CN113963387B - Finger multi-mode feature extraction and fusion method based on optimal coding bits - Google Patents
Finger multi-mode feature extraction and fusion method based on optimal coding bits Download PDFInfo
- Publication number
- CN113963387B CN113963387B CN202111186478.4A CN202111186478A CN113963387B CN 113963387 B CN113963387 B CN 113963387B CN 202111186478 A CN202111186478 A CN 202111186478A CN 113963387 B CN113963387 B CN 113963387B
- Authority
- CN
- China
- Prior art keywords
- finger
- tri
- image
- modal
- mode
- 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
- 238000007500 overflow downdraw method Methods 0.000 title claims abstract description 30
- 238000000605 extraction Methods 0.000 title description 4
- 238000000034 method Methods 0.000 claims abstract description 82
- 230000002708 enhancing effect Effects 0.000 claims abstract description 11
- 230000004927 fusion Effects 0.000 claims description 37
- 230000005284 excitation Effects 0.000 claims description 20
- 230000004044 response Effects 0.000 claims description 14
- 238000001914 filtration Methods 0.000 claims description 7
- 238000012163 sequencing technique Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 5
- 210000004204 blood vessel Anatomy 0.000 abstract description 3
- 238000003384 imaging method Methods 0.000 abstract description 3
- 210000003462 vein Anatomy 0.000 description 32
- 238000004364 calculation method Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 7
- 238000005286 illumination Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 230000001627 detrimental effect Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000005469 granulation Methods 0.000 description 2
- 230000003179 granulation Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000002792 vascular Effects 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 1
- 238000007499 fusion processing Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000000087 stabilizing effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Collating Specific Patterns (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
A finger multi-mode image coding and fusion method based on optimal coding bits. The method comprises the steps of enhancing an original finger three-mode image to obtain a finger three-mode enhanced image; binary coding is carried out on the optimal enhancement direction of the finger tri-modal enhancement image by using a direction coding method so as to extract the effective texture characteristics of the finger tri-modal and obtain the finger tri-modal characteristic code; and fusing the three-mode feature codes of the fingers by using a feature code fusion method to obtain a final fused image. The invention has the following effects: effectively highlights the finger blood vessel imaging area and realizes the stable enhancement of the degenerated finger image. The problems of redundant information and redundant feature encoding bits that may occur are solved in order to extract features. The three-mode information of the finger can be fully utilized, and the accuracy and the robustness of recognition are improved.
Description
Technical Field
The invention belongs to the technical field of finger multi-mode image recognition, and particularly relates to a finger multi-mode image coding and fusion method based on optimal coding bits.
Background
With the advent of the information age and the rapid development of computer technology, information security has become a prerequisite for social security. Products which adopt single-mode characteristics for identity authentication are widely used at present. In application of single-mode biological feature recognition, the recognition performance is easy to be hindered by intra-class variation and deception attack, and is easy to be limited by actual acquisition conditions and environments, so that the high-performance identification requirement of people in daily life cannot be met. Multi-modal biometric is always superior to single-modal methods in terms of versatility, accuracy, and security. The multi-mode fusion can extract complementary and common characteristics among a plurality of modes, so that the characteristic information of the main body can be more comprehensively and carefully described, and the stability and the safety of the identity recognition system are improved. Therefore, multi-modal identification is an important direction of current research in compliance with the development of the trend of the age.
Among the numerous combinations of biometric features, finger-based multimodal authentication techniques are of particular interest due to their high flexibility and user acceptance. The fingerprints, veins and knuckle marks of the finger have very high specificity, and the recognition accuracy far exceeding that of a single feature can be achieved through the combination of the features. In addition, the features are concentrated at the positions of fingers and can be collected uniformly, the requirements on equipment are low, the application cost is low, and the method is easy to accept by users, so that the method is beneficial to realizing the productization of the technology rapidly.
Since the conventional finger multi-modal feature expression method is insensitive to illumination and does not have gray scale invariance and rotation invariance. However, the encoding-based feature representation method provides higher performance in terms of illumination invariance, feature description capability, and feature matching efficiency. Therefore, it is a key problem in research to explore a robust feature encoding method that is insensitive to illumination changes and has high recognition accuracy. In addition, the existing multi-modal fusion methods can generate a large storage space, and the feature expression and fusion methods do not fully consider the distinguishing features of fingers, so that satisfactory recognition performance cannot be generated. Therefore, the research of the feature expression method with robustness has great value for improving the recognition performance of the finger multi-modal feature fusion system.
