WO2021151359A1 - 掌纹图像的识别方法、装置、设备及计算机可读存储介质 - Google Patents

掌纹图像的识别方法、装置、设备及计算机可读存储介质 Download PDF

Info

Publication number
WO2021151359A1
WO2021151359A1 PCT/CN2020/135844 CN2020135844W WO2021151359A1 WO 2021151359 A1 WO2021151359 A1 WO 2021151359A1 CN 2020135844 W CN2020135844 W CN 2020135844W WO 2021151359 A1 WO2021151359 A1 WO 2021151359A1
Authority
WO
WIPO (PCT)
Prior art keywords
palmprint
image
matching
registered
images
Prior art date
Application number
PCT/CN2020/135844
Other languages
English (en)
French (fr)
Inventor
刘翔
刘莹
Original Assignee
平安科技(深圳)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安科技(深圳)有限公司 filed Critical 平安科技(深圳)有限公司
Publication of WO2021151359A1 publication Critical patent/WO2021151359A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12Fingerprints or palmprints
    • G06V40/1347Preprocessing; Feature extraction
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00Pattern recognition
    • G06F18/20Analysing
    • G06F18/22Matching criteria, e.g. proximity measures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12Fingerprints or palmprints
    • G06V40/1365Matching; Classification

Definitions

  • This application relates to the field of image processing technology, and in particular to a palmprint image recognition method, device, equipment, and computer-readable storage medium.
  • palmprint can effectively identify a person's identity, so it can be widely used in various application scenarios that require person identification, such as unmanned supermarkets and workplace attendance.
  • the traditional palmprint recognition is usually realized by scanning the palm image of a fixed device, which belongs to palm contact recognition; long-term contact can easily cause equipment pollution, make the scanned palm image unclear, and affect the recognition effect.
  • a palmprint image recognition method includes the following steps:
  • a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images is determined.
  • a palmprint image recognition device comprising:
  • the reading module is used to read the registered palmprint image in the preset database when the palmprint image to be processed is obtained;
  • the matching module is used to block and match each of the registered palmprint images with the palmprint images to be processed to generate a plurality of matching data groups, wherein one of the registered palmprint images corresponds to generating a matching data Group;
  • the recognition module is configured to determine a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images according to a plurality of the matching data sets.
  • the palmprint image recognition device includes a memory, a processor, and a palmprint image recognition program stored in the memory and running on the processor.
  • the palmprint image When the image recognition program is executed by the processor, the following steps are implemented:
  • a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images is determined.
  • a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images is determined.
  • the present application can avoid matching with the entire palmprint image as a reference, reducing the amount of reference processing data in the matching process, and improving the efficiency of palmprint image recognition.
  • FIG. 1 is a schematic structural diagram of a palmprint image recognition device in a hardware operating environment involved in a solution of an embodiment of the application;
  • FIG. 2 is a schematic flowchart of a first embodiment of a palmprint image recognition method according to this application;
  • FIG. 3 is a schematic diagram of functional modules of a preferred embodiment of a palmprint image recognition device according to the present application.
  • FIG. 1 is a schematic diagram of the structure of a palmprint image recognition device of a hardware operating environment involved in a solution of an embodiment of the present application.
  • the palmprint image recognition device in the embodiment of the present application may be a PC, or a portable terminal device such as a tablet computer and a portable computer.
  • the palmprint image recognition device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, and a communication bus 1002.
  • the communication bus 1002 is used to implement connection and communication between these components.
  • the user interface 1003 may include a display screen (Display) and an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.
  • the network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).
  • the memory 1005 may be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as a magnetic disk memory.
  • the memory 1005 may also be a storage device independent of the aforementioned processor 1001.
  • the structure of the palmprint image recognition device shown in FIG. 1 does not constitute a limitation on the palmprint image recognition device, and may include more or less components than shown in the figure, or a combination of certain components. Components, or different component arrangements.
  • the memory 1005 which is a computer-readable storage medium, may include an operating system, a network communication module, a user interface module, and a detection program.
  • the network interface 1004 is mainly used to connect to the back-end server and communicate with the back-end server;
  • the user interface 1003 is mainly used to connect to the client (user side) and communicate with the client;
  • the processor 1001 can be used to call the detection program stored in the memory 1005 and perform the following operations:
  • a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images is determined.
  • each of the registered palmprint images exists in a preset database in the form of being divided into a preset number of registered images
  • the step of performing block matching of each of the registered palmprint images with the palmprint images to be processed, and generating a plurality of matching data groups includes:
  • the registered palmprint image and the palmprint image to be processed are segmented and matched to generate a match Data group.
  • the registered palmprint image and the palmprint image to be processed are divided into the registered palmprint image and the palmprint image to be processed according to each of the divided palmprint images and the registered divided images corresponding to each of the divided palmprint images.
  • the steps to generate a matching data group include:
  • the step of performing block matching on the registered palmprint image and the palmprint image to be processed according to the first feature value and the second feature value, and generating a matching data group includes:
  • the first matching unit, the second matching unit, the first feature value, and the second feature value are updated, and the second similarity value between the updated first feature value and the second feature value is calculated , And according to the second similarity value, determine 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;
  • the target data pair is generated for each of the block palmprint images in the palmprint image to be processed, the block matching between the registered palmprint image and the palmprint image to be processed is completed, And each of the target data pairs is generated as a matching data group.
  • step of updating the first matching unit, the second matching unit, the first characteristic value, and the second characteristic value includes:
  • the target segmented registration image is used as a new first matching unit, and the first feature value is updated according to the feature value corresponding to the new first matching unit.
  • the step of determining a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images according to a plurality of the matching data sets includes:
  • the processor 1001 may be used to call the detection program stored in the memory 1005 and perform the following operations :
  • the palmprint image is cropped to generate a palmprint image to be processed.
  • the first embodiment of the present application provides a schematic flowchart of a palmprint image recognition method.
  • the palmprint image recognition method includes the following steps:
  • Step S10 when the palmprint image to be processed is acquired, read the registered palmprint image in the preset database
  • the palmprint image recognition method in this embodiment is applied to a server, and is suitable for recognizing palmprint images through the server.
  • the server is connected to a terminal with a camera function such as a smart phone, a tablet computer, a video camera, a digital camera, and the like, and this type of smart terminal uploads the captured palmprint image to the server for recognition processing.
  • the palm is divided into the left palm and the right palm. Different palm prints have different characteristics.
  • the left and right hand attributes of the palm print images need to be recognized, which means that the palm print images taken are from Yu left palm or right palm.
  • the captured palmprint image may contain other background objects and affect the recognition effect, it is necessary to determine the effective area in the palmprint image before the recognition process, and the effective area is recognized to improve Accuracy of recognition processing, and reduce the amount of data for recognition processing.
  • the identification of the left and right hand attributes and the determination of the effective area are set as a preprocessing mechanism, which is performed before the palmprint image recognition processing by the server.
