of the finger is on the first knuckle. When collecting fingerprint images, not only the fingerprint area of ​​the first knuckle of the index finger, middle finger, ring finger and little finger will be collected, but there will be textures of other knuckle areas that are mistakenly collected. Therefore, it is necessary to accurately extract the fingerprint area of ​​the four fingers belonging to the index finger, middle finger, ring finger and little finger.

[0090] To extract the fingerprint area, we first need to perform image segmentation to obtain multiple connected regions G 1 ,G 2 ,..,G n .

[0091] Reference Figure 5 In step S300, when the binary image bin is a binary image with a white foreground and a black background, the step of marking a plurality of connected regions in the binary image based on the region growing method includes:

[0092] Step S301: Starting from the upper left corner of the binary image, the binary image is traversed row by row and column by column, and the first pixel point found during the traversal that is not marked as a region and has a pixel value of 255 is set as P 1 , with pixel P 1 is the center point O of the region growing.

[0093] Step S302: Find all pixel points P in the binary image that are pixels in the four regions of the center point O and whose pixel values ​​are 255, and push all the pixel points P into a stack.

[0094] Step S303: Take a pixel point from the stack, update the pixel point to the center point O of the region growth, find all the pixel points P in the binary image that are 4-region pixels belonging to the center point O and whose pixel values ​​are 255, and push all the pixel points P into the stack.

[0095] Step S304: Check whether the stack is empty; if so, execute step S305: mark all pixels in and out of the stack as belonging to the same connected area; otherwise, return to execute step S303.

[0096] Step S306: Check whether all pixels with a pixel value of 255 in the binary image have completed the marking of the connected region; if so, execute step S307: complete the image segmentation to obtain multiple connected regions G 1 ,G 2 ,..,G n ; If not, return to execute step S301.

[0097] In step S301, the pixel point P can be set 1 The pixel points need to meet the following conditions: when traversing the binary image, the first pixel point encountered that has not been marked as a region and has a pixel value of 255.

[0098] It can be understood that in the above steps, looping step S301 completes the marking of a connected region, that is, segmenting a connected region, and repeating looping steps S301 and S306 can complete the marking of multiple connected regions and segmentation of multiple connected regions.

[0099] Reference Figure 6 In step S400, calculate any connected region G i The steps of the minimum rectangular frame include:

[0100] Step S401: Rotate the binary image with an angle of 0 as the starting angle, and continue to rotate the binary image multiple times with an interval of a degrees and in the same direction until the binary image is rotated 90 degrees to obtain the connected region G i The rotation set G iR ;

[0101] Step S402: Calculate the connected area G i The coordinates of all pixels after each rotation;

[0102] Step S403: traverse the connected region G i The rotation set G iR , calculate the rotation set G iR The bounding rectangle of all connected regions in the , the area of ​​the bounding rectangle and the four vertices of the bounding rectangle;

[0103] Step S404: Select the circumscribed rectangular frame with the minimum area and record it as rectangular frame B im , with the rectangular frame B im The corresponding connected area is rotated relative to the binary image by an angle that the rectangular frame B im The four vertices of are rotated back to the original binary image to obtain the connected region G i The minimum rectangular box R in the binary image i ;

[0104] In the above steps, 1≤i≤n. i mentioned in the following content also falls within this range.

[0105] In one embodiment, the degree a is 1 degree, and the connected area G i The rotation set G iR There are 91 connected regions after rotation.

[0106] In step S403, the rotation set G is calculated iR The steps of determining the circumscribed rectangular frame and the area of ​​the circumscribed rectangular frame of each connected region in the image include:

[0107] Traverse the rotation set G iR For all pixels in each connected area, compare the horizontal coordinates of all pixels to obtain the minimum horizontal coordinate value x min and the maximum value of the horizontal axis x max , compare the coordinate tables of all pixels to obtain the left minimum value y of the ordinate min and the maximum value of the ordinate y max ;

[0108] Take the minimum value x of the horizontal axis min , the maximum value of the horizontal axis x max , the left minimum value of the vertical coordinate y min and the maximum value of the ordinate y max The circumscribed rectangular frame that constitutes the connected region;

[0109] Calculate the area of ​​the bounding rectangle as follows: max -y min )×(x max -x min ).

