CN111814535A

CN111814535A - Palm print image identification method, palm print image identification device, palm print image identification equipment and computer readable storage medium

Info

Publication number: CN111814535A
Application number: CN202010433723.6A
Authority: CN
Inventors: 刘翔; 刘莹
Original assignee: Ping An Technology Shenzhen Co Ltd
Current assignee: Ping An Technology Shenzhen Co Ltd
Priority date: 2020-05-20
Filing date: 2020-05-20
Publication date: 2020-10-23
Anticipated expiration: 2040-05-20
Also published as: CN111814535B; WO2021151359A1

Abstract

The invention relates to artificial intelligence, and provides a palm print image identification method, a palm print image identification device, palm print image identification equipment and a computer readable storage medium, wherein the method comprises the following steps: when a palm print image to be processed is obtained, reading a registered palm print image in a preset database; respectively carrying out block matching on each registered palm print image and the to-be-processed palm print image to generate a plurality of matching data sets, wherein one registered palm print image correspondingly generates one matching data set; and determining a target palm print image matched with the palm print image to be processed in each registered palm print image according to the plurality of matching data groups. According to the invention, each registered palm print image is respectively matched with the palm print image to be processed in a blocking way through an image processing technology, so that the matching with the whole palm print image as a reference is avoided, the data amount of reference processing in the matching process is reduced, and the palm print image identification efficiency is favorably improved.

Description

Palm print image identification method, palm print image identification device, palm print image identification equipment and computer readable storage medium

Technical Field

The present invention relates to the field of image processing technologies, and in particular, to a method, an apparatus, a device, and a computer-readable storage medium for identifying a palm print image.

Background

The development of science and technology has enabled the wide application of identity verification via biometric features. The palm print is a relatively stable biological characteristic, and can effectively identify the identity of a person, so that the palm print can be widely applied to various application scenes needing to identify the identity of the person, such as unmanned supermarkets, attendance checking in workplaces and the like. The traditional palm print recognition is generally realized by scanning a palm image by means of a fixing device, and belongs to palm contact type recognition; long-term contact easily causes equipment pollution, makes the palm image of scanning unclear, and then influences the effect of discernment.

With the development of artificial intelligence, a technology for identifying a palm print by identifying a palm print image shot by a digital camera or a camera appears, so that the problem that the identification effect is influenced due to the unclear scanned palm image in the traditional identification mode is solved. The palm print recognition based on palm print image recognition is mainly realized by extracting features in palm print images for matching. However, in this method, a large number of features are often extracted from the palm print image for the accuracy of recognition, and each feature is matched with the whole palm print image as a reference, so that the feature data required to be referred to for matching is numerous, and the efficiency of palm print image recognition is low.

Disclosure of Invention

The invention mainly aims to provide a palm print image identification method, a palm print image identification device, palm print image identification equipment and a computer readable storage medium, and aims to solve the technical problem of low palm print image identification efficiency in the prior art.

In order to achieve the above object, an embodiment of the present invention provides a method for identifying a palm print image, where the method for identifying a palm print image includes the following steps:

when a palm print image to be processed is obtained, reading a registered palm print image in a preset database;

respectively carrying out block matching on each registered palm print image and the to-be-processed palm print image to generate a plurality of matching data sets, wherein one registered palm print image correspondingly generates one matching data set;

and determining a target palm print image matched with the palm print image to be processed in each registered palm print image according to the plurality of matching data groups.

Preferably, each of the registered palm print images exists in a preset database in a form of being divided into a preset number of block-by-block registered images;

the step of respectively performing block matching on each registered palm print image and the palm print image to be processed to generate a plurality of matching data sets comprises the following steps:

dividing the palm print image to be processed into a preset number of block palm print images, and executing the following steps aiming at each registered palm print image:

determining block registration images corresponding to the block palm print images in the registered palm print images according to the arrangement positions of the block palm print images in the palm print image to be processed;

and according to each blocked palm print image and a blocked registration image corresponding to each blocked palm print image, performing blocked matching on the registered palm print image and the palm print image to be processed to generate a matching data set.

Preferably, the step of performing block matching on the registered palm print image and the to-be-processed palm print image according to each of the block palm print images and the block registration image corresponding to each of the block palm print images, and generating a matching data set includes:

calling each block palm print image, and executing the following steps aiming at each block palm print image:

judging whether a key data point exists in the segmented palm print image, if so, determining the segmented palm print image as a first matching unit, and reading a first characteristic value corresponding to the first matching unit;

searching a target block registration image corresponding to the first matching unit in block registration images respectively corresponding to the block palm print images and other block registration images adjacent to the target block registration image, and determining the target block registration image and the other block registration images as a second matching unit;

and reading a second characteristic value corresponding to the second matching unit, and performing block matching on the registered palm print image and the palm print image to be processed according to the first characteristic value and the second characteristic value to generate a matching data set.