Disclosure of Invention
In order to solve the problems, the invention aims to provide a finger multi-mode image coding and fusion method based on optimal coding bits.
In order to achieve the above object, the method for encoding and fusing multi-modal images of a finger based on optimal encoding bits provided by the present invention comprises the following steps performed in sequence:
1) Enhancing the original finger tri-modal image to obtain a finger tri-modal enhanced image;
2) Binary coding is carried out on the optimal enhancement direction of the finger tri-modal enhancement image by using a direction coding method so as to extract the effective texture characteristics of the finger tri-modal and obtain the finger tri-modal characteristic code;
3) And (3) fusing the finger three-mode feature codes obtained in the step (2) by using a feature code fusion method to obtain a final fused image.
In step 1), the method for enhancing the original finger tri-mode image to obtain the finger tri-mode enhanced image comprises the following steps: firstly, gabor filtering is carried out on an original finger tri-mode image by using a Gabor filter group with multiple scales, directivity and compatibility; then, based on weber's law, establishing directional weber differential excitation for the filtered image; finally, under the condition of multiple scales, an image with the strongest scale response and the strongest directional response, namely a finger three-mode enhanced image, is obtained.
In step 2), the method for binary encoding the optimal enhancement direction of the finger tri-modal enhanced image by using the direction encoding method to extract the effective texture features of the finger tri-modal, and the method for obtaining the finger tri-modal feature encoding is as follows:
First, 8 directions of a current pixel are represented using 8-bit binary coding; then defining a main direction for the current pixel and aligning the optimal enhancement direction with the main direction; then, only comparing the neighborhood value in the optimal enhancement direction of the finger tri-modal enhanced image with the current pixel value to obtain a binary code; if the neighborhood value is larger than the current pixel value, setting the binary code between two pixels to be 1, otherwise, setting the binary code to be 0; the binary code of the current pixel is always 1, and the values of other binary code bits are all set to 0, so that the finger tri-modal feature code is obtained.
In step 3), the method for fusing the finger tri-modal feature codes obtained in step 2) by using the feature code fusion method comprises the following steps: firstly, defining a main direction of the fusion of the three-mode feature codes of the finger; and then aligning the main direction in the direction coding method in the step 2) with the main direction of the finger tri-modal feature code fusion, and then fusing according to fusion sequencing by taking the main direction of the finger tri-modal feature code fusion as the center to obtain a final fusion image.
The finger multi-mode image coding and fusion method based on the optimal coding bit has the following beneficial effects:
1. Because the noise and redundant information exist at the edge of the captured original finger three-mode image, the imaging area quality is low and the like, the invention provides a finger vein blood vessel area stability enhancing method integrating Weber's law and Gabor filtering. The directional excitation capability of Wei Baju parts of descriptors is amplified through the multi-scale and multi-directional characteristics of Gabor filtering, and the interaction of the optimal response of Gabor filtering and the optimal excitation of Weber's law is realized, so that the finger vein imaging area is effectively highlighted, and the stable enhancement of the degraded finger vein image is realized. Experimental results show that the method is also applicable to fingerprints and knuckle prints.
2. A direction coding method based on Weber local descriptors is provided. In the multi-scale situation, the weber excitation response has the characteristics of strongest scale response and strongest directional response, so that the optimal enhancement direction is encoded, and the problems of redundant information (containing noise) and redundant feature encoding bits which can be generated are solved, so that features can be extracted. The optimal direction coding can adopt two coding methods of directivity and nondirectionality.
3. A feature code fusion method is provided. The method defines the main direction, aligns the direction of the finger trimodality with the main direction, and sequentially arranges the characteristic coding information of the finger trimodality, thereby fully utilizing the finger trimodality information and improving the accuracy and the robustness of the identification.
Drawings
fig. 1 is a 3*3 neighborhood schematic diagram of WLD.
Fig. 2 is a schematic diagram of three different neighbors of WLD.
Fig. 3 is an example of a portion of an original finger vein image and finger vein enhancement image. (a) an original finger vein image; (b) n=3; (c) n=5; (d) n=7.