  • the step of reading the registered palmprint image in the preset database includes:
  • Step a1 When a palmprint image is received, perform left and right hand recognition on the palmprint image according to a preset network model, and determine the left and right hand attributes of the palmprint image;
  • the server is pre-trained with a preset network model, which is trained through a large number of left palm image samples and right palm image samples to identify left and right hand attributes; at the same time, the preset network model also uses a large number of left palm patterns. The sample and the right palm pattern sample are trained to identify the effective area in the palmprint image.
  • the server After the server receives the uploaded palmprint image, it calls the preset network model, and the preset network model recognizes the palmprint image for left and right hands, and determines whether the palmprint image is consistent with the attributes of the left palm or the right palm; The left and right hand attributes of the palm print image.
  • Step a2 according to the attributes of the left and right hands, identify the pattern cutting points on the palmprint image
  • the pattern cutting points in the palmprint image are identified according to the pattern samples corresponding to the left and right hand attributes in the preset network model.
  • the texture cropping point is preferably the texture at the first joint of the four fingers of the palm.
  • the palmprint image is recognized through a preset network model, and the texture at the first joint of the four fingers is obtained as the texture cropping point.
  • Step a3 crop the palmprint image according to the texture cropping point to generate a palmprint image to be processed.
  • the palmprint image is cropped according to the texture cropping point, and the texture at the first joint of the index finger in the texture cropping point and the texture at the first joint of the little thumb are connected as two connection points to form a cropped connection line . Furthermore, the palmprint image is cropped according to the connecting line, that is, the effective area is obtained, and the effective area is used as the palmprint image to be processed for the server to recognize and process.
  • the server is connected to a preset database for storing palmprint images entered during registration. After the palmprint images to be processed are generated, the palmprint images entered during registration are used as the registered palmprint images in the preset data. Read it.
  • the palmprint image to be processed is recognized by the similarity between each registered palmprint image and the palmprint image to be processed.
  • Step S20 block matching each of the registered palmprint images with the palmprint images to be processed to generate a plurality of matching data groups, wherein one of the registered palmprint images corresponds to a matching data group;
  • the server presets a block matching mechanism for determining the similarity between palmprint images
  • the block matching is a mechanism for dividing the palmprint image into multiple block images for matching.
  • each registered palmprint image can be matched with the palmprint image to be processed respectively.
  • the palmprint image to be processed is divided into multiple block images, and each registered palmprint image is divided into multiple block images, and then the two types of block images are matched, and the matching between each block image Degree to determine the similarity between the palmprint image to be processed and each registered palmprint image.
  • the registered palmprint image and the registered palmprint image can be matched one by one.
  • the palmprint images to be processed can be matched, and each registered palmprint image can also be matched with the palmprint images to be processed at the same time. There is no restriction on this.
  • Each registered palmprint image is matched with the palmprint image to be processed, and a matching data set is generated to represent the difference between each palmprint feature point in the registered palmprint image and each palmprint feature point in the palmprint image to be processed Matching degree.
  • Step S30 Determine a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images according to a plurality of the matching data sets.
  • the matching data group characterizes the degree of matching of each palmprint feature point between the registered palmprint image and the palmprint image to be processed
  • block matching is performed on each registered palmprint image and the palmprint image to be processed.
  • the target matching data group with the highest degree of matching in each matching data group can be compared and searched, and the palmprint feature between the registered palmprint image of the target matching data group and the palmprint image to be processed can be generated
  • the point has the highest similarity, so it is determined as the target palmprint image in the registered palmprint image that matches the palmprint image to be processed, the palmprint image to be processed is recognized as the target palmprint image, and the palmprint image to be processed is completed Recognition.
  • the registered palmprint image is first read from the preset database, and each registered palmprint The image is separately matched with the palmprint image to be processed to obtain multiple matching data sets; each registered palmprint image is matched with the palmprint image to be processed, and a matching data set is generated to represent each registered palmprint The degree of matching between the image and the palmprint image to be processed; and based on multiple matching data sets, the target palmprint image that most closely matches the palmprint image to be processed is determined from each registered palmprint image, and the target palmprint image That is, the palmprint image to be processed is the palmprint image recognition, which realizes the recognition of the palmprint image to be processed.
  • block matching is a mechanism that divides the palmprint image into multiple block images for matching. By matching each registered palmprint image with the palmprint image to be processed, it can avoid using the entire palmprint image. For reference matching, the amount of reference processing data in the matching process is reduced, which is conducive to improving the efficiency of palmprint image recognition.
  • each registered palmprint image is divided into A preset number of registered images in blocks exist in a preset database, and the step of performing block matching of each of the registered palmprint images with the palmprint images to be processed, and generating a plurality of matching data groups includes :
  • Step S21 Divide the palmprint image to be processed into a preset number of divided palmprint images, and perform the following steps for each registered palmprint image:
  • the block image of each registered palmprint image is used as the block registration image of each registered palmprint image, and is stored in the preset database in the form of blocks.
  • the preset database sets a specific storage unit for each registered palmprint image, and divides the storage unit into blocks, and each block stores a block registered image.
  • search for the key data points contained therein and the feature values of the key data points and store the segmented registration image in the segment of the storage unit together with the segmented registered image.
  • the key data point is preferably the inflection point of the texture in the palmprint image
  • the characteristic value is a numerical value used to describe the inflection point, and may be a string of characters obtained by converting the pixel values of the points around the inflection point.
  • the feature value of the key data points corresponding to the registered image of the block can be directly read from the block for processing, avoiding the block and feature value extraction processing for the registered palmprint image each time to improve Processing efficiency.
  • the number of blocks for the registered palmprint image is determined by a preset number, that is, the data of the required blocks is set as the preset number according to requirements in advance, such as the number of blocks of n*n. According to the preset number, each registered palmprint image is divided into blocks in advance to obtain respective n*n block registered images, and the key data points and their characteristic values in each block registered image are extracted and stored in their respective counterparts. In the partition of the storage unit.
  • the registered palmprint image in the preset database is read, that is, the registered palmprint image is read.
  • the registered palmprint image is divided into blocks to register the feature value of the image for feature matching with the palmprint image to be processed.
  • the palmprint image to be processed is divided according to the preset number, and the block palmprint image of the preset data is obtained, so as to register the image for each block of the registered palmprint image and each block of the palmprint image to be processed The palmprint image is matched.
  • Step S22 according to the arrangement position of each of the divided palmprint images in the palmprint image to be processed, determine the registered divided image in the registered palmprint image that corresponds to each of the divided palmprint images. ;
  • the divided divided palmprint images are numbered according to the preset number, such as the aforementioned n*n number of divided palmprint images.
  • the numbers from top to bottom are [1*1], [1*2] ⁇ [2*1], [2*2] ⁇ ; each block of palmprint image is determined by each number The arrangement position in the palmprint image to be processed.
  • the registered images in the registered palmprint images are also numbered according to the preset number. In the process of matching each registered image and each palmprint image, the same between the two Numbers are matched, and the same number represents the same arrangement position to ensure the accuracy of matching.
  • each block of palmprint image in the palmprint image to be processed determine the block registration image corresponding to each block of palmprint image in the registered palmprint image, and each block of palmprint image and each block
  • the block registration images corresponding to the palmprint images have the same number, which indicates that the arrangement positions are the same.