Preferably, the step of performing block matching on the registered palm print image and the to-be-processed palm print image according to the first feature value and the second feature value to generate a matching data set includes:

calculating a first similarity value between the first characteristic value and the second characteristic value, and determining a first matching data pair between each registered data point in the second matching unit and each key data point in the first matching unit according to the first similarity value;

updating the first matching unit, the second matching unit, the first characteristic value and the second characteristic value, calculating a second similarity value between the updated first characteristic value and the updated second characteristic value, and determining a second matching data pair between each registered data point in the updated first matching unit and each key data point in the updated second matching unit according to the second similarity value;

determining a target data pair between the segmented palm print image and the target segmented registration image according to each first matching data pair and each second matching data pair;

after the target data pairs are generated in each blocked palm print image in the palm print image to be processed, block matching between the registered palm print image and the palm print image to be processed is completed, and each target data pair is generated into a matching data set.

Preferably, the step of updating the first matching unit, the second matching unit, the first feature value and the second feature value includes:

searching other blocked palm print images adjacent to the first matching unit, and taking the first matching unit and the other blocked palm print images as a new second matching unit;

updating the second characteristic value according to the characteristic value corresponding to the new second matching unit;

and taking the target block registration image as a new first matching unit, and updating the first characteristic value according to the characteristic value corresponding to the new first matching unit.

Preferably, the step of determining a target palm print image matched with the to-be-processed palm print image in each registered palm print image according to the plurality of matching data sets includes:

counting the number of target data pairs contained in each matching data group, and determining the target number with the maximum value in the numbers;

and searching a target matching data group corresponding to the target number in the plurality of matching data groups, and determining the registered palm print image corresponding to the target matching data group as the target palm print image.

Preferably, when the palm print image to be processed is acquired, the step of reading the registered palm print image in the preset database includes:

when a palm print image is received, according to a preset network model, performing left-hand and right-hand identification on the palm print image, and determining left-hand and right-hand attributes of the palm print image;

identifying a line cutting point on the palm print image according to the attributes of the left hand and the right hand;

and cutting the palm print image according to the line cutting points to generate a palm print image to be processed.

In order to achieve the above object, the present invention further provides a palm print image recognition apparatus, including:

the reading module is used for reading the registered palm print image in the preset database when the palm print image to be processed is obtained;

the matching module is used for respectively performing block matching on each registered palm print image and the to-be-processed palm print image to generate a plurality of matching data sets, wherein one registered palm print image correspondingly generates one matching data set;

and the identification module is used for determining a target palm print image matched with the palm print image to be processed in each registered palm print image according to the plurality of matching data groups.

Further, in order to achieve the above object, the present invention further provides a palm print image recognition apparatus, where the palm print image recognition apparatus includes a memory, a processor, and a palm print image recognition program stored in the memory and executable on the processor, and the palm print image recognition program implements the steps of the palm print image recognition method when executed by the processor.

In addition, in order to achieve the above object, the present invention further provides a computer readable storage medium, on which a palm print image recognition program is stored, and the palm print image recognition program realizes the steps of the palm print image recognition method when executed by a processor.

The invention provides a palm print image identification method, a palm print image identification device and a computer readable storage medium, wherein when a to-be-processed palm print image is obtained and a palm print identification requirement is represented, a registered palm print image is read from a preset database, and each registered palm print image is respectively matched with the to-be-processed palm print image in a blocking mode to obtain a plurality of matched data groups; each registered palm print image is matched with a palm print image to be processed in a blocking mode to generate a matching data set, and the matching degree between each registered palm print image and the palm print image to be processed is represented; and determining a target palm print image which is most matched with the palm print image to be processed from the registered palm print images according to the plurality of matching data groups, wherein the target palm print image is the palm print image for identifying the palm print image to be processed, and the identification of the palm print image to be processed is realized. The block matching is a mechanism for dividing the palm print image into a plurality of block images for matching, and each registered palm print image is respectively matched with the palm print image to be processed in a block mode, so that the matching with the whole palm print image as a reference can be avoided, the data amount of reference processing in the matching process is reduced, and the palm print image identification efficiency is improved.

Drawings

Fig. 1 is a schematic structural diagram of a palm print image recognition device of a hardware operating environment according to an embodiment of the present invention;

FIG. 2 is a flowchart illustrating a first embodiment of a method for identifying a palm print image according to the present invention;

fig. 3 is a functional block diagram of a palm print image recognition apparatus according to a preferred embodiment of the invention.

The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.

Detailed Description

It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

As shown in fig. 1, fig. 1 is a schematic structural diagram of a palm print image recognition device of a hardware operating environment according to an embodiment of the present invention.

In the following description, suffixes such as "module", "component", or "unit" used to denote elements are used only for facilitating the explanation of the present invention, and have no specific meaning in itself. Thus, "module", "component" or "unit" may be used mixedly.

The palm print image recognition device of the embodiment of the invention can be a PC, and can also be a mobile terminal device such as a tablet computer and a portable computer.

As shown in fig. 1, the palm print image recognition apparatus may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.

It will be understood by those skilled in the art that the palm print image recognition device configuration shown in fig. 1 does not constitute a limitation of the palm print image recognition device, and may include more or fewer components than those shown, or some components in combination, or a different arrangement of components.