Fig. 4 is an example of a finger tri-modal enhanced image. (a) finger vein enhancement images; (b) a knuckle print enhancement image; (c) fingerprint enhancing the image.
fig. 5 is a diagram illustrating a 5*5 neighborhood of the current pixel W.
xc
Fig. 6 is an example of encoding.
FIG. 7 is a schematic diagram of a finger tri-modal feature encoding fusion method; (a) directionality; (b) non-directional.
FIG. 8 is a fusion image of an original finger tri-modal image and a finger tri-modal feature code (a) an original finger tri-modal image and (b) a directional finger tri-modal feature code (d) fusion image; (c) The non-directional finger tri-modal feature encodes the fused image.
FIG. 9 is a comparison of recognition performance of different feature expression methods; (a) ROC curves for different feature expression methods; (b) 10 test accuracy curves for different feature expression methods.
Fig. 10 is a ROC curve for different sorting methods.
FIG. 11 is a comparison of recognition performance of different fusion methods; (a) ROC curves for different fusion methods; (b) 10 test accuracy curves for different fusion methods.
Detailed Description
The invention will now be described in detail with reference to the drawings and specific examples.
The finger multi-mode image coding and fusion method based on the optimal coding bit provided by the invention comprises the following steps in sequence:
1) Enhancing the original finger tri-modal image to obtain a finger tri-modal enhanced image;
The image content expression and enhancement method of the local descriptor is good at capturing local changes of the image content relative to the global descriptor, and is beneficial to describing local features of the target. In order to overcome the limitations of the existing local descriptors in illumination change, rotation transformation, scale singleness and the like, the method adopts local feature descriptors (multi-Gabor Weber Local Descriptor, MGWLD) with good generalization capability to enhance the original finger three-mode image. MGWLD organically integrates the neighborhood characteristics of a Wei Baju part descriptor (WLD) and the multi-scale multi-directional characteristics of a Gabor filter, and effectively considers the directional randomness of a finger vascular network and the local directional expression capability of the WLD.
WLD is characterized by simplicity, high efficiency and robustness to illumination variations. As shown in fig. 1, WLD may use 3*3 neighborhood filters to calculate the gray value of the current pixel, and the calculation formula of the original differential excitation intensity is shown in equation (1):
Where I represents the original excitation intensity, ΔI represents the excitation difference, Representing the spatial coordinates [ x, y ],/>Representing the current pixel/>P represents the number of neighbor pixels. If the original differential excitation intensity/>Indicating that the neighborhood pixel value is greater than the current pixel value; if the original differential excitation intensity/>Indicating that the brightness of the current pixel is low in this region. In the calculation process, in order to prevent the input value from being too large or too small, an arctan (·) function is utilized to map the output to a reasonable value range so as to prevent the ratio from being too large, thereby inhibiting the influence of partial noise.
The original differential excitation intensity will be poorRegarding as a scalar, only the gray differences in eight directions are summed, being essentially an isotropic laplace operator, resulting in insufficient applicability to gray scale variation information and sensitivity to image noise. For finger vein regions where the image content is more directional, the enhancement effect of the current pixel should be closely related to the directional excitation. Thus, the differential component simply represented as a scalar is detrimental to the directionality enhancement of the finger vein image.
In order to calculate the difference in directivity of the differential excitation, the calculation formula of redefining the intensity of the directivity differential excitation is shown as formula (2):
Wherein, theta k represents the kth direction of the WLD neighborhood, Representing the current pixel/>Differential excitation intensity in the kth direction. Since the finger vein image is seriously degraded, the blood vessel trend is not obvious, and the effect of the formula (2) on the aspect of describing the directional excitation difference is not ideal based on the original pixel information. Therefore, in order to highlight the directionality of the vascular network, it is necessary to perform directionality filtering on the original finger vein image.