[0110] It can be understood that when collecting the four-finger fingerprint image F, if the four fingers are placed on the fingerprint device in a vertical direction, the vertical direction is understood to be parallel to the Y-axis direction of the image coordinate system and the fingertips point to the X direction. The vertical position mentioned in the following steps can be understood as the four fingers are placed in a vertical direction. When the four fingers are imaged in a vertical direction, the order of the fingertips from high to low is the middle finger, ring finger, index finger and little finger.

[0111] Reference Figure 7 In the step S500, the step of obtaining the vertically corrected connected regions about the four fingerprints from the multiple connected regions based on the vertical correction and prior knowledge includes:

[0112] Step S501: traverse the minimum rectangular frame set R = {R 1 ,R 2 ,..,R n}, calculate the lengths of the four sides of each minimum rectangular frame, and rotate the minimum rectangular frame to a vertical position based on the inclination angle of the longest side of the four sides of each minimum rectangular frame to obtain the rotated minimum rectangular frame set R′={R′ 1 ,R′ 2 ,..,R′ n}, R 1 ,R 2 ,..,R n The rotation angles are denoted as α 1 , α 2 ,…,α n .

[0113] Step S502: Obtain all minimum rectangular frames R' in the minimum rectangular frame set R' 1 ,R′ 2 ,..,R′ n The four vertices of .

[0114] Step S503: Obtain the minimum value of the ordinates of the four vertices of each minimum rectangular frame in the minimum rectangular frame set R′, and sort all the rectangular frames in the minimum rectangular frame set R′ in ascending order according to the minimum ordinate values ​​of the four vertices of the rectangular frame to obtain a rectangular frame set

[0115] Step S504: Set the comparison mark value k=1, select R s The second smallest rectangle in Calculate the center C of the second minimum rectangular box 2 The coordinates of .

[0116] Step S505: Determine C 2 Whether the horizontal coordinate falls into the rectangular box The minimum and maximum values ​​of the horizontal coordinates of the four vertices are within the range of the horizontal coordinates; if so, delete the rectangle Otherwise, keep the rectangle

[0117] Step S506: Update the comparison mark value k=k+1, and determine whether k is greater than n-1; if so, execute step S507; otherwise, execute step S508.

[0118] Step S507: Count the rectangular frame set R s The number of rectangular boxes retained in the image. If the number of rectangular boxes retained is 4, then the four connected areas corresponding to the four rectangular boxes are the connected areas of the four fingerprints f 1 、f 2 、f 3 and f 4 .

[0119] Step S508: Select the original rectangular frame set R s The k+1th rectangular box in Calculate the minimum rectangle Center C k+1 The coordinates of .

[0120] Step S509: Determine C k+1 Whether the horizontal coordinate falls into the original rectangular frame set R s The minimum value of the horizontal coordinates of the four vertices of any rectangular box from the first to the kth and not deleted is within the range of the maximum value of the horizontal coordinates; if so, delete the rectangular box If you do not keep the rectangular frame

[0121] In the above steps, R 1 ,R 2 ,..,R n With α 1 , α 2 ,…,α n One-to-one correspondence.

[0122] In step S501, the tilt angle is the angle between the longest side and the Y axis.

[0123] In step S502, since R can be obtained through step S400 i The coordinates of the four vertices, R i By rotating the angle α i Rotate to the vertical position to obtain R′ i , so we can i The coordinates of the four vertices and the rotation angle α i Find R′ i The coordinates of the four vertices.

[0124] In the above steps, when step S509 is completed, the process proceeds to step S506.

[0125] In the above steps, it can be understood that step S506, step S508 and step S509 need to be repeatedly executed in a loop until the value of k is greater than or equal to n, that is, the value of k is greater than n-1.

