Fingerprint image enhancement method, electronic device and computer readable storage medium

Abstract

The application discloses a fingerprint image enhancement method, an electronic device and a computer readable storage medium. The method comprises the following steps: obtaining a to-be-processed fingerprint image, performing denoising processing on the to-be-processed fingerprint image to obtain a texture feature image, performing filtering processing on the texture feature image to obtain a filtering response image corresponding to the texture feature image, determining a quality score corresponding to the texture feature image, and fusing the texture feature image and the filtering response image based on the quality score to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image. In the foregoing manner, the application can improve the accuracy of fingerprint image enhancement.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image processing, in particular to a fingerprint image enhancement method, an electronic device and a computer readable storage medium. BACKGROUND

[0002] Fingerprint is the pattern formed by the skin of the human finger, and the fingerprint characteristics of each person are different, which has good distinguishability. Therefore, fingerprint recognition is increasingly valued. However, when collecting a fingerprint image, the dirt on the finger tip or the shaking during the collection process will reduce the value of the fingerprint image for fingerprint recognition. In order to improve the value of the fingerprint image for fingerprint recognition, fingerprint image enhancement has become an extremely important link in fingerprint recognition, which can restore some broken and missing ridge lines in the fingerprint, so as to make the fingerprint characteristics more obvious. Therefore, the precision of the fingerprint image enhancement greatly affects the quality of the fingerprint minutia feature extraction and the accuracy of the fingerprint recognition. Therefore, how to improve the precision of the fingerprint image enhancement has become a problem to be solved. SUMMARY

[0003] The technical problem solved by the present application is to provide a fingerprint image enhancement method, an electronic device and a computer readable storage medium, which can improve the precision of the fingerprint image enhancement.

[0004] To solve the above technical problem, the first aspect of the present application provides a fingerprint image enhancement method, which comprises: obtaining a to-be-processed fingerprint image, performing denoising processing on the to-be-processed fingerprint image to obtain a texture feature image; performing filtering processing on the texture feature image to obtain a filter response image corresponding to the texture feature image; determining a quality score corresponding to the texture feature image, and fusing the texture feature image and the filter response image based on the quality score to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image.

[0005] To solve the above technical problem, the second aspect of the present application provides an electronic device, which comprises a memory and a processor coupled with each other, wherein the memory stores program data, and the processor calls the program data to execute the method of the first aspect.

[0006] To solve the above technical problem, the third aspect of the present application provides a computer readable storage medium, which stores program data, and the program data is executed by a processor to implement the method of the first aspect.

[0007] The scheme is characterized in that after obtaining the to-be-processed fingerprint image, the to-be-processed fingerprint image is denoised in a first stage to obtain a texture feature image, so as to improve the accuracy of the texture feature image; after obtaining the texture feature image, the texture feature image is filtered in a second stage, based on the characteristic that the response value at a false texture is small and the response value at a real texture is large, to obtain a filtered response image, so that the texture feature image is further enhanced through two progressive stages, the quality score of the texture feature image is determined, and the texture feature image and the filtered response image are fused based on the quality score, so that the texture feature image and the corresponding filtered response image of different qualities can be fused, the texture feature image is adaptively enhanced based on the quality score, and the target fingerprint image corresponding to the to-be-processed fingerprint image is obtained, thereby improving the precision of the fingerprint image enhancement. BRIEF DESCRIPTION OF DRAWINGS

[0008] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor. Among them:

[0009] Figure 1 is a flowchart of an embodiment of the fingerprint image enhancement method of the present application;

[0010] Figure 2 is a flowchart of another embodiment of the fingerprint image enhancement method of the present application;

[0011] Figure 3 is Figure 2 is a flowchart of an embodiment corresponding to step S202 in

[0012] Figure 4 is a structural diagram of an embodiment of the electronic device of the present application;

[0013] Figure 5 is a structural diagram of an embodiment of the computer-readable storage medium of the present application. DETAILED DESCRIPTION

[0014] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0015] The terms "system" and "network" are often used interchangeably herein. The term "and / or", whether used in the context of "comprising", "containing", "including", "having" or any other variation thereof, is used to mean a non-exclusive inclusion, such that a process, method, article or apparatus that comprises, contains, includes or has an item or items does not include only those items unless specifically stated otherwise. The term "or" is used in the context of features that are described as "any of" or "any of at least one of" and means either one of the features, or any combination of the features. The term "about" means approximately or nearly as understood by those of ordinary skill in the art. The term "comprising", whether used in the context of "comprising", "containing", "including", "having" or any other variation thereof, is used to mean a non-exclusive inclusion, such that a process, method, article or apparatus that comprises, contains, includes or has an item or items does not include only those items but can also include other items not specifically listed or other structural equivalents thereof. Other designated terms are defined within the specification.