As shown in fig. 1, a memory 1005, which is a kind of computer-readable storage medium, may include therein an operating system, a network communication module, a user interface module, and a detection program.

In the device shown in fig. 1, the network interface 1004 is mainly used for connecting to a backend server and performing data communication with the backend server; the user interface 1003 is mainly used for connecting a client (user side) and performing data communication with the client; and the processor 1001 may be configured to call the detection program stored in the memory 1005 and perform the following operations:

Further, each registered palm print image exists in a preset database in a form of being divided into a preset number of block registered images;

Further, the step of performing block matching on the registered palm print image and the to-be-processed palm print image according to each of the block palm print images and the block registration images corresponding to the block palm print images, respectively, and generating a matching data set includes:

Further, the step of performing block matching on the registered palm print image and the to-be-processed palm print image according to the first feature value and the second feature value to generate a matching data set includes:

Further, the step of updating the first matching unit, the second matching unit, the first feature value, and the second feature value includes:

Further, the step of determining a target palm print image matched with the to-be-processed palm print image in each registered palm print image according to the plurality of matching data sets includes:

Further, before the step of reading the registered palm print image in the preset database when the to-be-processed palm print image is acquired, the processor 1001 may be configured to call the detection program stored in the memory 1005, and perform the following operations:

The specific implementation of the palm print image recognition device of the present invention is basically the same as the following embodiments of the palm print image recognition method, and is not described herein again.

For a better understanding of the above technical solutions, exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

In order to better understand the technical solution, the technical solution will be described in detail with reference to the drawings and the specific embodiments.

Referring to fig. 2, a first embodiment of the invention provides a flowchart illustrating a method for recognizing a palm print image. In this embodiment, the method for identifying a palm print image includes the following steps:

step S10, when acquiring the palm print image to be processed, reading the registered palm print image in the preset database;

the method for identifying the palm print image in the embodiment is applied to the server and is suitable for identifying the palm print image through the server. The server is connected with terminals with camera shooting functions, such as a smart phone, a tablet personal computer, a video camera and a digital camera, and the taken palm print images are uploaded to the server by the intelligent terminals for identification processing.

Understandably, the palm is divided into a left palm and a right palm, different palm textures have different characteristics, and in order to identify the palm print image in a targeted manner, left-hand and right-hand attribute identification needs to be performed on the palm print image, that is, whether the shot palm print image is from the left palm or the right palm is determined. In addition, considering that the captured palm print image may include other background objects to affect the recognition effect, it is necessary to determine an effective area in the palm print image before the recognition processing, and improve the accuracy of the recognition processing and reduce the data amount of the recognition processing by performing the recognition processing on the effective area. The left-right hand attribute identification and the effective area determination are set as a preprocessing mechanism and are carried out before the palm print image is identified and processed by the server. Specifically, when the palm print image to be processed is acquired, the step of reading the registered palm print image in the preset database includes:

a1, when receiving a palm print image, identifying the left hand and the right hand of the palm print image according to a preset network model, and determining the left hand and the right hand attributes of the palm print image;

further, the server is trained with a preset network model in advance, and the preset network model is trained by a large number of left palm image samples and right palm image samples to identify left and right hand attributes; meanwhile, the preset network model is trained through a large number of left-hand palm print line samples and right-hand palm print line samples so as to identify effective areas in the palm print images. After receiving the uploaded palm print image, the server calls a preset network model, the preset network model identifies the left hand and the right hand of the palm print image, and determines whether the palm print image is consistent with the left palm attribute or the right palm attribute; thus, the left and right hand attributes of the palm print image are determined.

Step a2, identifying a texture cutting point on the palm print image according to the attributes of the left hand and the right hand;

furthermore, after the left-hand and right-hand attributes of the palm print image are determined, the texture clipping points in the palm print image are identified according to the texture samples corresponding to the left-hand and right-hand attributes in the preset network model. The pattern cutting point is preferably a pattern of a first joint of four fingers of a palm, and the palm pattern image is identified through a preset network model to obtain the pattern of the first joint of the four fingers as the pattern cutting point.

Step a3, cutting the palm print image according to the line cutting points to generate a palm print image to be processed.

Furthermore, the palm print image is cut according to the print cutting point, and the print of the first joint of the index finger and the print of the first joint of the little finger in the print cutting point are connected as two connecting points to form a cutting connecting line. And then cutting the palm print image according to the connecting line to obtain an effective area, and taking the effective area as the palm print image to be processed for the server to identify and process.

Furthermore, the server is interfaced with a preset database for storing the palm print image entered at the time of registration, and after the palm print image to be processed is generated, the palm print image entered at the time of registration is read as the registered palm print image in the preset data. And identifying the palm print image to be processed according to the similarity between each registered palm print image and the palm print image to be processed.