  • Step S23 Perform block matching between the registered palmprint image and the palmprint image to be processed according to each of the divided palmprint images and the divided registered images corresponding to each of the divided palmprint images. To generate a matching data set.
  • the registered palmprint image and the palmprint image to be processed are matched in blocks according to the matching between each block of palmprint images and their respective corresponding block registration images.
  • the segmented matching between the registered palmprint image and the palmprint image to be processed is completed, and a matching data group between the two is generated.
  • a plurality of matching data sets are generated to characterize the similarity between each registered palmprint image and the palmprint image to be processed.
  • Step S231 call each of the divided palmprint images, and perform the following steps for each of the divided palmprint images:
  • each segmented palmprint image and its corresponding can be called Register the image in blocks to match.
  • Step S232 Determine whether there are key data points in the divided palmprint image. If there are key data points, determine the divided palmprint image as the first matching unit, and read the corresponding first matching unit. The first eigenvalue;
  • each called block palmprint image it is first judged whether it contains a key data point, and if there is a key data point, the called block palmprint image is used as the first matching unit. Thereafter, the feature value of the key data point contained therein is extracted as the first feature value corresponding to the first matching unit.
  • Step S233 Search for the target block registration image corresponding to the first matching unit in the block registration images respectively corresponding to the block palmprint images, and other block registration images adjacent to the target block registration image. Block registration images, and determine the target block registration image and the other block registration images together as the second matching unit;
  • search for the block registration image corresponding to each block of palmprint images determine the block registration image corresponding to the first matching unit, and use the block registration image obtained by the search as the target block registration image .
  • search for other block registration images adjacent to the target block registration image and then form the searched target block registration image and other block registration images together as a second matching unit.
  • the number of adjacent other block registration images is set according to requirements, such as setting two block registration images adjacent to the left and right, or four block registration images adjacent to the top, bottom, left and right, or eight surrounding blocks Registered images, etc.; in this embodiment, taking into account the accuracy of matching, the number is set to the surrounding eight segmented registered images.
  • the arrangement position of the divided palmprint image is represented by a number as [5*5], then the number of the corresponding target divided registration image is also [5*5], and the other divisions adjacent to the target divided registration image
  • the number of the block registration image is [4*4], [4*5], [4*6], [5*4], [5*6], [6*4], [6*5], [ 6*6], thereby forming a block palmprint image with the serial number [5*5] as the first matching unit, and will have the serial numbers [4*4], [4*5], [4*6 ], [5*4], [5*5], [5*6], [6*4], [6*5], [6*6] nine registered image blocks are formed as the second matching unit .
  • the number of adjacent registration images of other blocks varies depending on the location. .
  • the number of adjacent other block registration images is three, and the three other block registration images and the target block registration image are formed as the second matching unit; for four sides
  • the number of adjacent other block registration images is five, and the five other block registration images and the target block registration image are formed together as the second matching unit.
  • Step S234 Read a second feature value corresponding to the second matching unit, and compare the registered palmprint image and the palmprint image to be processed according to the first feature value and the second feature value. Perform block matching to generate matching data groups.
  • search for the block that stores the registered image of each block that forms the second matching unit and extract the respective feature value from the searched block, and form each feature value into a second matching unit corresponding to the second matching unit.
  • Eigenvalues According to the key data points represented by the first feature value and the second feature value, the registered palmprint image and the palmprint image to be processed are matched in blocks. After the first eigenvalues formed are matched with the second eigenvalues formed by the respective corresponding block registration images in the registered image, the block matching between the registered palmprint image and the palmprint image to be processed is completed, Generate matching data sets between the two.
  • block matching is performed on the registered palmprint image and the palmprint image to be processed, and the step of generating a matching data group includes:
  • Step b1 Calculate the first similarity value between the first characteristic value and the second characteristic value, and determine that each registered data point in the second matching unit matches the first match according to the first similarity value The first matching data pair between the key data points in the unit;
  • the server calculates the first similarity value between the first feature value and the second feature value through a preset preset method.
  • the preset method can be set as the cosine distance or as Euclidean distance.
  • the first similarity calculation between the first feature value and the second feature value is between the feature value of the key data point contained in the first matching unit and the feature value of the registered data point contained in the second matching unit Calculation of similarity.
  • the calculated similarity is used to find each registered data point most similar to each key data point of the first matching unit among the registered data points contained in the second matching unit. 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.
  • the number of key data points contained in the first matching unit may be single or multiple, and the number of registered data points contained in the second matching unit may also be single or multiple.
  • the feature value of the included key data point forms each first element of the first feature value
  • the feature value of the registered data point contained in the second matching unit forms each second element of the second feature value.
  • similarity calculations are performed on the first element and each second element one by one to obtain multiple first similarity values for each first element.
  • the multiple first similarity values are compared to determine the first similarity value with the largest numerical value, and the second element that generates the largest first similarity value has the highest similarity with the first element.
  • the registration data point corresponding to the second element and the key data point corresponding to the first element can be formed as the first matching data between the registration data point in the second matching unit and the key data point in the first matching unit Correct.
  • Each first element contained in the first feature value is based on the respective first similarity value to find the second element with the highest degree of similarity to each, and then each registered data point in the second matching unit is matched with the first A plurality of first matching data pairs between each key data point in the unit, so as to obtain each registration data point most similar to each key data point of the first matching unit among the registration data points included in the second matching unit.
  • Step b2 the first matching unit, the second matching unit, the first characteristic value and the second characteristic value are updated, and the second characteristic value between the updated first characteristic value and the second characteristic value is calculated.
  • a similarity value and according to the second similarity value, determine 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;
  • this 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.
  • the generation mode of the second matching unit is exchanged.
  • the step of updating the first matching unit, the second matching unit, the first characteristic value, and the second characteristic value includes:
  • Step b21 searching for other block palmprint images adjacent to the first matching unit, and using the first matching unit and the other block palmprint images together as a new second matching unit;
  • Step b22 updating the second characteristic value according to the characteristic value corresponding to the new second matching unit
  • the key data points contained in each block of palmprint images forming the new second matching unit are extracted, and the feature values contained in the extracted key data points are identified, and each key data obtained by the identification is
  • the feature value of the point is the feature value corresponding to the new second matching unit, and the original second feature value is replaced with the corresponding feature value to realize the update of the second feature value.
  • Step b23 Use the target block registration image as a new first matching unit, and update the first feature value according to the feature value corresponding to the new first matching unit.
  • the target segment registration image corresponding to the original first matching unit is taken as the new first matching unit, and the feature value of the segment storing the target segment registration image is extracted as the new first matching unit.
  • the feature value corresponding to the unit is replaced with the corresponding feature value to replace the original first feature value to realize the update of the first feature value.
  • the similarity between the updated first eigenvalue and the second eigenvalue is calculated in a preset manner, and the calculated result is the second similarity between the two, and the calculated similarity is used to calculate the similarity between the two.
  • the same as the first similarity is that the feature value of each registered data point is used as the first element and the second element formed by the feature value of each key data point to generate multiple second similarity values.
  • the second similarity value with the largest numerical value is determined among the two similarity values, and the second element that generates the largest second similarity value has the highest degree of similarity with the first element.