[0016] The fingerprint image enhancement method provided by the present application is used for processing a fingerprint image, and the corresponding execution subject is a processor capable of calling an image.

[0017] Referring to Figure 1 , Figure 1 is a flowchart of an embodiment of the fingerprint image enhancement method of the present application, which comprises the following steps:

[0018] S101: obtaining a to-be-processed fingerprint image, performing feature extraction and denoising on the to-be-processed fingerprint image, and obtaining a texture feature image.

[0019] Specifically, after obtaining the to-be-processed fingerprint image, the to-be-processed fingerprint image is denoised in the first stage to obtain the texture feature image, so as to improve the accuracy of the texture feature image.

[0020] In an application mode, the to-be-processed fingerprint image is obtained, a convolution operation is performed on the to-be-processed fingerprint image to obtain a texture feature, the texture feature is compared with a texture feature threshold, texture features higher than the texture feature threshold are filtered to obtain a texture correction feature, and an inverse convolution operation is performed on the texture correction feature to obtain the texture feature image. In this way, in the process of feature extraction, the features related to noise are found out and filtered, so as to realize the denoising process of the to-be-processed fingerprint image and obtain a more accurate texture feature image.

[0021] In another application mode, the to-be-processed fingerprint image is obtained, a wavelet transform is performed on the to-be-processed fingerprint image to obtain a texture transform feature, the texture transform feature is compared with a texture feature threshold, texture transform features higher than the texture feature threshold are filtered to obtain a texture filtering feature, and an inverse wavelet transform is performed on the texture filtering feature to obtain the texture feature image. In this way, the wavelet denoising is used to remove invalid noise in the to-be-processed fingerprint image while the texture information is well preserved, so as to improve the accuracy of the texture feature image.

[0022] S102: performing filtering processing on the texture feature image to obtain a filter response image corresponding to the texture feature image.

[0023] Specifically, the texture feature image is filtered in the second stage, based on the characteristic that the response value is smaller at the false texture and the response value is larger at the real texture, to obtain a filtered response image, so that the texture feature image is further enhanced through the two progressive stages to obtain the filtered response image.

[0024] In an application mode, the texture feature image is divided into multiple image regions, and based on the ridge line direction and frequency of each image region, a Gabor filter matched with the image region is constructed and the matched image region is filtered, so that the filter matched with the local ridge line direction and frequency can effectively remove the features with smaller response values and maintain the real ridge valley structure of the fingerprint, to obtain the filtered image region. All the filtered image regions are spliced to obtain the filtered response image corresponding to the texture feature image.

[0025] In another application mode, the texture feature image is divided into multiple image blocks according to a preset step length, wherein the preset step length is smaller than the length of the image block, so that the multiple image blocks obtained are partially overlapped between adjacent image blocks. The image blocks are filtered using a short-time Fourier filter to highlight the response values generated at the real texture and filter the smaller response values at the false texture, to obtain the filtered image blocks. All the filtered image blocks are spliced, and the maximum value of the response values at the same position in the overlapping regions of the image blocks is taken to obtain the filtered response image corresponding to the texture feature image.

[0026] S103: Determine the quality score corresponding to the texture feature image, and fuse the texture feature image and the filtered response image based on the quality score to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image.

[0027] Specifically, the quality score of the texture feature image is determined, and the texture feature image and the filtered response image are fused based on the quality score, so that the texture feature image with different qualities and the corresponding filtered response image can be fused to adaptively enhance the texture feature image based on the quality score, to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image, and improve the accuracy of the fingerprint image enhancement.

[0028] In an application mode, the texture feature image is scored based on a preset reference dimension to obtain the quality score corresponding to the texture feature image, wherein the preset reference dimension at least includes local features and global features.

[0029] Further, the texture feature image and the filtering response image are fused according to the fusion weight to obtain the target fingerprint image, so that the target fingerprint image can obtain more accurate identification results when applied to fingerprint identification.