Step S20, respectively performing block matching on each registered palm print image and the to-be-processed palm print image to generate a plurality of matching data sets, wherein one registered palm print image correspondingly generates one matching data set;

furthermore, the server is preset with a block matching mechanism for determining the similarity between the palm print images, and the block matching is a mechanism for dividing the palm print images into a plurality of block images for matching. After the to-be-processed palm print image is generated and the registered palm print image is read, each registered palm print image can be respectively matched with the to-be-processed palm print image in a blocking mode. Dividing the palm print image to be processed into a plurality of block images, dividing each registered palm print image into a plurality of block images, matching the two block images, and determining the similarity between the palm print image to be processed and each registered palm print image according to the matching degree between the block images. The registered palm print images can be matched with the to-be-processed palm print images one by one or matched with the to-be-processed palm print images simultaneously in the process of matching the block images of the to-be-processed palm print images with the block images of the registered palm print images because of the large number of the registered palm print images, and the method is not limited to this. And matching each registered palm print image with the palm print image to be processed to generate a matching data group, and representing the matching degree between each palm print characteristic point in the registered palm print image and each palm print characteristic point in the palm print image to be processed.

Step S30, determining a target palm print image matched with the to-be-processed palm print image in each registered palm print image according to the plurality of matching data sets.

Furthermore, because the matching data group represents the matching degree of each palm print feature point between the registered palm print image and the palm print image to be processed, after each registered palm print image and each palm print image to be processed are subjected to block matching, and respective matching data groups are generated, a target matching data group with the highest matching degree in each matching data group can be searched in a comparison mode, the palm print feature point between the registered palm print image and the palm print image to be processed of the target matching data group has the highest similarity, so that the target matching data group is determined as the target palm print image matched with the palm print image to be processed in the registered palm print image, the palm print image to be processed is recognized as the target palm print image, and the recognition of the palm print image to be processed is completed.

In the method for identifying a palm print image of the embodiment, when a to-be-processed palm print image is obtained and a palm print identification requirement is represented, registered palm print images are read from a preset database, and each registered palm print image is respectively matched with the to-be-processed palm print image in a blocking manner to obtain a plurality of matched data sets; each registered palm print image is matched with a palm print image to be processed in a blocking mode to generate a matching data set, and the matching degree between each registered palm print image and the palm print image to be processed is represented; and determining a target palm print image which is most matched with the palm print image to be processed from the registered palm print images according to the plurality of matching data groups, wherein the target palm print image is the palm print image for identifying the palm print image to be processed, and the identification of the palm print image to be processed is realized. The block matching is a mechanism for dividing the palm print image into a plurality of block images for matching, and each registered palm print image is respectively matched with the palm print image to be processed in a block mode, so that the matching with the whole palm print image as a reference can be avoided, the data amount of reference processing in the matching process is reduced, and the palm print image identification efficiency is improved.

Further, based on the first embodiment of the method for identifying a palm print image of the present invention, a second embodiment of the method for identifying a palm print image of the present invention is provided, in the second embodiment, each registered palm print image exists in a preset database in a form of being divided into a preset number of block-wise registered images, and the step of performing block-wise matching on each registered palm print image and the palm print image to be processed to generate a plurality of matching data sets includes:

step S21, dividing the palm print image to be processed into a preset number of block palm print images, and executing the following steps for each registered palm print image:

in this embodiment, the block images of the registered palm print images are used as the block registration images of the registered palm print images, and are stored in the preset database in a block form. The preset database sets a specific storage unit for each registered palm print image, blocks the storage unit, and stores one blocked registered image in each block. And simultaneously searching the key data points and the characteristic values of the key data points contained in each block registration image, and storing the key data points and the block registration images into the blocks of the storage unit. The key data point is preferably a line inflection point in the palm print image, and the characteristic value is a numerical value for describing the inflection point, and may be a string of characters converted from pixel values of points around the inflection point, and the like. In each matching process, the characteristic values of the key data points corresponding to the block registration images can be directly read from the blocks for processing, and the blocking and characteristic value extraction processing aiming at the registered palm print images at each time is avoided, so that the processing efficiency is improved. The number of blocks of the registered palm print image is determined by a preset number, that is, data of the blocks required to be registered is preset as the preset number according to requirements, for example, the number of blocks of n × n. And according to the preset quantity, carrying out blocking processing on each registered palm print image in advance to obtain respective n × n blocked registered images, and extracting key data points and characteristic values of the key data points in each blocked registered image to store in blocks of respective corresponding storage units.

Further, after the palm print image has the identification requirement and is preprocessed to generate a to-be-processed palm print image, reading the registered palm print images in the preset database, namely reading the characteristic values of the respective blocked registered images of the registered palm print images for characteristic matching with the to-be-processed palm print image. And dividing the palm print images to be processed according to the preset number to obtain block palm print images of preset data, and performing matching processing on each block registration image of each registered palm print image and each block palm print image of the palm print images to be processed.