  • the key data point corresponding to the second element and the registered data point corresponding to the first element can be formed as the difference between the registered data point in the new first matching unit and the key data point in the new second matching unit.
  • the second matching data pair After the first element formed by the characteristic value of each registered data point is based on the respective second similarity value, after finding the second element with the highest degree of similarity to each, the registration data in the new first matching unit is formed A plurality of second matching data pairs between a point and each key data point in the new second matching unit.
  • Step b3 determining a target data pair between the block palmprint image and the target block registration image according to each of the first matching data pairs and each of the second matching data pairs;
  • each first matching data pair represents the key data points between the block registration image corresponding to the block palmprint image and its adjacent block registration image based on the block palmprint image Matching relationship with each registration data point; each key data point has a unique registration data point to match it.
  • Each second matching data pair represents the registration data point and each key data between the block palmprint image corresponding to the block registration image and its imminent block palmprint image based on the block registration image The matching relationship between points; each registered data point has a unique key data point to match it.
  • the match between the key data point and the registered data point is a valid match, and there is a high similarity between the two.
  • the registration data point matched by the key data point a1 is b1
  • the registration data point matched by the key data point a1 is still b1
  • the difference between a1 and b1 is determined
  • the similarity is high, and the two are effective matches. If there is a change, it means that the match between the key data point and the registered data point is an invalid match, and the similarity between the two is revealed.
  • the registration data point matched by the key data point a1 is b1
  • the registration data point matched by the key data point a1 is b2
  • the similarities with b2 are both low, and the two are invalid matches.
  • the key data points and registration data points for effective matching are determined, and the key data points and registration data points for effective matching are used as the block palmprint image and Register target data pairs between images in blocks.
  • Step b4 after each of the segmented palmprint images in the palmprint image to be processed generates the target data pair, complete the division between the registered palmprint image and the palmprint image to be processed Block matching, and generate each of the target data pairs as a matching data group.
  • each block of the palmprint image in the palmprint image to be processed matches each block of the registered palmprint image corresponding to each block of the registered palmprint image.
  • the registered palmprint image is completed. Block matching between the pattern image and the palmprint image to be processed, and each target matching pair is generated as a matching data set.
  • the registered palmprints are characterized by the similarity between each block of the palmprint image in the palmprint image to be processed represented by each target matching pair and each block of the registered image in the corresponding registered palmprint image.
  • the overall similarity between the image and the palmprint image to be processed that is, the matching data group formed by each target data pair, represents the overall similarity between the registered palmprint image and the palmprint image to be processed.
  • a block matching mechanism is set, and a smaller block and a larger block are performed during the block matching process. Matching to compensate for the displacement error in the palmprint image cropping process and avoid the corresponding matching points from falling outside the matching area.
  • the original smaller block is formed into a larger block, and the original larger block is formed into the smallest block, and the matching is performed again to avoid a single time
  • the matching error improves the accuracy of the matching, and makes the determined similarity between the registered palmprint image and the palmprint image to be processed more accurate.
  • the step of matching the data set to determine the target palmprint image matching the palmprint image to be processed in each of the registered palmprint images includes:
  • Step S31 Count the number of target data pairs included in each matching data group, and determine the target number with the largest value in each of the numbers;
  • the registered palmprint images and the palmprint to be processed represented by each matching data set can be used.
  • the overall similarity between the images is used to determine the registered palmprint image with the greatest similarity to the palmprint image to be processed. Specifically, the number of target data pairs contained in each matching data group is counted, and the more the number of target data pairs included in the matching data group, the more the number of target data pairs contained in the matching data group, the more that each segmented palmprint image in the palmprint image to be processed and the registered palmprint image.
  • the higher the similarity between each block of registered images in the palmprint image the higher the overall similarity between the registered palmprint image and the palmprint image to be processed.
  • the data is compared to determine the number of targets with the largest value, and the target data pair is used to determine the registered palmprint image with the highest similarity to the palmprint image to be processed. Palmprint image.
  • Step S32 searching for a target matching data group corresponding to the number of targets among the plurality of matching data groups, and determining a registered palmprint image corresponding to the target matching data group as the target palmprint image.
  • the number of matching data groups containing the target data pairs is the target number
  • the matching data group is regarded as the target matching data group
  • the target matching data is searched and generated
  • the registered palmprint image of the group is used as the registered palmprint image corresponding to the target matching data group.
  • the corresponding registered palmprint image is the target palmprint image that matches the palmprint image to be processed, so that the palmprint to be processed is realized Image recognition.
  • the corresponding registered palmprint image is determined and recognized as the target palmprint image matching the palmprint image to be processed. Since the matching data group containing the largest number of target data pairs represents the greatest similarity between the registered palmprint image and the palmprint image to be processed, the determined target palmprint image and the palmprint image to be processed have the largest The similarity ensures the accuracy of palmprint image recognition to be processed.
  • the present application also provides a palmprint image recognition device.
  • FIG. 3 is a schematic diagram of the functional modules of the first embodiment of the palmprint image recognition device of the present application.
  • the palmprint image recognition device includes:
  • the reading module 10 is used for reading the registered palmprint image in the preset database when the palmprint image to be processed is obtained;
  • the matching module 20 is configured to perform block matching of each of the registered palmprint images with the palmprint images to be processed to generate a plurality of matching data groups, wherein one of the registered palmprint images corresponds to a matching Data group
  • the recognition module 30 is configured to determine a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images according to a plurality of the matching data sets.
  • the reading module 10 when the palmprint image to be processed is obtained to indicate the palmprint recognition requirement, the reading module 10 first reads the registered palmprint image from the preset database, and then The matching module 20 performs block matching of each registered palmprint image with the palmprint image to be processed to obtain a plurality of matching data groups; each registered palmprint image is matched with the palmprint image to be processed in blocks, and a matching data is generated Group, which characterizes the degree of matching between each registered palmprint image and the palmprint image to be processed; and then the recognition module 30 determines the palmprint image from each registered palmprint image according to a plurality of matching data sets.
  • the most matching target palmprint image is the palmprint image to be processed for recognition of the palmprint image, and the recognition of the palmprint image to be processed is realized.
  • block matching is a mechanism that divides the palmprint image into multiple block images for matching. By matching each registered palmprint image with the palmprint image to be processed, it can avoid using the entire palmprint image. For reference matching, the amount of reference processing data in the matching process is reduced, which is conducive to improving the efficiency of palmprint image recognition.
  • each of the registered palmprint images exists in a preset database in the form of being divided into a preset number of registered images, and the matching module 20 includes:
  • the dividing unit is configured to divide the palmprint image to be processed into a preset number of divided palmprint images, and perform the following steps for each registered palmprint image:
  • the determining unit is configured to determine, according to the arrangement position of each of the divided palmprint images in the palmprint image to be processed, the divided blocks in the registered palmprint image that correspond to each of the divided palmprint images, respectively Registered image
  • the matching unit is configured to divide the registered palmprint image and the palmprint image to be processed according to each of the divided palmprint images and the divided registered images corresponding to each of the divided palmprint images. Block matching, generate matching data group.