[0030] In another application mode, the texture feature image is scored based on a preset reference dimension to obtain a quality score corresponding to the texture feature image, wherein the preset reference dimension at least includes average deviation, maximum deviation and relative deviation corresponding to the feature information.

[0031] Further, the texture feature image corresponding to the base weight is obtained, the texture image corresponding to the bonus weight is determined based on the quality score, so that the first weight corresponding to the texture feature image is obtained based on the base weight and the bonus weight, the second weight corresponding to the filtering response image is determined based on the first weight, and the texture feature image and the filtering response image are fused according to the first weight and the second weight to obtain the target fingerprint image, wherein the first weight and the second weight are negatively correlated, so that the proportion of the texture feature image is increased when the quality of the texture feature image is high, and the proportion of the filtering response image is increased when the quality of the texture feature image is low, so that the texture feature image is adaptively enhanced based on the quality score to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image.

[0032] The above scheme obtains the to-be-processed fingerprint image, denoises the to-be-processed fingerprint image in the first stage to obtain the texture feature image to improve the accuracy of the texture feature image, obtains the texture feature image, filters the texture feature image in the second stage, obtains the filtering response image after filtering based on the characteristics that the response value at the false texture is small and the response value at the real texture is large, and determines the quality score of the texture feature image, so that the texture feature image and the filtering response image can be fused for different quality texture feature images and corresponding filtering response images, so that the texture feature image is adaptively enhanced based on the quality score to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image, and the precision of the fingerprint image enhancement is improved.

[0033] Please refer to Figure 2 , Figure 2 is a flowchart of another embodiment of the fingerprint image enhancement method of the present application, which comprises:

[0034] S201: obtaining a to-be-processed fingerprint image, extracting features and denoising the to-be-processed fingerprint image to obtain a texture feature image.

[0035] Specifically, after obtaining the to-be-processed fingerprint image, the to-be-processed fingerprint image is denoised by using the wavelet filter to obtain a texture feature image, so as to improve the accuracy of the texture feature image, wherein the wavelet filter comprises a wavelet transform module and a wavelet inverse transform module.

[0036] In an application scenario, a to-be-processed fingerprint image is obtained, the to-be-processed fingerprint image is wavelet transformed by using the wavelet transform module of the wavelet filter to obtain a texture transform feature, the texture transform feature is low-pass filtered to obtain a texture filtered feature, and the texture filtered feature is wavelet inverse transformed by using the wavelet inverse transform module of the wavelet filter to obtain a texture feature image.

[0037] Specifically, after obtaining the to-be-processed fingerprint image, the to-be-processed fingerprint image is wavelet transformed by using the wavelet transform module to obtain a texture transform feature, the wavelet coefficients obtained by the wavelet transformation in the texture transform feature are processed by low-pass filtering, and the texture transform feature corresponding to the wavelet coefficients greater than the coefficient threshold is filtered to obtain a texture filtered feature, and the texture filtered feature is wavelet inverse transformed by using the wavelet inverse transform module to obtain a texture feature image. Thus, the wavelet denoising can be used to remove invalid noise in the to-be-processed fingerprint image while the texture information is well preserved, the accuracy of the texture feature image is improved, and the subsequent stage is facilitated.

[0038] It should be noted that the wavelet filter is obtained after being pre-trained based on a training fingerprint image, and the training fingerprint image comprises a preset noise. The wavelet filter is iteratively optimized based on the difference between the result output by the wavelet inverse transform module and a reference fingerprint image. The reference fingerprint image comprises the training fingerprint image from which the preset noise is removed.

[0039] Specifically, the wavelet filter is trained by using the training fingerprint image in a training process. The training fingerprint image is input into the wavelet transform module, the training fingerprint image is wavelet transformed to obtain a training texture transform feature, the training texture transform feature is low-pass filtered to obtain a training texture filtered feature, and the training texture filtered feature is wavelet inverse transformed by using the wavelet inverse transform module to obtain a training texture feature image. Therefore, the difference between the training texture feature image and the reference fingerprint image is constrained in the training process, the wavelet filter is iteratively cycled, the coefficient threshold is gradually optimized and dynamically adjusted in the training process, the trained wavelet filter is obtained, so that the trained wavelet filter can more effectively remove noise, restore the real fingerprint image, and achieve a more optimal denoising effect.