[0126] In the above steps, if The coordinates of the four vertices are changed to and Where 1≤j≤n; C k+1 The coordinates are represented by C k+1 (x, y). Understandably, The coordinates of the four vertices are and Judgment C2 Whether the horizontal coordinate of falls into the rectangular box The range from the minimum value of the horizontal coordinates of the four vertices to the maximum value of the horizontal coordinates can be understood as: C 2 (x) whether it falls within the range W, Among them, min() means finding the minimum value, and max() means finding the maximum value.

[0127] When collecting four-finger fingerprint images, not only the fingerprint of the first joint is collected, but also the textures of other joints may be collected. Step S500 can eliminate the interference of the textures of other joints and extract the connected area of ​​the fingerprint texture belonging to the first joint.

[0128] Under normal circumstances, the fingerprints of the four fingers in the image are lined up from left to right. However, sometimes, the palm is tilted on the fingerprint device, and the fingerprint image will appear tilted. Therefore, the positions of the little finger, ring finger, middle finger and index finger cannot be simply determined by left-right sorting. In particular, the second and third knuckles of the fingers may be photographed at the same time during shooting. The influence of the second and third knuckles must be excluded when determining the fingerprints. After the vertical correction of rotating to the vertical direction in step S500, the fingerprint connection areas of the four fingers are rotated to the vertical direction, which facilitates the subsequent preparation for determining whether the four fingerprints belong to the left and right and the accurate determination of the positions of the four fingerprints.

[0129] From step S500, it can be seen that the connected area f of the four fingerprints 1 、f 2 、f 3 and f 4 is a rectangle set R s obtained in, while R s The fingerprint connected area corresponding to each rectangular frame in is placed in the vertical direction. When imaging in the vertical direction, the height order of the fingertips of the index finger, middle finger, ring finger and little finger can be determined.

[0130] Reference Figure 8 In step S600, the steps of determining the left and right hands of the four fingerprints and sorting the positions of the four fingerprints include:

[0131] Step S601: For the connected regions f of the four fingerprints 1 、f 2 、f 3 and f 4 The corresponding four minimum rectangular boxes, the vertical coordinates of the four vertices of the minimum rectangular box of each fingerprint connected area are compared and sorted to obtain the minimum vertical coordinate value, and the connected area f is recorded 1 、f 2 、f 3 and f 4The minimum values ​​of the ordinates of the four vertices of their respective minimum rectangular boxes are A, B, C and D respectively.

[0132] Step S602: sort A, B, C and D in order, and mark the connected areas of the four fingerprints with symbols 4, 3, 2 and 1 in order from the maximum to the minimum value of the sorting results.

[0133] Step S603: According to the connected area f 1 、f 2 、f 3 and f 4 The symbol marking results are output from left to right in the horizontal coordinate direction of the image, and the symbol marking results are sorted.

[0134] Step S604: determine the left and right hands of the four-finger fingerprints and the positions of the four fingerprints according to the symbol marking result sort: if sort = {4,2,1,3}, it means that the collected four-finger image is the four fingerprints of the left hand and they are the little finger, ring finger, middle finger and index finger from left to right in the direction of the horizontal coordinate of the image; if sort = {3,1,2,4}, it means that the collected four-finger image is the four fingerprints of the right hand and they are the index finger, middle finger, ring finger and little finger from left to right in the direction of the horizontal coordinate of the image; otherwise, re-acquire the fingerprint image F.

[0135] The horizontal axis direction is Figure 2 The X-axis direction of the image coordinate system, that is, the symbols are input from left to right to mark the results.

[0136] like Fig. 9 As shown, Fig. 9 It is the position determination result of the four fingerprints of the left and right fingers.