  • the matching unit is also used for:
  • the matching unit is also used for:
  • the first matching unit, the second matching unit, the first feature value, and the second feature value are updated, and the second similarity value between the updated first feature value and the second feature value is calculated , And according to the second similarity value, determine 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;
  • the target data pair is generated for each of the block palmprint images in the palmprint image to be processed, the block matching between the registered palmprint image and the palmprint image to be processed is completed, And each of the target data pairs is generated as a matching data group.
  • the matching unit is also used for:
  • the target segmented registration image is used as a new first matching unit, and the first feature value is updated according to the feature value corresponding to the new first matching unit.
  • identification module 30 further includes:
  • a statistical unit configured to count the number of target data pairs included in each of the matching data groups, and determine the target number with the largest value in each of the numbers
  • the searching unit is configured to search for target matching data sets corresponding to the number of targets among the plurality of matching data sets, and determine a registered palmprint image corresponding to the target matching data set as the target palmprint image.
  • the palmprint image recognition device further includes:
  • the determining module is configured to, when the palm print image is received, perform left and right hand recognition on the palm print image according to a preset network model, and determine the left and right hand attributes of the palm print image;
  • the recognition module is further configured to recognize the cropping points of the lines on the palmprint image according to the attributes of the left and right hands;
  • the generating module is configured to cut the palmprint image according to the texture cutting point to generate the palmprint image to be processed.
  • the embodiment of the present application also proposes a computer-readable storage medium.
  • the computer-readable storage medium may be volatile or non-volatile.
  • a palmprint image recognition program is stored on the computer-readable storage medium, and the palmprint image recognition program is executed by the processor to implement the following steps:
  • a target palmprint image matching the palmprint image to be processed in each of the registered palmprint images is determined.
  • the palmprint image recognition method provided in the present application further ensures the privacy and security of all the above-mentioned data that appears, all the above-mentioned data can also be stored in a node of a blockchain.
  • all the above-mentioned data can also be stored in a node of a blockchain.
  • palmprint images to be processed and target palmprint images, etc. these data can be stored in the blockchain node.
  • Blockchain essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information for verification. The validity of the information (anti-counterfeiting) and the generation of the next block.
  • the blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.
  • the technical solution of this application essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product is stored in a computer-readable storage medium as described above (such as The ROM/RAM, magnetic disk, optical disk) includes several instructions to make a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) execute the methods described in the various embodiments of the present application.
  • a terminal device which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Human Computer Interaction (AREA)