The Gabor function has outstanding performance as a directional filter in terms of enhancement of the finger vein region. The combination of weber's law and Gabor filter is of great value for stabilizing and enhancing the finger vein region. In the kth direction θ k of the WLD neighborhood, the Gabor filtered finger vein image is:
wherein, Representing a directional multiscale Gabor filter bank, symbol/>Representing a 2D convolution,/>Representing the original finger vein image,/>Representing Gabor filtered finger vein images. Since the original Gabor wavelet is not compatible, in order to weaken the response deviation caused by the image illumination variation, the Gabor wavelet needs to be made compatible in directivity. A multiscale, directional, capacitive Gabor filter bank is defined as:
Wherein m represents the scale change of the Gabor wavelet, Δφε [1,1.5] represents the half-amplitude bandwidth of octaves, A= (diag [1, vsin (pi/16) (2 ln 2) -0.5 ]) represents a diagonal matrix of 2 x 2, embodying the anisotropy of Gabor wavelet,/>Σ m (m=1, 2, or 3) corresponds to the scale of the gabor wavelet, θ k represents the kth direction of the gabor wavelet, which is the same as the kth direction of the WLD neighborhood expressed by formula (2), k=1, 2,3, …, K, k=2 (n-1), n (=3, 5, or 7), and represents the neighborhood size of the WLD. When n=3, the neighborhood of WLD expresses 4 directions, when n=5 expresses 8 directions, and when n=7 expresses 12 directions, as shown in fig. 2.
The calculation formula for obtaining the new WLD differential excitation intensity by combining the formula (3) and the formula (4) is shown as the formula (6):
Since the main excitation lobe of Gabor wavelet is locally perpendicular to finger vein ridge line, the filter response is strongest, at this time the new WLD differential excitation intensity And should perform optimally. Thus, at a certain scale m, the optimal function of the finger vein enhancement image is:
thus, by using equation (7), taking the scale m=1, a finger vein enhancement image with the optimal single scale direction can be obtained, as shown in fig. 3. As can be seen from fig. 3, MGWLD has a very obvious effect on enhancing the vein region of the finger, especially under the 7*7 neighborhood condition, the calculation of the directional excitation by fusing gabor filtering and weber in 12 directions effectively suppresses noise, and the effect on enhancing the main blood vessel of the finger is very prominent.
Because the diameter of the finger vein is random in diameter variation in a neighborhood, the Gabor wavelet has the characteristics of multiple scales, and the optimal filter response can be obtained, as shown in formulas (3) and (4). Thus, in the multi-scale case, considering that the weber stimulus response should have both the strongest scale response and the strongest directional response, the enhancement function of the finger vein image is:
Thus, using equation (8), a finger vein enhancement image can be obtained; similarly, fingerprint and knuckle print enhanced images, i.e., finger trimodal enhanced images, may be obtained, as shown in FIG. 4.
2) Binary coding is carried out on the optimal enhancement direction of the finger tri-modal enhancement image by using a direction coding method so as to extract the effective texture characteristics of the finger tri-modal and obtain the finger tri-modal characteristic code;
Taking the 5 x 5 neighborhood of WLD as an example, 8 directions are defined. The 8 directions of the current pixel are represented by using 8-bit binary coding to obtain efficient texture features. The main ideas of the direction coding method are: the binary code is obtained by comparing only the neighborhood value in the optimal enhancement direction with the size of the current pixel value. If the neighborhood value is greater than the current pixel value, the binary code between the two pixels is set to 1, otherwise, to 0. The direction encoding method can improve recognition performance and reduce time cost compared to other encoding methods. In addition, the above-described direction encoding method includes two types: (1) a directional coding method; (2) non-directional coding method. The specific process is as follows:
(1) Directional coding method
To distinguish the binary encoding of the optimal enhancement direction for the current pixel, a main direction is defined for the current pixel, any direction may be defined as the main direction. The present invention defines direction 5 as the main direction, as shown in fig. 5. The specific coding process is as follows:
when 1< theta k < 8,
Wherein,
Wherein, θ k (=1, 2,3 …) represents the optimal enhancement direction of the image,Representing the enhanced current pixel. As is clear from FIG. 5,/>And/>Representing two different pixels in the same direction in a 5 x 5 neighborhood of WLD, respectively. TQ and TP represent the difference between two adjacent pixel values of the optimal enhancement direction on both sides of the main direction and the current pixel value, respectively. /(I)And c c denotes the binary encoding of the current pixel. c kl and c kr represent two different binary encodings of the optimal enhancement direction θ k in the WLD neighborhood, respectively. The binary coding of the optimal enhancement direction θ k is always 1. When 1< θ k <8, its binary coding depends on the sizes of TP and TQ. Binary code c kl is 1 if TP >0, and similarly binary code c kr is 1 if TQ > 0. For 8-bit binary coding, θ k =1 and θ k =8 are special directions. Thus, if the optimal enhancement direction is 1, the binary coding of the current pixel needs to be shifted one bit to the high in the binary coding. If the optimal enhancement direction is 8, the binary coding of the current pixel needs to be shifted one bit toward the lower position in the binary coding. Then, the calculation is carried out according to the rule of 1< theta k < 8. The feature code F (θ k) of the current pixel in the optimal enhancement direction θ k is to sum the binary codes, and the calculation formula is shown in formula (12):
Fig. 6 is an example of encoding. Assuming that the optimal enhancement direction of the current pixel is 3, the feature code of the current pixel can be obtained by using the formulas (9) - (12). Similarly, for other direction encodings, the same is true of the calculation process.