[0040] S202: performing filtering processing on the texture feature image to obtain a filter response image corresponding to the texture feature image.

[0041] Specifically, the texture feature image is filtered based on a short-time Fourier transform (STFT) in a frequency domain to obtain a filtered response image, so that the filtered response image is further enhanced in two progressive stages.

[0042] In an application mode, refer to Figure 3 , Figure 3 is Figure 2 The flowchart of an embodiment corresponding to step S202 in

[0043] S301: The texture feature image is divided into blocks to obtain a plurality of image blocks, wherein adjacent image blocks partially overlap.

[0044] Specifically, a preset size of a to-be-moved region is determined on any corner of the texture feature image, and the to-be-moved region is translated on the texture feature image at a preset step length, wherein the preset step length is smaller than the preset size, so as to obtain a plurality of image blocks, and adjacent image blocks partially overlap. There is an overlapping region between adjacent image blocks, so in the subsequent step, the overlapping region is enhanced in different image blocks when image enhancement is performed, thereby improving the accuracy of fingerprint image enhancement.

[0045] Further, the plurality of image blocks are subjected to mean-variance normalization processing, so that the feature values in the image blocks are normalized to a distribution with a mean value of 0 and a variance of 1, thereby reducing the influence of abnormal values in the image blocks.

[0046] Optionally, a Fourier transform of the image block is calculated, the image block subjected to square root filtering and high-frequency signal truncation after the Fourier transform, so as to update the image block to improve the quality of the image block.

[0047] S302: A short-time Fourier transform is performed on each image block to obtain a direction field image and a frequency field image corresponding to each image block.

[0048] Specifically, a short-time Fourier transform is performed on each image block to obtain an energy field, a direction field and a frequency field corresponding to each image block, so as to determine a direction field image and a frequency field image corresponding to each image block.

[0049] S303: A short-time filter is constructed based on the direction field image and the frequency field image corresponding to the image block, and the image is filtered by using the short-time filter to obtain a filtered image block.

[0050] Specifically, a short-time filter is constructed based on the direction field image and the frequency field image corresponding to the image block, and the image is filtered by using the short-time filter to obtain a filtered image block, so as to enhance the real texture, especially the fingerprint ridge line, in the image block.

[0051] In an application scenario, at least one smoothing processing is performed on the direction field image to obtain a modified direction field image, anisotropic diffusion is performed on the frequency field image to obtain a modified frequency field image; an angle filter is constructed based on the modified direction field image, and a radius filter is constructed based on the modified frequency field image; wherein the short-time filter includes the angle filter and the radius filter; the image block is filtered by using the angle filter and the radius filter to obtain a filtered image block.

[0052] Specifically, the direction field image is denoted as O(x, y), the frequency field image is denoted as F(x, y), mean value smoothing and Gaussian smoothing are performed on the direction field image O(x, y) to obtain a modified direction field image O'(x, y) after smoothing processing, and anisotropic diffusion is performed on the frequency field image F(x, y) to obtain a modified frequency field image F'(x, y), so as to improve the accuracy of the modified direction field image O'(x, y) and the modified frequency field image F'(x, y).

[0053] Optionally, the modified direction field image O ′ (x, y) is subjected to secondary Gaussian filtering for smoothing processing, so as to update the modified direction field image O ′ (x, y).

[0054] Further, the angle filter F ′ is constructed with the direction field point in the modified direction field image O A (x, y) as the core, the radius filter F R is constructed with the frequency field point in the modified frequency field image F'(x, y) as the core, the angle filter F A and the radius filter F R comprise a short-time filter, so as to filter the image block by using the angle filter F A and the radius filter F R , to obtain a filtered image block F' = F*F R *F A , wherein F is the image block, so as to enhance the image block in two dimensions of the direction field and the frequency field, and improve the accuracy of the filtered image block F'.

[0055] S304: Based on the filtered image blocks corresponding to all the image blocks, a filtered response image corresponding to the texture feature image is obtained.

[0056] Specifically, the filtered image block after filtering processing is used to obtain a filtered response image corresponding to the texture feature image, so as to perform two-stage processing on the texture feature image, and further enhance the filtered response image on the basis of the texture feature image.