  • Multimedia (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Collating Specific Patterns (AREA)

Abstract

一种掌纹图像的识别方法、装置、设备及计算机可读存储介质,涉及人工智能领域。所述方法包括:当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像(S10);将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组(S20);根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像(S30)。该方法通过图像处理技术将各个已注册掌纹图像分别和待处理掌纹图像进行分块匹配,避免以整张掌纹图像为参照进行匹配,减少了匹配过程中参照处理的数据量,有利于掌纹图像识别效率的提高。

Description

掌纹图像的识别方法、装置、设备及计算机可读存储介质
本申请要求于2020年5月20日提交中国专利局、申请号为CN202010433723.6,发明名称为“掌纹图像的识别方法、装置、设备及计算机可读存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及图像处理技术领域,尤其涉及一种掌纹图像的识别方法、装置、设备及计算机可读存储介质。
背景技术
科技的发展,使得通过生物特征所进行的身份验证得到了广泛的应用。掌纹作为一种相对稳定的生物特征,能有效识别出人物的身份,因而可大量应用于无人超市、职场考勤等各种需要进行人物身份识别的应用场景。传统的掌纹识别一般借助固定设备扫描手掌图像实现,属于手掌接触式的识别;长期的接触容易导致设备污染,使所扫描的手掌图像不清晰,进而影响识别的效果。
随着人工智能的发展,出现了通过对数码相机或摄像头所拍摄掌纹图像的识别来进行掌纹识别的技术,以此避免传统识别方式中因扫描的手掌图像不清晰而影响识别效果的问题。发明人意识到当前基于掌纹图像识别所实现的掌纹识别,主要通过提取掌纹图像中的特征进行匹配的方式实现。但是,该方式为了识别的准确性往往会从掌纹图像中提取大量特征,每个特征均以整张掌纹图像为参照进行匹配,如此一来,匹配所需要参照处理的特征数据众多,导致了掌纹图像识别的效率低。
发明内容
一种掌纹图像的识别方法,所述掌纹图像的识别方法包括以下步骤:
当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
一种掌纹图像的识别装置,所述掌纹图像的识别装置包括:
读取模块,用于当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
匹配模块,用于将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
识别模块,用于根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
一种掌纹图像的识别设备,所述掌纹图像的识别设备包括存储器、处理器以及存储在所述存储器上并可在所述处理器上运行的掌纹图像的识别程序,所述掌纹图像的识别程序被所述处理器执行时实现如下步骤:
当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
一种计算机可读存储介质,所述计算机可读存储介质上存储有掌纹图像的识别程序, 所述掌纹图像的识别程序被处理器执行时实现如下步骤:
当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
本申请可避免以整张掌纹图像为参照进行匹配,减少了匹配过程中参照处理的数据量,有利于掌纹图像识别效率的提高。
附图说明
图1为本申请实施例方案涉及的硬件运行环境的掌纹图像的识别设备结构示意图;
图2为本申请掌纹图像的识别方法第一实施例的流程示意图;
图3为本申请掌纹图像的识别装置较佳实施例的功能模块示意图。
本申请目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
应当理解,此处所描述的具体实施例仅仅用以解释本申请,并不用于限定本申请。
如图1所示,图1是本申请实施例方案涉及的硬件运行环境的掌纹图像的识别设备结构示意图。
在后续的描述中,使用用于表示元件的诸如“模块”、“部件”或“单元”的后缀仅为了有利于本申请的说明,其本身没有特定的意义。因此,“模块”、“部件”或“单元”可以混合地使用。
本申请实施例掌纹图像的识别设备可以是PC,也可以是平板电脑、便携计算机等可移动式终端设备。
如图1所示,该掌纹图像的识别设备可以包括:处理器1001,例如CPU,网络接口1004,用户接口1003,存储器1005,通信总线1002。其中,通信总线1002用于实现这些组件之间的连接通信。用户接口1003可以包括显示屏(Display)、输入单元比如键盘(Keyboard),可选用户接口1003还可以包括标准的有线接口、无线接口。网络接口1004可选的可以包括标准的有线接口、无线接口(如WI-FI接口)。存储器1005可以是高速RAM存储器,也可以是稳定的存储器(non-volatile memory),例如磁盘存储器。存储器1005可选的还可以是独立于前述处理器1001的存储装置。
本领域技术人员可以理解,图1中示出的掌纹图像的识别设备结构并不构成对掌纹图像的识别设备的限定,可以包括比图示更多或更少的部件,或者组合某些部件,或者不同的部件布置。
如图1所示,作为一种计算机可读存储介质的存储器1005中可以包括操作系统、网络通信模块、用户接口模块以及检测程序。
在图1所示的设备中,网络接口1004主要用于连接后台服务器,与后台服务器进行数据通信;用户接口1003主要用于连接客户端(用户端),与客户端进行数据通信;而处理器1001可以用于调用存储器1005中存储的检测程序,并执行以下操作:
当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
进一步地,每一所述已注册掌纹图像以划分为预设数量分块注册图像的形式存在于预 设数据库中;
所述将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组的步骤包括:
将所述待处理掌纹图像划分为预设数量的分块掌纹图像,并针对每一所述已注册掌纹图像均执行以下步骤:
根据各所述分块掌纹图像在所述待处理掌纹图像中的排列位置,确定所述已注册掌纹图像中与各所述分块掌纹图像分别对应的分块注册图像;
根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
进一步地,所述根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
调用各所述分块掌纹图像,并针对每一所述分块掌纹图像执行以下步骤:
判断所述分块掌纹图像中是否存在关键数据点,若存在关键数据点,则将所述分块掌纹图像确定为第一匹配单位,并读取所述第一匹配单位对应的第一特征值;
查找与各所述分块掌纹图像分别对应的分块注册图像中与所述第一匹配单位对应的目标分块注册图像,以及与所述目标分块注册图像相邻的其他分块注册图像,并将所述目标分块注册图像和所述其他分块注册图像一并确定为第二匹配单位;
读取与所述第二匹配单位对应的第二特征值,根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
进一步地,所述根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
计算所述第一特征值和所述第二特征值之间的第一相似度值,并根据所述第一相似度值,确定第二匹配单位中各注册数据点与第一匹配单位中各所述关键数据点之间的第一匹配数据对;
对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新,计算更新后的所述第一特征值和所述第二特征值之间的第二相似度值,并根据所述第二相似度值,确定更新后第一匹配单位中各注册数据点与更新后第二匹配单位中各关键数据点之间的第二匹配数据对;
根据各所述第一匹配数据对和各所述第二匹配数据对,确定所述分块掌纹图像和所述目标分块注册图像之间的目标数据对;
在所述待处理掌纹图像中的各所述分块掌纹图像均生成所述目标数据对之后,完成所述已注册掌纹图像和所述待处理掌纹图像之间的分块匹配,并将各所述目标数据对生成为匹配数据组。
进一步地,所述对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新的步骤包括:
查找与第一匹配单位相邻的其他分块掌纹图像,并将所述第一匹配单位和所述其他分块掌纹图像一并作为新的第二匹配单位;
根据与新的所述第二匹配单位对应的特征值,对所述第二特征值进行更新;