(2) Non-directional coding method
The definition of the main direction described above also applies to the non-directional coding method. It is noted that the non-directional coding method also takes direction 5 as the main direction. The coding calculation process is as follows:
wherein,
Unlike the directional coding method, in the non-directional coding method, the binary coding of the main direction 5 is always 1. If TQ > 0, binary code c 6 is 1. If TP > 0, binary code c 4 is 1. In other cases the binary code is 0.
At this time, the calculation formula of the feature code F (θ k) of the current pixel is shown in formula (16):
F(θk)=c4×23+c5×24+c6×25。 (16)
The main difference between the two coding methods is that the optimal enhancement direction of the directional coding method always corresponds to 8-bit binary coding, while the optimal enhancement direction of the non-directional coding method always aligns with the main direction, thus reducing the imbalance problem of binary coding caused by the direction. Other coding rules of the non-directional coding method are the same as those of the directional coding method, except that the non-directional coding method.
3) And (3) fusing the finger three-mode feature codes obtained in the step (2) by using a feature code fusion method to obtain a final fused image.
The main idea of the step is as follows: (1) Since the 5×5 neighborhood of WLD has 8 directions, 8-bit binary encoding ([ c1, c2, c3, c4, c5, c6, c7, c8 ]) is also used to represent the fused information of the finger trimodality. (2) In the binary encoding process of step 2), the direction 5 is taken as the main direction. The step is to take the direction 5 as the main direction of the fusion of the three-mode feature codes of the finger. (3) In the process of feature coding fusion of finger veins, fingerprints and knuckle veins, six fusion sequences can be generated according to the rule of permutation and combination, namely: FV, FP, FKP; FV, FKP, FP; FP, FV, FKP; FP, FKP, FV; FKP, FV, FP; FKP, FP, FV. Taking the first fusion ordering (FV, FP, FKP) as an example, (c 4, c 6), (c 3, c 7), (c 2, c 8) represent the feature codes on the left and right sides of the finger vein, fingerprint, and knuckle, respectively, of the current pixel.
The specific fusion method comprises the following steps: (1) When the directional finger tri-modal feature codes are fused, as shown in fig. 7 (a), the first line and the second line are finger tri-modal codes according to the coding method of step 2). In the directional coding method, the direction 1 and the direction 8 are special directions, and one bit of movement to the left or the right is required. And finally, aligning the optimal enhancement direction of each mode with the main direction in the multi-mode fusion frame, and fusing the codes on the left side and the right side of the pixels before the multi-mode by taking the main direction as the center according to the fusion sequence to obtain a final fusion image.
(2) When the non-directional finger tri-modal feature codes are fused, as shown in fig. 7 (b), in the non-directional coding method, all the modes are coded centering on the main direction 5. Therefore, during fusion, the main direction of the codes is aligned with the main direction in the multi-mode fusion frame, and the codes at the left side and the right side of the current pixel are fused by taking the main direction as the center according to the fusion sequence, so that a final fusion image is obtained.