[0057] In an application scenario, inverse Fourier transform is performed on all the filter image blocks corresponding to the image blocks to obtain reconstructed filter response image blocks; the reconstructed filter response image blocks include a plurality of filter response values; the reconstructed filter response image blocks are spliced according to the positions of the image blocks, in the overlapping areas of adjacent reconstructed filter response image blocks, the filter response value with a larger absolute value is taken as a target response value, and in the non-overlapping areas of adjacent reconstructed filter response image blocks, the filter response value is taken as the target response value, to obtain a filter response image corresponding to the texture feature image; the filter response image includes a plurality of target response values.

[0058] Specifically, inverse Fourier transform is performed on all the filter image blocks corresponding to the image blocks to obtain reconstructed filter response image blocks on the spatial domain, and each reconstructed filter response image block includes a plurality of filter response values.

[0059] Further, the image blocks correspond to positions on the texture feature image, the reconstructed filter response image blocks corresponding to the image blocks are spliced according to the positions of the image blocks, the non-overlapping areas of adjacent reconstructed filter response image blocks correspond to non-overlapping areas of adjacent image blocks, for the non-overlapping areas of adjacent reconstructed filter response image blocks, the filter response value is taken as the target response value, and for the overlapping areas of adjacent reconstructed filter response image blocks, the filter response value with a larger absolute value is taken as the target response value, to obtain a filter response image corresponding to the texture feature image, so as to maximize the target response value in the filter response image and improve the enhancement effect of the filter response image.

[0060] S203: determining a quality score corresponding to the texture feature image based on the feature information of a plurality of preset dimensions corresponding to the texture feature image.

[0061] Specifically, the quality of the texture feature image is scored based on the feature information of a plurality of preset dimensions, to obtain a quality score corresponding to the texture feature image, so as to improve the accuracy of the quality score.

[0062] In an application scenario, a feature vector corresponding to the texture feature image is constructed based on the feature information of a plurality of preset dimensions corresponding to the texture feature image; the feature vector is mapped by a classifier to obtain a quality score corresponding to the texture feature image.

[0063] Specifically, a feature vector corresponding to the texture feature image is constructed based on the feature information of a plurality of preset dimensions corresponding to the texture feature image, and the feature vector is mapped to a quality score by a classifier, to obtain a quality score in a preset score interval, so as to improve the accuracy of the quality score.

[0064] In a specific application scenario, the image quality is evaluated based on a second generation fingerprint image quality scoring standard (NIST Fingerprint Image Quality 2, NFIQ2), the local features and the global features of the texture feature image are selected as the feature information of the fingerprint image based on the prediction ability, the calculation time and the correlation of the texture feature image, the feature vectors are constructed by connecting the feature information, the feature vectors are mapped to a fingerprint quality score score through training of a random forest classifier, the score is in an interval [1, 100], and thus the quality score corresponding to the texture feature image is obtained.

[0065] S204: Based on the quality score corresponding to the texture feature image, a first weight corresponding to the texture feature image and a second weight corresponding to the filtered response image are determined, wherein the first weight is positively correlated with the quality score, the second weight is negatively correlated with the first weight, and the first weight is greater than the second weight.

[0066] Specifically, based on the quality score corresponding to the texture feature image, the first weight corresponding to the texture feature image is determined, wherein the first weight includes a basic weight and an added weight, the added weight is positively correlated with the quality score, and therefore the first weight is positively correlated with the quality score, and based on the first weight, the second weight corresponding to the filtered response image is determined, wherein the first weight and the second weight correspond to a weight sum, and the basic weight of the first weight is at least half of the weight sum.

[0067] In a specific application scenario, the interval of the quality score is [1, 100], the added weight is determined based on the quality score and is added to the basic weight to obtain the first weight corresponding to the texture feature image, the sum of the first weight and the second weight is 1, and the above process is represented by the following formula:

[0068]

[0069] wherein w(x) is the first weight, 0.5 is the basic weight, score is the quality score, and therefore the second weight is 1-w(x).

[0070] S205: Based on the first weight and the second weight, the texture feature image and the filtered response image are fused to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image.

[0071] Specifically, based on the first weight corresponding to the texture feature image and the second weight corresponding to the filtered response image, the texture feature image and the filtered response image are fused to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image. The above process is represented by the following formula:

[0072] f out (x)=wf ori (x)+(1-w)*c*f enh(x) (2)

[0073] wherein, f out (x) is a target fingerprint image, w is a first weight, f ori (x) is a texture feature image, (1-w) is a second weight, c is a normalized restoration coefficient, f enh (x) is a filter response image.