将所述目标分块注册图像作为新的第一匹配单位,并根据与新的所述第一匹配单位对应的特征值,对所述第一特征值进行更新。
进一步地,所述根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像的步骤包括:
统计每一所述匹配数据组中包含目标数据对的数量,并确定各所述数量中数值最大的目标数量;
查找多个所述匹配数据组中与所述目标数量对应的目标匹配数据组,并将与目标匹配数据组对应的已注册掌纹图像确定为所述目标掌纹图像。
进一步地,所述当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像的步骤之前,处理器1001可以用于调用存储器1005中存储的检测程序,并执行以下操作:
当接收到掌纹图像时,根据预设网络模型,对所述掌纹图像进行左右手识别,确定所述掌纹图像的左右手属性;
根据所述左右手属性,识别所述掌纹图像上的纹路裁剪点;
根据所述纹路裁剪点,对所述掌纹图像进行裁剪,生成待处理掌纹图像。
本申请掌纹图像的识别设备的具体实施方式与下述掌纹图像的识别方法各实施例基本相同,在此不再赘述。
为了更好的理解上述技术方案,下面将参照附图更详细地描述本公开的示例性实施例。虽然附图中显示了本公开的示例性实施例,然而应当理解,可以以各种形式实现本公开而不应被这里阐述的实施例所限制。相反,提供这些实施例是为了能够更透彻地理解本公开,并且能够将本公开的范围完整的传达给本领域的技术人员。
为了更好的理解上述技术方案,下面将结合说明书附图以及具体的实施方式对上述技术方案进行详细的说明。
参照图2,本申请第一实施例提供一种掌纹图像的识别方法的流程示意图。该实施例中,所述掌纹图像的识别方法包括以下步骤:
步骤S10,当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
本实施例中掌纹图像的识别方法应用于服务器,适用于通过服务器来对掌纹图像进行识别。其中,服务器与智能手机、平板电脑、摄像机、数码相机等具有摄像功能的终端连接,由该类智能终端将拍摄的掌纹图像上传到服务器进行识别处理。
可理解地,手掌分为左手掌和右手掌,不同的手掌纹路具有不同的特性,为了针对性的识别掌纹图像,需要对掌纹图像进行左右手属性识别,即确定所拍摄的掌纹图像来自于左手掌还是右手掌。此外,考虑到所拍摄的掌纹图像中可能包含有其他背景物体而影响识别效果,故而在识别处理前,还需要确定出掌纹图像中的有效区域,通过对有效区域的识别处理,来提高识别处理的准确性,并减少识别处理的数据量。将该左右手属性识别和有效区域的确定,设置为预处理机制,由服务器对掌纹图像进行识别处理前进行。具体地,当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像的步骤之前包括:
步骤a1,当接收到掌纹图像时,根据预设网络模型,对所述掌纹图像进行左右手识别,确定所述掌纹图像的左右手属性;
进一步地,服务器预先训练有预设网络模型,该预设网络模型经由大量左手掌图像样本和右手掌图像样本训练,以对左右手属性进行识别;同时,预设网络模型还经由大量的左手掌纹路样本和右手掌纹路样本进行训练,以对掌纹图像中的有效区域进行识别。服务器在接收到上传的掌纹图像后,调用预设网络模型,由预设网络模型对掌纹图像进行左右手识别,确定掌纹图像与左手掌属性一致还是与右手掌属性一致;以此,确定掌纹图像的左右手属性。
步骤a2,根据所述左右手属性,识别所述掌纹图像上的纹路裁剪点;
更进一步地,在确定掌纹图像的左右手属性后,则依据预设网络模型中与左右手属性对应的纹路样本,识别掌纹图像中的纹路裁剪点。其中纹路裁剪点优选为手掌四根手指中第一个关节处的纹路,通过预设网络模型对掌纹图像进行识别,得到其中四根手指第一个关节处的纹路作为纹路裁剪点。
步骤a3,根据所述纹路裁剪点,对所述掌纹图像进行裁剪,生成待处理掌纹图像。
进一步地,依据纹路裁剪点对掌纹图像进行裁剪,将纹路裁剪点中的食指第一个关节处的纹路和小拇指第一个关节处的纹路作为两个连接点进行连接,形成裁剪的连接线。进 而依据连接线对掌纹图像进行裁剪,即得到有效区域,将有效区域作为待处理掌纹图像,以供服务器进行识别处理。
更进一步地,服务器对接有用于存储注册时所录入掌纹图像的预设数据库,在生成待处理掌纹图像之后,将注册时所录入的掌纹图像作为预设数据中的已注册掌纹图像进行读取。通过各已注册掌纹图像和待处理掌纹图像之间的相似性,来对待处理掌纹图像进行识别。
步骤S20,将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
更进一步地,服务器针对掌纹图像之间相似性的确定预先设置有分块匹配机制,分块匹配即为将掌纹图像划分为多个分块图像进行匹配的机制。在生成待处理掌纹图像,并读取到已注册掌纹图像后,即可将各项已注册掌纹图像分别和待处理掌纹图像进行分块匹配。将待处理掌纹图像划分为多个分块图像,并且将各已注册掌纹图像划分为多个分块图像,再两类分块图像之间进行匹配,由各个分块图像之间的匹配度来确定待处理掌纹图像和各项已注册掌纹图像之间的相似性。其中,因已注册掌纹图像的数量众多,在将待处理掌纹图像的各分块图像和已注册掌纹图像的各分块图像进行匹配的过程中,可逐一将已注册掌纹图像和待处理掌纹图像进行匹配,也可同时将各已注册掌纹图像分别和待处理掌纹图像进行匹配,对此不做限制。每个已注册掌纹图像和待处理掌纹图像进行匹配,均生成一个匹配数据组,表征已注册掌纹图像中各个掌纹特征点与待处理掌纹图像中各个掌纹特征点之间的匹配程度。
步骤S30,根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
进一步地,因匹配数据组表征了已注册掌纹图像与待处理掌纹图像之间各个掌纹特征点的匹配程度,在各个已注册掌纹图像和待处理掌纹图像均进行分块匹配,生成各自的匹配数据组后,可对比查找各个匹配数据组中表征匹配程度最高的目标匹配数据组,生成该目标匹配数据组的已注册掌纹图像与待处理掌纹图像之间的掌纹特征点具有最高的相似性,故而将其确定为已注册掌纹图像中与待处理掌纹图像匹配的目标掌纹图像,将待处理掌纹图像识别为目标掌纹图像,完成对待处理掌纹图像的识别。
本实施例的掌纹图像的识别方法,在获取到待处理掌纹图像,表征具有掌纹识别需求时,先从预设数据库中读取出已注册掌纹图像,并将各个已注册掌纹图像分别和待处理掌纹图像进行分块匹配,得到多个匹配数据组;每一已注册掌纹图像与待处理掌纹图像分块匹配,生成一个匹配数据组,表征每一已注册掌纹图像与待处理掌纹图像之间的匹配程度;进而依据多个匹配数据组,从各个以注册掌纹图像中确定出与待处理掌纹图像最为匹配的目标掌纹图像,该目标掌纹图像即为对待处理掌纹图像识别的掌纹图像,实现待处理掌纹图像的识别。其中,分块匹配为将掌纹图像划分为多个分块图像进行匹配的机制,通过将各个已注册掌纹图像分别和待处理掌纹图像进行分块匹配,可避免以整张掌纹图像为参照进行匹配,减少了匹配过程中参照处理的数据量,有利于掌纹图像识别效率的提高。
进一步的,基于本申请掌纹图像的识别方法第一实施例,提出本申请掌纹图像的识别方法第二实施例,在第二实施例中,每一所述已注册掌纹图像以划分为预设数量分块注册图像的形式存在于预设数据库中,所述将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组的步骤包括:
步骤S21,将所述待处理掌纹图像划分为预设数量的分块掌纹图像,并针对每一所述已注册掌纹图像均执行以下步骤:
本实施例中,将各已注册掌纹图像的分块图像作为各已注册掌纹图像的分块注册图像,以分块的形式存储于预设数据库中。预设数据库针对每一已注册掌纹图像设定特定的存储单元,并将存储单元进行分块,每一分块存储一项分块注册图像。同时对于每一分块注册 图像,查找其中包含的关键数据点以及关键数据点的特征值,和分块注册图像一并存储至存储单元的分块中。关键数据点优选为掌纹图像中的纹路拐点,特征值为用于描述拐点的数值,可以是拐点周围各点的像素值所转换到得到的一串字符等。在每次匹配过程中,可直接从分块中读取分块注册图像对应关键数据点的特征值进行处理,避免每次均针对已注册掌纹图像进行分块和特征值提取处理,以提高处理效率。其中,对于已注册掌纹图像的分块数量通过预设数量确定,即预先依据需求设定所需要分块的数据作为预设数量,如n*n的分块数量。根据预设数量,预先对各已注册掌纹图像进行分块处理,得到各自的n*n个分块注册图像,并提取各分块注册图像中的关键数据点及其特征值存储在各自对应存储单元的分块中。
进一步地,在具有对掌纹图像的识别需求,并将掌纹图像经预处理生成为待处理掌纹图像后,对预设数据库中的已注册掌纹图像进行读取,即读取各已注册掌纹图像各自分块注册图像的特征值,以用于和待处理掌纹图像进行特征匹配。并且,按照预设数量对待处理掌纹图像进行划分,得到预设数据的分块掌纹图像,以针对每一已注册掌纹图像的各分块注册图像和待处理掌纹图像的各分块掌纹图像进行匹配处理。