[0137] In the four-finger fingerprint position recognition method, the four-finger fingerprint image F is preprocessed to obtain a binary image, the binary image is segmented based on the region growing method to obtain multiple connected regions, the connected regions of the four fingerprints are obtained from the multiple connected regions through vertical correction and prior knowledge, and finally the connected regions of the four fingerprints after vertical correction are judged to belong to the left hand or the right hand, and the four fingerprints are respectively identified to belong to the index finger, middle finger, ring finger and little finger. The four-finger fingerprint recognition method can effectively judge whether the four-finger fingerprint image is collected from the left hand or the right hand, and can effectively identify which finger of the index finger, middle finger, ring finger and little finger each fingerprint region in the four-finger fingerprint image F belongs to; this is conducive to reducing the number of fingerprint comparisons in the subsequent four-finger fingerprint recognition, improving fingerprint recognition efficiency, saving user time and thus improving user experience.

[0138] Example 2, refer to Fig.10The four-finger fingerprint position recognition device 100 provided in the second embodiment of the present application includes an image acquisition unit 10, an image preprocessing unit 20, a plurality of connected area acquisition units 30, a four-fingerprint connected area acquisition unit 40 and a four-fingerprint position determination unit 50. The implementation principle of the image acquisition unit 10 is as shown in step S100 of the first embodiment, the principle of the image preprocessing unit 20 is as shown in step S200 of the first embodiment, the implementation principle of the plurality of connected area acquisition units 30 is as shown in steps S300 and S400 of the first embodiment, the implementation principle of the four-fingerprint connected area acquisition unit 40 is as shown in step S500 of the first embodiment, and the implementation principle of the four-fingerprint position determination unit 50 is as shown in step S600 of the first embodiment, which will not be repeated here.

[0139] Specifically, the image acquisition unit 10 is used to acquire a four-finger fingerprint image F. The image preprocessing unit 20 is used to perform binarization processing on the four-finger fingerprint image F to obtain a binary image. The multiple connected region acquisition unit 30 is used to perform image segmentation on the binary image based on the region growing method to obtain multiple connected regions, and to calculate the minimum rectangular frame corresponding to each connected region. The four-fingerprint connected region acquisition unit 40 is used to obtain connected regions about four fingerprints from the multiple connected regions based on vertical correction and prior knowledge. The four-fingerprint position determination unit 50 is used to sort the positions of the four fingerprints and determine whether the four-finger fingerprints belong to the left hand or the right hand.

[0140] In the four-finger fingerprint position recognition device 100, it is possible to effectively determine whether the four-finger fingerprint image is collected from the left hand or the right hand, and to effectively identify to which finger of the index finger, middle finger, ring finger and little finger each fingerprint area in the four-finger fingerprint image F belongs; this helps to reduce the number of fingerprint comparisons during subsequent four-finger fingerprint recognition, improve fingerprint recognition efficiency, save user time and thus improve user experience.

[0141] Embodiment 3: The device provided in Embodiment 3 of the present application includes at least one memory for storing program instructions and at least one processor for executing the program instructions; when the program instructions are executed by the at least one processor, the steps of the above-mentioned four-finger fingerprint position recognition method are implemented.

[0142] In the device, the processor executes the program instructions to determine whether the four-finger fingerprint image is collected from the left hand or the right hand, and to identify which finger of the index finger, middle finger, ring finger and little finger each fingerprint area in the four-finger fingerprint image F belongs to. This helps to reduce the number of fingerprint comparisons during subsequent four-finger fingerprint recognition, improve fingerprint recognition efficiency, save user time and thus improve user experience.