[0074] Further, the first weight is greater than the second weight, so that the target fingerprint image is obtained by using the filter response image to adaptively enhance the texture feature image based on the denoised texture feature image and the quality score of the texture feature image, so that the target fingerprint image can obtain more accurate recognition results when applied to fingerprint recognition.

[0075] In this embodiment, the wavelet filter is used to denoise the fingerprint image to be processed to obtain the texture feature image to improve the accuracy of the texture feature image, the short-time Fourier transform based on the frequency domain is used to filter the texture feature image to obtain the filter response image after filtering, so that the filter response image further enhanced by the texture feature image is obtained through two progressive stages, the quality of the texture feature image is scored based on the feature information of a plurality of preset dimensions, so that the quality score corresponding to the texture feature image is obtained, the accuracy of the quality score is improved, the first weight corresponding to the texture feature image and the second weight corresponding to the filter response image are determined based on the quality score, and the target fingerprint image is obtained by using the filter response image to adaptively enhance the texture feature image based on the denoised texture feature image and the quality score of the texture feature image, so that the target fingerprint image can obtain more accurate recognition results when applied to fingerprint recognition.

[0076] Please refer to Figure 4 , Figure 4 is a structural schematic diagram of an embodiment of an electronic device of the present application. The electronic device 40 comprises a memory 401 and a processor 402 coupled with each other. The memory 401 stores program data (not shown in the figure), and the processor 402 invokes the program data to implement the method in any of the above embodiments. For related content, please refer to the detailed description of the above method embodiments, which will not be repeated here.

[0077] Please refer to Figure 5 , Figure 5 is a structural schematic diagram of an embodiment of a computer readable storage medium of the present application. The computer readable storage medium 50 stores program data 500, which is executed by a processor to implement the method in any of the above embodiments. For related content, please refer to the detailed description of the above method embodiments, which will not be repeated here.

[0078] It should be noted that the units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, i.e. may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment scheme.

[0079] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present alone, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0080] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the technical scheme of the present application essentially or the part that contributes to the prior art or the whole or part of the technical scheme can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to execute all or part of the steps of the method of each embodiment of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0081] The above is only the embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation, or direct or indirect application in other related technical fields, is also included in the patent protection scope of the present application.

Claims

1. A method of enhancing a fingerprint image, characterized by, The method comprises: obtaining a to-be-processed fingerprint image, performing denoising processing on the to-be-processed fingerprint image to obtain a texture feature image; performing filtering processing on the texture feature image to obtain a filtering response image corresponding to the texture feature image; determining a quality score corresponding to the texture feature image, and fusing the texture feature image and the filtering response image based on the quality score to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image; specifically comprising: determining the quality score corresponding to the texture feature image based on feature information of multiple preset dimensions corresponding to the texture feature image; determining a first weight corresponding to the texture feature image and a second weight corresponding to the filtering response image based on the quality score corresponding to the texture feature image; wherein the first weight is positively correlated with the quality score, the second weight is negatively correlated with the first weight, and the first weight is greater than the second weight; fusing the texture feature image and the filtering response image based on the first weight and the second weight to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image.

2. The method of claim 1, wherein, The method comprises: constructing a feature vector corresponding to the texture feature image based on the feature information of the multiple preset dimensions corresponding to the texture feature image; mapping the feature vector using a classifier to obtain the quality score corresponding to the texture feature image.

3. The method of claim 1, wherein, The method comprises: obtaining a to-be-processed fingerprint image, performing denoising processing on the to-be-processed fingerprint image to obtain a texture feature image; performing filtering processing on the texture feature image to obtain a filtering response image corresponding to the texture feature image; 4. The method of claim 3, wherein, determining a quality score corresponding to the texture feature image, and fusing the texture feature image and the filtering response image based on the quality score to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image; specifically comprising: determining the quality score corresponding to the texture feature image based on feature information of multiple preset dimensions corresponding to the texture feature image; determining a first weight corresponding to the texture feature image and a second weight corresponding to the filtering response image based on the quality score corresponding to the texture feature image; wherein the first weight is positively correlated with the quality score, the second weight is negatively correlated with the first weight, and the first weight is greater than the second weight; fusing the texture feature image and the filtering response image based on the first weight and the second weight to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image.