Step S22, determining block registration images corresponding to the block palm print images in the registered palm print images according to the arrangement positions of the block palm print images in the palm print image to be processed;

furthermore, in the process of dividing the palmprint image to be processed into a preset number of segmented palmprint images, numbering the divided segmented palmprint images according to the preset number, wherein the numbering is [1 × 1], [1 × 2] · [2 × 1], [2 × 2 · · according to the number of the n × n segments; and determining the arrangement position of each block palm print image in the palm print image to be processed through each serial number. And numbering each block registration image in the registered palm print images according to a preset number, performing matching processing according to the same number between each block registration image and each block palm print image in the process of matching the block registration images and the block palm print images, wherein the same number represents the same arrangement position so as to ensure the matching accuracy. And determining the blocked registration images corresponding to the blocked palm print images in the registered palm print images according to the arrangement positions of the blocked palm print images in the palm print image to be processed, wherein the blocked palm print images and the blocked registration images corresponding to the blocked palm print images have the same serial numbers and the same representation arrangement positions.

Step S23, performing block matching on the registered palm print image and the to-be-processed palm print image according to each of the block palm print images and the block registration images corresponding to each of the block palm print images, respectively, to generate a matching data set.

Further, according to the matching between each block palm print image and the corresponding block registration image, block matching is carried out on the registered palm print image and the palm print image to be processed. And after each block palm print image is matched with the corresponding block registration image, completing block matching between the registered palm print image and the palm print image to be processed, and generating a matching data set between the registered palm print image and the palm print image to be processed. And after each registered palm print image is matched with the palm print image to be processed in a blocking mode, a plurality of matching data sets are generated to represent the similarity between each registered palm print image and the palm print image to be processed. The method comprises the following steps of performing block matching on a registered palm print image and a palm print image to be processed according to each block palm print image and a block registration image corresponding to each block palm print image respectively, and generating a matching data set, wherein the step comprises the following steps of:

step S231, calling each of the segmented palm print images, and executing the following steps for each of the segmented palm print images:

furthermore, after the palm print image to be processed is divided into all the block palm print images and all the block registration images with the same positions as the palm print images are determined, each block palm print image and the corresponding block registration image can be called to be matched.

Step S232, judging whether a key data point exists in the segmented palm print image, if so, determining the segmented palm print image as a first matching unit, and reading a first characteristic value corresponding to the first matching unit;

furthermore, whether the called segmented palm print image contains the relevant key data point or not is judged firstly, and if the key data point exists, the called segmented palm print image is used as a first matching unit. Thereafter, the feature value of the key data point contained therein is extracted as a first feature value corresponding to the first matching unit.

Step S233, searching for a target block registration image corresponding to the first matching unit in block registration images corresponding to the respective block palm print images, and other block registration images adjacent to the target block registration image, and determining the target block registration image and the other block registration images as a second matching unit;

further, the block registration images corresponding to the block palm print images are searched, the block registration image corresponding to the first matching unit is determined, and the block registration image obtained through searching is used as the target block registration image. And meanwhile, searching other block registration images adjacent to the target block registration image, and further forming the searched target block registration image and other block registration images into a second matching unit. The number of other adjacent block registration images is set according to requirements, for example, two adjacent block registration images in the left-right direction, or four adjacent block registration images in the up-down, left-right direction, or eight adjacent block registration images in the periphery are set; the present embodiment sets the number to eight block registration images around in consideration of the accuracy of matching. If the arrangement position of the segmented palm print image is characterized as [5 × 5] through the number, the number of the corresponding target segmented registration image is also [5 × 5], and the numbers of other segmented registration images adjacent to the target segmented registration image are [4 × 4], [4 × 5], [4 × 6], [5 × 4], [5 × 6], [6 × 4], [6 × 5], [6 × 6], so that one segmented palm print image with the number [5 × 5] is formed into a first matching unit, and the two segmented registration images with the numbers [4 × 4], [4 × 5], [4 × 6], [5 × 4], [6 × 4], [5 × 6], [ 6] are formed into a ninth matching unit.

It should be noted that, for the segmented palm print images located at the vertex positions of the four corners of the palm print image and the segmented palm print images located at the four sides of the palm print image, the number of the other adjacent segmented registered images is different according to the position. For the segmented palm print images at the four corner vertex positions, the number of other adjacent segmented registration images is 3, and the 3 other segmented registration images and the target segmented registration image are formed into a second matching unit; for the segmented palm print images at the four sides, the number of other adjacent segmented registration images is 5, and the 5 other segmented registration images and the target segmented registration image are formed together as the second matching unit.

Step S234, reading a second feature value corresponding to the second matching unit, and performing block matching on the registered palm print image and the to-be-processed palm print image according to the first feature value and the second feature value to generate a matching data set.

Further, the blocks storing the block registration images forming the second matching unit are searched, respective feature values are extracted from the searched blocks, and the respective feature values are formed into second feature values corresponding to the second matching unit. And then according to key data points represented by the first characteristic value and the second characteristic value, block matching is carried out on the registered palm print image and the palm print image to be processed, after the first characteristic value formed by each block palm print image of the palm print image to be processed is matched with the second characteristic value formed by the corresponding block registered image in the registered image, block matching between the registered palm print image and the palm print image to be processed is completed, and a matching data group between the registered palm print image and the palm print image to be processed is generated.