[0143] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A four-finger fingerprint position recognition method, characterized in that: include: Obtain four-finger fingerprint image F; Performing image preprocessing on the four-finger image to obtain a binary image; The binary image is segmented based on the region growing method to obtain multiple connected regions G1, G2, .., G n , calculate the minimum rectangular box corresponding to each connected area, and obtain the minimum rectangular box set R = {R1, R2, .., R n }; Acquire connected regions about four fingerprints from the plurality of connected regions based on vertical correction and prior knowledge; Sort the positions of the four fingerprints and determine whether the four fingerprints belong to the left hand or the right hand; Among them, the connected region set G = {G1, G2, .., G n }, n≥4, multiple connected regions in the connected region set G correspond one-to-one to multiple minimum rectangular frames in the minimum rectangular frame set R; The step of obtaining vertically rectified connected regions about four fingerprints from the plurality of connected regions based on the vertical rectification and prior knowledge comprises: Traverse the minimum rectangular box set R = {R1, R2, .., R n }, calculate the lengths of the four sides of each minimum rectangular box, and rotate the minimum rectangular box to a vertical position based on the inclination angle of the longest side of the four sides of each minimum rectangular box to obtain the rotated minimum rectangular box set R′={R′1,R′2,..,R′ n }, R1, R2, .., R n The rotation angles are denoted as α1, α2,…,α n ; Find all the minimum rectangular boxes R′1, R′2, .., R′ in the minimum rectangular box set R′ n The four vertices of Get the minimum value of the ordinates of the four vertices of each minimum rectangular frame in the minimum rectangular frame set R′, and sort all the rectangular frames in the minimum rectangular frame set R′ in ascending order according to the minimum ordinates of the four vertices of the rectangular frame to obtain a rectangular frame set Set the comparison mark value k = 1, select R s The second smallest rectangle in Calculate the coordinates of the center C2 of the second smallest rectangular frame and determine whether the horizontal coordinate of C2 falls into the rectangular frame The minimum and maximum values ​​of the horizontal coordinates of the four vertices are within the range of the horizontal coordinates; if so, delete the rectangle Otherwise, keep the rectangle Update the comparison mark value k = k + 1, select the original rectangular frame set R s The k+1th rectangular box in Calculate the minimum rectangle Center C k+1 The coordinates of C k+1 Whether the horizontal coordinate falls into the original rectangular frame set R s The minimum value of the horizontal coordinates of the four vertices of any rectangular box from the first to the kth and not deleted is within the range of the maximum value of the horizontal coordinates; if so, delete the rectangular box If you do not keep the rectangular frame Repeat this step until k is greater than n-1; Statistical rectangular frame set R s The number of rectangular frames retained in the image. If the number of rectangular frames retained is 4, the four connected areas corresponding to the four rectangular frames are the connected areas f1, f2, f3 and f4 of the four fingerprints.

2. The four-finger fingerprint position recognition method according to claim 1, characterized in that: When the binary image is a binary image with a white foreground and a black background, the step of marking a plurality of connected regions in the binary image based on a region growing method comprises: Starting from the upper left corner of the binary image, the binary image is traversed row by row and column by column, and the first pixel point found during the traversal that is not marked as a region and has a pixel value of 255 is set as P1, and the pixel point P1 is used as the center point O of the region growth; Find all pixel points P in the binary image that belong to the four-region pixels of the center point O and whose pixel values ​​are 255, and push all the pixel points P into a stack; Take a pixel from the stack, update the pixel to the center point O of the region growth, find all the pixel points P in the binary image that belong to the four-domain pixels of the center point O and whose pixel value is 255, and push all the pixel points P into the stack; repeat this step until the stack is empty; then mark all the pixel points in and out of the stack as belonging to the same connected region; Continue to traverse the binary image, and repeat the above steps until all the pixels of the binary image with a pixel value of 255 have completed the marking of the connected region attribution.

3. The four-finger fingerprint position recognition method according to claim 1, characterized in that: Calculate a connected region G i The steps of the minimum rectangular frame include: The binary image is rotated with an angle of 0 as the starting angle, and the binary image is rotated multiple times with an interval of a degrees and in the same direction until the binary image is rotated 90 degrees to obtain the connected region G. i The rotation set G iR ; Calculate the connected area G i The coordinates of all pixels after each rotation; Traverse the rotation set G iR , calculate the rotation set G iR The bounding rectangle of all connected regions in the , the area of ​​the bounding rectangle and the four vertices of the bounding rectangle; Select the circumscribed rectangular box with the smallest area and record it as rectangular box B im , with the rectangular frame B im The corresponding connected area is rotated relative to the binary image by an angle that the rectangular frame B im The four vertices of are rotated back to the original binary image to obtain the connected region G i The minimum rectangular box R in the binary image i ; where 1≤i≤n.