5. The method of claim 1, wherein, The method comprises: constructing a feature vector corresponding to the texture feature image based on the feature information of the multiple preset dimensions corresponding to the texture feature image; mapping the feature vector using a classifier to obtain the quality score corresponding to the texture feature image. The method comprises: obtaining a to-be-processed fingerprint image, performing denoising processing on the to-be-processed fingerprint image to obtain a texture feature image; performing filtering processing on the texture feature image to obtain a filtering response image corresponding to the texture feature image; determining a quality score corresponding to the texture feature image, and fusing the texture feature image and the filtering response image based on the quality score to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image; specifically comprising: determining the quality score corresponding to the texture feature image based on feature information of multiple preset dimensions corresponding to the texture feature image; determining a first weight corresponding to the texture feature image and a second weight corresponding to the filtering response image based on the quality score corresponding to the texture feature image; wherein the first weight is positively correlated with the quality score, the second weight is negatively correlated with the first weight, and the first weight is greater than the second weight; fusing the texture feature image and the filtering response image based on the first weight and the second weight to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image. The method comprises: constructing a feature vector corresponding to the texture feature image based on the feature information of the multiple preset dimensions corresponding to the texture feature image; mapping the feature vector using a classifier to obtain the quality score corresponding to the texture feature image. The method comprises: obtaining a to-be-processed fingerprint image, performing denoising processing on the to-be-processed fingerprint image to obtain a texture feature image; performing filtering processing on the texture feature image to obtain a filtering response image corresponding to the texture feature image; determining a quality score corresponding to the texture feature image, and fusing the texture feature image and the filtering response image based on the quality score to obtain a target fingerprint image corresponding to the to-be-processed fingerprint image; specifically comprising: determining the quality score corresponding to the texture feature image based on feature information of multiple preset dimensions corresponding to the texture feature image; determining a first weight corresponding to the texture feature image and a second weight corresponding to the filtering response image based on the quality score corresponding to the texture feature image; wherein the first weight is positively correlated with the quality score, the second weight is negatively correlated with the first weight, and the first weight is greater than the second weight; fusing the texture feature image and the filtering response image based on the first weight and the second weight to obtain the target fingerprint image corresponding to the to-be-processed fingerprint image. The method comprises: constructing a feature vector corresponding to the texture feature image based on the feature information of the multiple preset dimensions corresponding to the texture feature image; mapping the feature vector using a classifier to obtain the quality score corresponding to the texture feature image.

6. The method of enhancing a fingerprint image according to claim 5, wherein, The short-time filter is constructed based on the direction field image and the frequency field image corresponding to the image block, and the image block is filtered by using the short-time filter to obtain a filtered image block, including: The direction field image is smoothed at least once to obtain a modified direction field image, and the frequency field image is anisotropically diffused to obtain a modified frequency field image; An angle filter is constructed based on the modified direction field image, and a radius filter is constructed based on the modified frequency field image; wherein the short-time filter includes the angle filter and the radius filter; The image block is filtered by using the angle filter and the radius filter to obtain the filtered image block.

7. The method of claim 5, wherein the step of enhancing the fingerprint image is performed by using a method of enhancing a fingerprint image according to any one of claims 1 to 6. The filtered response image corresponding to the texture feature image is obtained based on the filtered image blocks corresponding to all the image blocks, including: Inverse Fourier transform is performed on the filtered image blocks corresponding to all the image blocks to obtain reconstructed filtered response image blocks; wherein the reconstructed filtered response image blocks include a plurality of filtered response values; The reconstructed filtered response image blocks are spliced according to the positions of the image blocks, in the overlapping areas of adjacent reconstructed filtered response image blocks, the filtered response value with a larger absolute value is taken as a target response value, and in the non-overlapping areas of adjacent reconstructed filtered response image blocks, the filtered response value is taken as a target response value, to obtain the filtered response image corresponding to the texture feature image; wherein the filtered response image includes a plurality of target response values.

8. An electronic device, comprising: including: The memory and the processor are coupled to each other, wherein the memory stores program data, and the processor calls the program data to execute the method in any one of claims 1-7.

9. A computer readable storage medium having stored thereon program data, wherein, The program data is executed by the processor to implement the method in any one of claims 1-7.

Images