In fig. 8, the original finger tri-modal image, (b) the directional finger tri-modal feature encoded fusion image; (c) The non-directional finger tri-modal feature encodes the fused image. It is apparent that the non-directional encoding method results in a fused image with relatively sharp texture features. This is because for directional coding, some directions always correspond to high coded bits in the binary coding, which corresponds to the binary coding being always larger. Conversely, other directions always correspond to low-encoded bits in the binary code, which correspond to the binary code being smaller at all times. This phenomenon can produce unbalanced binary coding and asymmetric coded images. Furthermore, unclear image features would be detrimental to the identification of biological features. Therefore, when the three-mode feature codes of the fingers are fused, the invention preferably uses a non-directional coding method to binary code the optimal enhancement direction of each single-mode enhanced image, and then carries out multi-mode feature fusion on the coded result.
To fully demonstrate the feasibility and effectiveness of the method of the present invention, the inventors performed a series of experiments using a laboratory-created finger trimodal database for evaluating the identification performance of the method of the present invention. This database contains 585 categories, in which 10 finger vein, fingerprint and knuckle images were collected as raw finger trimodal images, namely 5850 Zhang Shouzhi vein images, 5850 fingerprint images and 5850 Zhang Zhijie fingerprint images, respectively. And normalizing all original finger three-mode images in the self-made database to 80 x 180 pixels, wherein the experimental environment is a PC, and the method is completed in a Matlab R2018a environment.
The recognition capability of the feature code fusion method provided by the invention applied to the finger single-mode features is compared with several common single-mode feature expression methods (LGS, SLGS, LBP, LLBP, compcode and SCW-LGS). FIG. 9 is an experimental result of different expression methods on a single-mode database (finger vein), and EER, AVE (with STD) and time taken for single feature extraction of different methods are shown in Table 1.
TABLE 1 identification results of different feature expression methods
According to the rule of permutation and combination, six fusion sequences can be generated in the fusion process of feature codes of finger veins, fingerprints and knuckle veins: (FV, FP, FKP), (FV, FKP, FP), (FP, FV, FKP), (FP, FKP, FV), (FKP, FV, FP), (FKP, FP, FV). To determine the most suitable fusion ordering method to fuse the three-mode feature codes of the fingers, the recognition performance of 6 methods is compared, and the ROC curve is shown in figure 10.
As can be seen from the ROC curve, the identification performance of the six different fusion ordering methods is the same, which indicates that the identification performance of the fusion of the three-mode feature codes is irrelevant to the fusion ordering method of the single-mode feature codes. The feature coding fusion method provided by the invention has robustness to the three-mode feature coding fusion of the finger. Therefore, any feature code ordering method can be used in the finger tri-modal feature code fusion experiment.
The proposed FEF-OBC code fusion algorithm is compared with the fusion method (circle granulation, triangle granulation, GOM, graph_code_fusion, MRRID,) proposed in recent years. FIG. 11 shows the results of different fusion methods. Table 2 is the EER, STD and feature extraction time for individual pictures for different fusion methods. It can be seen that the EER result of the FEF-OBC provided by the invention is the lowest in the six fusion methods, the average accuracy AVE is about 99.93% and is relatively close to that of the graph_code_fusion, but the STD provided by the invention is much lower than that of the graph_code_fusion, which indicates that the FEF-OBC coding fusion method provided by the invention is more stable. In addition, compared with other methods, the method has the highest recognition efficiency, and the time spent for single recognition is less. In conclusion, the FEF-OBC multi-mode coding fusion algorithm provided by the invention is more accurate and has highest recognition efficiency. Therefore, the reliability of fusion feature recognition of the finger tri-mode FV, FP and FKP can be improved remarkably by using the proposed algorithm.