According to the first characteristic value and the second characteristic value, the registered palm print image and the palm print image to be processed are subjected to block matching, and the step of generating the matching data set comprises the following steps:

step b1, calculating a first similarity value between the first characteristic value and the second characteristic value, and determining a first matching data pair between each registered data point in the second matching unit and each key data point in the first matching unit according to the first similarity value;

furthermore, the server calculates a first similarity value between the first characteristic value and the second characteristic value in a preset mode, wherein the preset mode can be set as a cosine distance or an Euclidean distance according to requirements. The first similarity calculation between the first characteristic value and the second characteristic value is a similarity calculation between the characteristic value of the key data point included in the first matching unit and the characteristic value of the registration data point included in the second matching unit. And searching each registration data point which is most similar to each key data point of the first matching unit in each registration data point contained in the second matching unit through the calculated similarity. It should be noted that the registered data points included in the second matching unit are the key data points in the second matching unit.

Understandably, the number of the key data points included in the first matching unit may be single or multiple, and the number of the registration data points included in the second matching unit may be single or multiple, the feature values of the key data points included in the first matching unit form first elements of the first feature value, and the feature values of the registration data points included in the second matching unit form second elements of the second feature value. During calculation, similarity calculation is performed on the first elements and each second element one by one on the basis of the first elements contained in the first characteristic values, so that a plurality of first similarity values of each first element are obtained. And comparing the plurality of first similarity values, determining the first similarity value with the largest value, and generating the second element with the largest first similarity value and the first element with the highest similarity degree. Thus, the registration data point corresponding to the second element and the key data point corresponding to the first element can be formed as a first matching data pair between the registration data point in the second matching unit and the key data point in the first matching unit. And searching each first element contained in the first characteristic value according to the respective first similarity value to obtain a second element with the highest similarity degree, and forming a plurality of first matching data pairs between each registration data point in the second matching unit and each key data point in the first matching unit so as to obtain each registration data point which is most similar to each key data point in the first matching unit in each registration data point contained in the second matching unit.

Step b2, updating the first matching unit, the second matching unit, the first eigenvalue and the second eigenvalue, calculating a second similarity value between the updated first eigenvalue and the updated second eigenvalue, and determining a second matching data pair between each registered data point in the updated first matching unit and each key data point in the updated second matching unit according to the second similarity value;

further, in order to improve the matching accuracy, the present embodiment is provided with a mechanism for updating the first matching unit, the second matching unit, the first feature value, and the second feature value, wherein the updating is implemented by interchanging the generation manners of the first matching unit and the second matching unit. Specifically, the step of updating the first matching unit, the second matching unit, the first feature value, and the second feature value includes:

b21, searching other block palm print images adjacent to the first matching unit, and taking the first matching unit and the other block palm print images as a new second matching unit;

and searching other blocked palm print images adjacent to the first matching unit, and forming the first matching unit and other searched blocked registered palm print images into a new second matching unit together so as to update the second characteristic value by updating the second matching unit.

Step b22, updating the second characteristic value according to the characteristic value corresponding to the new second matching unit;

furthermore, extracting key data points contained in each segmented palm print image forming a new second matching unit, identifying a feature value contained in each extracted key data point, wherein the feature value of each identified key data point is the feature value corresponding to the new second matching unit, and replacing the original second feature value with the corresponding feature value to update the second feature value.

And b23, taking the target block registration image as a new first matching unit, and updating the first characteristic value according to the characteristic value corresponding to the new first matching unit.

Further, a target block registration image corresponding to the original first matching unit is used as a new first matching unit, a feature value of the target block registration image is extracted from blocks storing the target block registration image and used as a feature value corresponding to the new first matching unit, and the original first feature value is replaced by the corresponding feature value, so that the first feature value is updated.

Furthermore, the similarity between the updated first characteristic value and the second characteristic value is calculated in a preset mode, the obtained calculation result is the second similarity between the updated first characteristic value and the second characteristic value, and each registered data point in each key data point contained in the new second matching unit and the new first matching unit is searched as each similar key data point according to the calculated similarity. The similarity between the first element and the second element is determined by taking the feature value of each registered data point as the first element and the second element formed by the feature values of the key data points, and determining the second similarity value with the largest value from the plurality of second similarity values. Thus, the key data point corresponding to the second element and the registration data point corresponding to the first element may be formed as a second matching data pair between the registration data point in the new first matching unit and the key data point in the new second matching unit. And after the first elements formed by the characteristic values of the registered data points are searched and obtained according to the respective second similarity values to obtain the second elements with the highest respective similarity degree, forming a plurality of second matching data pairs between the registered data points in the new first matching unit and the key data points in the new second matching unit.