4. The four-finger fingerprint position recognition method according to claim 3, characterized in that: Calculate the rotation set G iR The steps of determining the circumscribed rectangular frame and the area of ​​the circumscribed rectangular frame of each connected region in the example include: Traverse the rotation set G iR For all pixels in each connected area, compare the horizontal coordinates of all pixels to obtain the minimum horizontal coordinate value x min and the maximum value of the horizontal axis x max , compare the coordinates of all pixels to obtain the minimum ordinate value y min and the maximum value of the ordinate y max ; Take the minimum value x of the horizontal axis min , the maximum value of the horizontal axis x max , minimum value of y coordinate min and the maximum value of the ordinate y max The circumscribed rectangular frame that constitutes the connected area; Calculate the area of ​​the bounding rectangle as follows: max -y min )×(x max -x min ).

5. The four-finger fingerprint position recognition method according to claim 3, characterized in that: The degree a is 1 degree, and the connected area G i The rotation set G iR There are 91 connected regions after rotation.

6. The four-finger fingerprint position recognition method according to claim 1, characterized in that: The connected areas of the four fingerprints are recorded as f1, f2, f3 and f4. The steps of determining the left and right hands of the four fingerprints and sorting the positions of the four fingerprints include: For the four minimum rectangular boxes corresponding to the connected areas f1, f2, f3 and f4 of the four fingerprints, the ordinates of the four vertices of the minimum rectangular box of each fingerprint connected area are compared and sorted to obtain the minimum ordinate value. The minimum ordinate values ​​of the four vertices of the minimum rectangular box of each connected area f1, f2, f3 and f4 are recorded as A, B, C and D respectively; Sort A, B, C and D by size, and mark the connected areas of the four fingerprints with symbols 4, 3, 2 and 1 in order from the largest to the smallest value of the sorting results; Output the symbol marking results from left to right in the horizontal coordinate direction of the image according to the connected areas f1, f2, f3 and f4, and the symbol marking result is sort; If sort = {4,2,1,3}, it means that the four-finger images collected are the four fingerprints of the left hand and are the little finger, ring finger, middle finger and index finger from left to right in the horizontal coordinate direction of the image; If sort = {3, 1, 2, 4}, it means that the collected four-finger images are the four fingerprints of the right hand and are the index finger, middle finger, ring finger and little finger from left to right in the horizontal coordinate direction of the image.

7. The four-finger fingerprint position recognition method according to claim 1, characterized in that: The step of performing image preprocessing on the four-finger image to obtain a binary image includes: Performing mean smoothing filtering on the four-finger images; Perform threshold binarization processing on the result after mean smoothing filtering; A closing operation is performed on the result of the threshold binarization processing to obtain the binarized image.

8. A four-finger fingerprint position recognition device, used to implement the steps of the four-finger fingerprint position recognition method according to any one of claims 1 to 7, characterized in that: include: An image acquisition unit, used to acquire four-finger fingerprint images F; An image preprocessing unit, used for performing binarization processing on the four-finger fingerprint image F to obtain a binarized image; A plurality of connected region acquisition units, used for performing image segmentation on the binary image based on a region growing method to obtain a plurality of connected regions, and for calculating a minimum rectangular frame corresponding to each connected region; A four-fingerprint connected region acquisition unit, used for acquiring connected regions about four fingerprints from the multiple connected regions based on vertical correction and prior knowledge; as well as The four-fingerprint position determination unit is used to sort the positions of the four fingerprints and determine whether the four fingerprints belong to the left hand or the right hand.

9. A device, characterized in that: It comprises at least one memory for storing program instructions and at least one processor for executing the program instructions; when the program instructions are executed by the at least one processor, the steps of the four-finger fingerprint position recognition method as described in any one of claims 1 to 7 are implemented.

Images