TABLE 2 identification results of different fusion methods
Claims (1)
1. A finger multi-mode image coding and fusion method based on optimal coding bits is characterized in that: the method comprises the following steps performed in sequence:
1) Enhancing the original finger tri-modal image to obtain a finger tri-modal enhanced image;
2) Binary coding is carried out on the optimal enhancement direction of the finger tri-modal enhancement image by using a direction coding method so as to extract the effective texture characteristics of the finger tri-modal and obtain the finger tri-modal characteristic code;
3) Fusing the finger three-mode feature codes obtained in the step 2) by using a feature code fusion method to obtain a final fused image;
In step 1), the method for enhancing the original finger tri-mode image to obtain the finger tri-mode enhanced image comprises the following steps: firstly, gabor filtering is carried out on an original finger tri-mode image by using a Gabor filter group with multiple scales, directivity and compatibility; then, based on weber's law, establishing directional weber differential excitation for the filtered image; finally, under the condition of multiple scales, obtaining an image with the strongest scale response and the strongest directional response, namely a finger three-mode enhanced image;
In step 2), the method for binary encoding the optimal enhancement direction of the finger tri-modal enhanced image by using the direction encoding method to extract the effective texture features of the finger tri-modal, and the method for obtaining the finger tri-modal feature encoding is as follows:
First, 8 directions of a current pixel are represented using 8-bit binary coding; then defining a main direction for the current pixel and aligning the optimal enhancement direction with the main direction; then, only comparing the neighborhood value in the optimal enhancement direction of the finger tri-modal enhanced image with the current pixel value to obtain a binary code; if the neighborhood value is larger than the current pixel value, setting the binary code between two pixels to be 1, otherwise, setting the binary code to be 0; the binary code of the current pixel is constantly 1, and the values of other binary code bits are all set to 0, so that the finger three-mode characteristic code is obtained;
In step 3), the method for fusing the finger tri-modal feature codes obtained in step 2) by using the feature code fusion method comprises the following steps: firstly, defining a main direction of the fusion of the three-mode feature codes of the finger; and then aligning the main direction in the direction coding method in the step 2) with the main direction of the finger tri-modal feature code fusion, and then fusing according to fusion sequencing by taking the main direction of the finger tri-modal feature code fusion as the center to obtain a final fusion image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111186478.4A CN113963387B (en) | 2021-10-12 | 2021-10-12 | Finger multi-mode feature extraction and fusion method based on optimal coding bits |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111186478.4A CN113963387B (en) | 2021-10-12 | 2021-10-12 | Finger multi-mode feature extraction and fusion method based on optimal coding bits |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113963387A CN113963387A (en) | 2022-01-21 |
CN113963387B true CN113963387B (en) | 2024-04-19 |
Family
ID=79464025
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111186478.4A Active CN113963387B (en) | 2021-10-12 | 2021-10-12 | Finger multi-mode feature extraction and fusion method based on optimal coding bits |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113963387B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN118196099B (en) * | 2024-05-17 | 2024-07-23 | 南京阿吉必信息科技有限公司 | Manufacturing method of fiber mechanical reinforced LED display array |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4562467A (en) * | 1983-04-01 | 1985-12-31 | Itek Corporation | Data compression apparatus and method for encoding and decoding multi-line signals |
US6741748B1 (en) * | 1997-06-19 | 2004-05-25 | Samsung Electronics Co. | Scan interleaving method based on context values of neighboring pixels, improved encoding/decoding method of binary shape data and encoding/decoding apparatus suitable for the same |
CN1667639A (en) * | 2004-03-10 | 2005-09-14 | Lg电子有限公司 | Binary coding method of fingerprint image |
CN103310411A (en) * | 2012-09-25 | 2013-09-18 | 中兴通讯股份有限公司 | Image local reinforcement method and device |
CN109902585A (en) * | 2019-01-29 | 2019-06-18 | 中国民航大学 | A kind of three modality fusion recognition methods of finger based on graph model |
CN109993142A (en) * | 2019-04-10 | 2019-07-09 | 黑龙江大学 | Two dimensional code identity identifying method based on finger portion multi-modal biological characteristic |