Step b3, determining a target data pair between the segmented palm print image and the target segmented registration image according to each first matching data pair and each second matching data pair;

understandably, each first matching data pair represents the matching relationship between the block registration image corresponding to the block palm print image and the adjacent block registration image thereof and between each key data point and each registration data point on the basis of the block palm print image; each key data point has a unique registered data point to match. Each second matching data pair represents the matching relationship between the segmented palm print image corresponding to the segmented registration image and the temporary segmented palm print image thereof and between each registration data point and each key data point on the basis of the segmented registration image; each registered data point has a unique key data point to match. With each first matching data pair and each second matching data pair, it can be determined whether the matching relationship between the key data points and the registered data points generated in different ways has changed. If the key data point is not changed, the matching between the key data point and the registered data point is a valid matching, and the key data point and the registered data point have higher similarity. If the key data point a1 matches the registration data point b1 in the first matching data pair, and the key data point a1 matches the registration data point b1 in the second matching data pair, the similarity between a1 and b1 is determined to be high, and the two are valid matches. If the data is changed, the matching between the key data point and the registered data point is invalid, and the similarity between the key data point and the registered data point is crossed. If the key data point a1 matches the registration data point b1 in the first matching data pair and the key data point a1 matches the registration data point b2 in the second matching data pair, then the similarity between a1 and b1 and between a1 and b2 are determined to be low, and the two are invalid matches.

Further, according to each first matching data pair and each second matching data pair, effective matching key data points and registration data points are determined, and each effective matching key data point and registration data point is used as a target data pair between the segmented palm print image and the segmented registration image.

And b4, after each blocked palm print image in the palm print image to be processed generates the target data pair, completing the blocked matching between the registered palm print image and the palm print image to be processed, and generating each target data pair into a matched data set.

Furthermore, after each segmented palm print image in the palm print image to be processed is matched with each corresponding segmented registered image in the registered palm print image in a segmented manner to generate respective target data pairs, the segmented matching between the registered palm print image and the palm print image to be processed is completed, and each target matching pair is generated into a matching data set. Therefore, the overall similarity between the registered palm print image and the palm print image to be processed is represented through the similarity of the block images between each block palm print image in the palm print image to be processed represented by each target matching pair and each block registered image in the corresponding registered palm print image, namely the overall similarity between the registered palm print image and the palm print image to be processed is represented by the matching data group formed by each target data pair.

In the embodiment, a block matching mechanism is set in the process of matching the registered palm print image and the to-be-processed palm print image, and a smaller block and a larger block are matched in the block matching process, so that the displacement error in the cutting process of the palm print image is compensated, and the corresponding matching point is prevented from falling outside the matching area. Meanwhile, after the smaller blocks are matched with the larger blocks, the original smaller blocks are formed into larger blocks, the original larger blocks are formed into the smallest blocks, and matching is performed again, so that single matching error is avoided, the matching accuracy is improved, and the determined similarity between the registered palm print image and the palm print image to be processed is more accurate.

Further, based on the first embodiment or the second embodiment of the method for identifying a palm print image of the present invention, a third embodiment of the method for identifying a palm print image of the present invention is provided, and in the third embodiment, the step of determining a target palm print image in each registered palm print image, which matches the palm print image to be processed, according to a plurality of the matching data sets includes:

step S31, counting the number of target data pairs contained in each matching data group, and determining the target number with the maximum value in the numbers;

after obtaining a plurality of matching data sets between each registered palm print image and the to-be-processed palm print image, the embodiment may determine the registered palm print image with the maximum similarity to the to-be-processed palm print image according to the overall similarity between each registered palm print image and the to-be-processed palm print image represented by each matching data set. Specifically, the number of the target data pairs included in each matching data group is counted, and the greater the number of the target data pairs included in the matching data group is, the higher the similarity between each segmented palm print image in the palm print image to be processed and each segmented registered image in the registered palm print image is, the higher the overall similarity between the registered palm print image and the palm print image to be processed is. And after the number of target data pairs contained in each matching data group is obtained through statistics, comparing the data, determining the target number with the maximum value, and determining the registered palm print image with the highest similarity to the palm print image to be processed according to the target data pairs.

Step S32, finding a target matching data set corresponding to the target number from the multiple matching data sets, and determining a registered palm print image corresponding to the target matching data set as the target palm print image.

Further, according to the target number, searching a plurality of matching data sets, searching the matching data sets containing the target number of the target data pairs as the target number, using the matching data sets as the target matching data sets, further searching the registered palm print images generating the target matching data sets as the registered palm print images corresponding to the target matching data sets, wherein the corresponding registered palm print images are the target palm print images matched with the palm print images to be processed, and realizing the identification of the palm print images to be processed.

In the embodiment, the corresponding registered palm print image is determined by the matching data group with the largest number of target data pairs, and is identified as the target palm print image matched with the palm print image to be processed. The matching data group with the largest number of target data represents the maximum similarity between the registered palm print image and the palm print image to be processed, so that the determined target palm print image and the palm print image to be processed have the maximum similarity, and the accuracy of identifying the palm print image to be processed is ensured.

Furthermore, the invention also provides a palm print image recognition device.