CN111353472A (en) * | 2020-03-30 | 2020-06-30 | 蚌埠学院 | Finger vein identification method based on LBP and Gabor direction Weber local descriptor |
CN111507206A (en) * | 2020-03-29 | 2020-08-07 | 杭州电子科技大学 | Finger vein identification method based on multi-scale local feature fusion |
US11068693B1 (en) * | 2020-12-29 | 2021-07-20 | King Abdulaziz University | Liveness detection in fingerprint-based biometric systems |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1952344B1 (en) * | 2005-11-23 | 2011-06-08 | Cedara Software Corp. | Method and system for enhancing digital images |
EP1868369B1 (en) * | 2006-06-12 | 2013-08-21 | STMicroelectronics (Research & Development) Limited | Digital filter |
-
2021
- 2021-10-12 CN CN202111186478.4A patent/CN113963387B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4562467A (en) * | 1983-04-01 | 1985-12-31 | Itek Corporation | Data compression apparatus and method for encoding and decoding multi-line signals |
US6741748B1 (en) * | 1997-06-19 | 2004-05-25 | Samsung Electronics Co. | Scan interleaving method based on context values of neighboring pixels, improved encoding/decoding method of binary shape data and encoding/decoding apparatus suitable for the same |
CN1667639A (en) * | 2004-03-10 | 2005-09-14 | Lg电子有限公司 | Binary coding method of fingerprint image |
CN103310411A (en) * | 2012-09-25 | 2013-09-18 | 中兴通讯股份有限公司 | Image local reinforcement method and device |
CN109902585A (en) * | 2019-01-29 | 2019-06-18 | 中国民航大学 | A kind of three modality fusion recognition methods of finger based on graph model |
CN109993142A (en) * | 2019-04-10 | 2019-07-09 | 黑龙江大学 | Two dimensional code identity identifying method based on finger portion multi-modal biological characteristic |
CN111507206A (en) * | 2020-03-29 | 2020-08-07 | 杭州电子科技大学 | Finger vein identification method based on multi-scale local feature fusion |
CN111353472A (en) * | 2020-03-30 | 2020-06-30 | 蚌埠学院 | Finger vein identification method based on LBP and Gabor direction Weber local descriptor |
US11068693B1 (en) * | 2020-12-29 | 2021-07-20 | King Abdulaziz University | Liveness detection in fingerprint-based biometric systems |
Non-Patent Citations (3)
Title |
---|
"融合Weber定律与Gabor滤波的手指静脉图像增强方法";杨玉清等;《激光与光电子学进展》;20210731;第58卷(第4期);1410003-1至1410003-7 * |
基于双Gabor方向韦伯局部描述子的掌纹识别;王华彬等;电子与信息学报;20180415(第04期);936-943 * |
杨玉清.手指多模态图像方向性增强和编码方法研究.2022,摘要、目录、绪论部分. * |
Also Published As
Publication number | Publication date |
---|---|
CN113963387A (en) | 2022-01-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Xin et al. | Multimodal feature-level fusion for biometrics identification system on IoMT platform | |
Zhao et al. | High resolution partial fingerprint alignment using pore–valley descriptors | |
CN105095880B (en) | A kind of multi-modal Feature fusion of finger based on LGBP coding | |
CN108509927B (en) | Finger vein image identification method based on local symmetrical graph structure | |
CN105138974B (en) | A kind of multi-modal Feature fusion of finger based on Gabor coding | |
CN102521575A (en) | Iris identification method based on multidirectional Gabor and Adaboost | |
CN107408195B (en) | Iris identification method and device | |
CN107092874A (en) | Personal identification method, apparatus and system based on electrocardio and fingerprint fusion feature | |
He et al. | Iris feature extraction method based on LBP and chunked encoding | |
Fei et al. | Enhanced minutiae extraction for high-resolution palmprint recognition | |
CN113963387B (en) | Finger multi-mode feature extraction and fusion method based on optimal coding bits | |
Patel et al. | An improved approach in fingerprint recognition algorithm | |
Bahaa-Eldin | A medium resolution fingerprint matching system | |
Kumar et al. | Non-overlapped blockwise interpolated local binary pattern as periocular feature | |
El-Abed et al. | Quality assessment of image-based biometric information | |
Rana et al. | MSRD-CNN: Multi-scale residual deep CNN for general-purpose image manipulation detection | |
Hassin et al. | Machine Learning System for Human–Ear Recognition Using Scale Invariant Feature Transform | |
Kanjan et al. | A comparative study of fingerprint matching algorithms | |
Nachar et al. | Hybrid minutiae and edge corners feature points for increased fingerprint recognition performance | |
Al Tamimi et al. | Offline signature recognition system using oriented FAST and rotated BRIEF | |
Kumar et al. | FPR using machine learning with multi‐feature method | |
George et al. | Transform domain fingerprint identification based on DTCWT | |
Jaswal et al. | Deformable multi-scale scheme for biometric personal identification | |
DR et al. | Fingerprint verification based on fusion of minutiae and ridges using strength factors | |
Khammari | Weber pattern and Binarized Statistical Image Features encoded CNN for fingerprint liveness detection |
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 |