Referring to fig. 3, fig. 3 is a functional module diagram of a first embodiment of the palm print image recognition apparatus of the present invention. The palm print image recognition device comprises:

the reading module 10 is used for reading the registered palm print image in the preset database when the palm print image to be processed is obtained;

a matching module 20, configured to perform block matching on each registered palm print image and the to-be-processed palm print image, respectively, to generate a plurality of matching data sets, where one registered palm print image generates one matching data set correspondingly;

and the identification module 30 is configured to determine, according to the plurality of matching data sets, a target palm print image that matches the to-be-processed palm print image in each registered palm print image.

In the palm print image recognition device of the embodiment, when a to-be-processed palm print image is obtained and a palm print recognition requirement is represented, a reading module 10 reads a registered palm print image from a preset database, and a matching module 20 performs block matching on each registered palm print image and the to-be-processed palm print image respectively to obtain a plurality of matching data sets; each registered palm print image is matched with a palm print image to be processed in a blocking mode to generate a matching data set, and the matching degree between each registered palm print image and the palm print image to be processed is represented; and then the identification module 30 determines a target palm print image most matched with the palm print image to be processed from the registered palm print images according to the plurality of matching data sets, wherein the target palm print image is the palm print image identified by the palm print image to be processed, and the identification of the palm print image to be processed is realized. The block matching is a mechanism for dividing the palm print image into a plurality of block images for matching, and each registered palm print image is respectively matched with the palm print image to be processed in a block mode, so that the matching with the whole palm print image as a reference can be avoided, the data amount of reference processing in the matching process is reduced, and the palm print image identification efficiency is improved.

Further, each of the registered palm print images exists in a preset database in a form of being divided into a preset number of block registration images, and the matching module 20 includes:

a dividing unit, configured to divide the to-be-processed palm print image into a preset number of block palm print images, and execute the following steps for each registered palm print image:

a determining unit, configured to determine, according to an arrangement position of each of the segmented palm print images in the to-be-processed palm print image, a segmented registration image corresponding to each of the segmented palm print images in the registered palm print image;

and the matching unit is used for performing block matching on the registered palm print image and the palm print image to be processed according to each block palm print image and the block registration images corresponding to the block palm print images respectively to generate a matching data set.

Further, the matching unit is further configured to:

Further, the identification module 30 further includes:

a counting unit, configured to count the number of target data pairs included in each of the matching data sets, and determine a maximum target number among the numbers;

and the searching unit is used for searching a target matching data group corresponding to the target number in the matching data groups and determining the registered palm print image corresponding to the target matching data group as the target palm print image.

Further, the device for recognizing a palm print image further comprises:

the determining module is used for identifying the left hand and the right hand of the palm print image according to a preset network model when the palm print image is received, and determining the left hand and the right hand attributes of the palm print image;

the identification module is also used for identifying the line cutting points on the palm print image according to the left-hand and right-hand attributes;

and the generating module is used for cutting the palm print image according to the line cutting points to generate a palm print image to be processed.

The specific implementation of the palm print image recognition device of the present invention is basically the same as the above palm print image recognition method, and is not described herein again.

In addition, the embodiment of the invention also provides a computer readable storage medium.

The computer readable storage medium has stored thereon a palm print image recognition program, which when executed by a processor implements the steps of the palm print image recognition method as described above.

The specific implementation of the computer-readable storage medium of the present invention is substantially the same as the embodiments of the palm print image recognition method described above, and will not be described herein again.

It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.

The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.

Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a computer-readable storage medium (such as ROM/RAM, magnetic disk, optical disk) as described above, and includes several instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.

The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.

Claims

1. A palm print image recognition method is characterized by comprising the following steps:

2. The method for recognizing palm print images according to claim 1, wherein each of the registered palm print images exists in a predetermined database in the form of being divided into a predetermined number of block-wise registered images;

3. The method for identifying a palm print image according to claim 2, wherein the step of performing block matching on the registered palm print image and the palm print image to be processed according to each of the block palm print images and a block registration image corresponding to each of the block palm print images, respectively, to generate a matching data set comprises:

4. The palm print image recognition method according to claim 3, wherein the step of performing block matching on the registered palm print image and the palm print image to be processed according to the first feature value and the second feature value to generate a matching data set comprises:

5. The palm print image recognition method according to claim 4, wherein the step of updating the first matching unit, the second matching unit, the first feature value, and the second feature value includes:

6. The method for identifying a palm print image according to any one of claims 1 to 5, wherein the step of determining a target palm print image matching the palm print image to be processed in each of the registered palm print images based on a plurality of the matching data sets comprises:

7. The method for identifying a palm print image according to any one of claims 1 to 5, wherein the step of reading the registered palm print image in the preset database when the palm print image to be processed is acquired comprises:

8. A palm print image recognition apparatus, comprising:

9. A palm print image recognition apparatus, characterized in that the palm print image recognition apparatus comprises a memory, a processor and a palm print image recognition program stored on the memory and executable on the processor, the palm print image recognition program, when executed by the processor, implementing the steps of the palm print image recognition method according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that the computer-readable storage medium has stored thereon a palm print image recognition program, which when executed by a processor implements the steps of the palm print image recognition method according to any one of claims 1 to 7.