步骤S22,根据各所述分块掌纹图像在所述待处理掌纹图像中的排列位置,确定所述已注册掌纹图像中与各所述分块掌纹图像分别对应的分块注册图像;
更进一步地,在将待处理掌纹图像划分为预设数量的分块掌纹图像的过程中,根据预设数量对划分的分块掌纹图像进行编号,如上述n*n个分块数量,则编号从上至下分别为[1*1]、[1*2]···[2*1]、[2*2]···;通过各个编号来确定各分块掌纹图像在待处理掌纹图像中的排列位置。同样的对于已注册掌纹图像中的各分块注册图像也依据预设数量进行编号,在将各分块注册图像和各分块掌纹图像进行匹配的过程中,依据两者之间同样的编号进行匹配处理,同样的编号表征同样的排列位置,以确保匹配的准确性。依据各分块掌纹图像在待处理掌纹图像中的排列位置,确定已注册掌纹图像中与各个分块掌纹图像对应的分块注册图像,各分块掌纹图像以及与各分块掌纹图像对应的分块注册图像之间具有相同的编号,表征排列位置相同。
步骤S23,根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
进一步地,依据各分块掌纹图像及其各自对应的分块注册图像之间的匹配性,来对已注册掌纹图像和待处理掌纹图像进行分块匹配。在各个分块掌纹图像均和各自对应的分块注册图像进行匹配之后,完成已注册掌纹图像和待处理掌纹图像之间的分块匹配,生成两者之间的匹配数据组。并且,在各个已注册掌纹图像均和待处理掌纹图像进行分块匹配后,则生成多个匹配数据组,表征各已注册掌纹图像与待处理掌纹图像之间的相似性高低。其中,根据各分块掌纹图像以及与各分块掌纹图像分别对应的分块注册图像,将已注册掌纹图像和待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
步骤S231,调用各所述分块掌纹图像,并针对每一所述分块掌纹图像执行以下步骤:
进一步地,在将待处理掌纹图像划分为各分块掌纹图像,并确定与各掌纹图像具有相同位置的各分块注册图像后,即可调用每一分块掌纹图像和其对应的分块注册图像进行匹配。
步骤S232,判断所述分块掌纹图像中是否存在关键数据点,若存在关键数据点,则将所述分块掌纹图像确定为第一匹配单位,并读取所述第一匹配单位对应的第一特征值;
更进一步地,针对每一调用的分块掌纹图像先判断其中是否包含有关键数据点,若存在关键数据点,则将调用的分块掌纹图像作为第一匹配单位。此后,提取其中所包含关键数据点的特征值,作为与第一匹配单位对应的第一特征值。
步骤S233,查找与各所述分块掌纹图像分别对应的分块注册图像中与所述第一匹配单位对应的目标分块注册图像,以及与所述目标分块注册图像相邻的其他分块注册图像, 并将所述目标分块注册图像和所述其他分块注册图像一并确定为第二匹配单位;
进一步地,对与各分块掌纹图像分别对应的分块注册图像进行查找,确定第一匹配单位在其中所对应的分块注册图像,并将查找得到分块注册图像作为目标分块注册图像。同时,查找与目标分块注册图像相邻的其他分块注册图像,进而将查找得到的目标分块注册图像和其他分块注册图像一并形成为第二匹配单位。其中,相邻的其他分块注册图像数量依据需求设定,如设定左右相邻的两个分块注册图像,或者上下左右相邻的四个分块注册图像,或者周围的八个分块注册图像等;本实施例考虑到匹配的准确性,将数量设定为周围的八个分块注册图像。如对于分块掌纹图像的排列位置通过编号表征为[5*5],则其对应目标分块注册图像的编号也为[5*5],与该目标分块注册图像相邻的其他分块注册图像的编号则为[4*4]、[4*5]、[4*6]、[5*4]、[5*6]、[6*4]、[6*5]、[6*6],由此将具有编号[5*5]的一个分块掌纹图像形成为第一匹配单位,而将分别具有编号[4*4]、[4*5]、[4*6]、[5*4]、[5*5]、[5*6]、[6*4]、[6*5]、[6*6]的九个分块注册图像形成为第二匹配单位。
需要说明的是,对于位于掌纹图像四角顶点位置的分块掌纹图像,以及位于掌纹图像四边位置的分块掌纹图像,则相邻的其他分块注册图像数量依据位置的不同而不同。对于四角顶点位置的分块掌纹图像,相邻的其他分块注册图像数量为3个,将该3个其他分块注册图像和目标分块注册图像一并形成为第二匹配单位;对于四边位置的分块掌纹图像,相邻的其他分块注册图像数量为5个,将该5个其他分块注册图像和目标分块注册图像一并形成为第二匹配单位。
步骤S234,读取与所述第二匹配单位对应的第二特征值,根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
进一步地,查找对形成第二匹配单位的各分块注册图像进行存储的分块,并从查找得到的分块中提取各自的特征值,将各特征值形成与第二匹配单位对应的第二特征值。进而依据第一特征值和第二特征值各自所表征的关键数据点,对已注册掌纹图像和待处理掌纹图像进行分块匹配,在待处理掌纹图像的各个分块掌纹图像所形成的第一特征值,均和已注册图像中各自对应分块注册图像所形成的第二特征值进行匹配之后,则完成已注册掌纹图像和待处理掌纹图像之间的分块匹配,生成两者之间的匹配数据组。
其中,根据第一特征值和第二特征值,对已注册掌纹图像和待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
步骤b1,计算所述第一特征值和所述第二特征值之间的第一相似度值,并根据所述第一相似度值,确定第二匹配单位中各注册数据点与第一匹配单位中各所述关键数据点之间的第一匹配数据对;
更进一步地,服务器通过预先设定的预设方式计算第一特征值和第二特征值之间的第一相似度值,该预设方式可以依据需求设定为余弦距离、也可以设定为欧氏距离。其中,第一特征值和第二特征值之间的第一相似度计算,为第一匹配单位中所包含关键数据点的特征值和第二匹配单位中所包含注册数据点的特征值之间的相似度计算。通过计算的相似度来查找第二匹配单位包含的各注册数据点中与第一匹配单位的各关键数据点最为相似的各个注册数据点。需要说明的是,第二匹配单位中所包含的注册数据点即为第二匹配单位中的关键数据点。
可理解地,第一匹配单位中包含的关键数据点数量可能为单个也可能为多个,第二匹配单位中包含的注册数据点数量也可能为单个也可能为多个,第一匹配单位中所包含关键数据点的特征值形成第一特征值的各第一元素,第二匹配单位中所包含注册数据点的特征值形成第二特征值的各第二元素。在计算时,以第一特征值中所包含的第一元素为基础,逐一将第一元素和各项第二元素进行相似度计算,得到每个第一元素的多个第一相似度值。进而在多个第一相似度值之间对比,确定其中数值最大的第一相似度值,生成该最大第一 相似度值的第二元素与第一元素之间的相似程度最高。因此,可将该第二元素对应的注册数据点和该第一元素对应的关键数据点,形成为第二匹配单位中注册数据点与第一匹配单位中关键数据点之间的第一匹配数据对。在第一特征值中所包含的各第一元素均依据各自的第一相似度值,查找得到与各自相似程度最高的第二元素,则形成第二匹配单位中各注册数据点与第一匹配单位中各关键数据点之间的多个第一匹配数据对,以此得到第二匹配单位包含的各注册数据点中与第一匹配单位的各关键数据点最为相似的各个注册数据点。
步骤b2,对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新,计算更新后的所述第一特征值和所述第二特征值之间的第二相似度值,并根据所述第二相似度值,确定更新后第一匹配单位中各注册数据点与更新后第二匹配单位中各关键数据点之间的第二匹配数据对;
进一步地,为了提高匹配的准确性,本实施例设置有对第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新的机制,其中,更新通过对第一匹配单位和第二匹配单位的生成方式进行互换实现。具体地,对第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新的步骤包括:
步骤b21,查找与第一匹配单位相邻的其他分块掌纹图像,并将所述第一匹配单位和所述其他分块掌纹图像一并作为新的第二匹配单位;
对与第一匹配单位相邻的其他分块掌纹图像进行查找,并将第一匹配单位和查找得到的其他分块注册掌纹图像一并形成为新的第二匹配单位,以通过更新第二匹配单位来更新第二特征值。
步骤b22,根据与新的所述第二匹配单位对应的特征值,对所述第二特征值进行更新;
更进一步地,对形成新的第二匹配单位的各分块掌纹图像中包含的关键数据点进行提取,并识别所提取的各关键数据点中包含的特征值,该识别得到的各关键数据点的特征值值即为与新的第二匹配单位对应的特征值,用该对应的特征值对原始的第二特征值进行替换,实现对第二特征值的更新。
步骤b23,将所述目标分块注册图像作为新的第一匹配单位,并根据与新的所述第一匹配单位对应的特征值,对所述第一特征值进行更新。
进一步地,将与原始的第一匹配单位对应的目标分块注册图像作为新的第一匹配单位,并从存储该目标分块注册图像的分块中提取其特征值作为与新的第一匹配单位对应的特征值,用该对应的特征值对原始的第一特征值进行替换,实现对第一特征值的更新。
更进一步地,通过预设方式对更新后第一特征值和第二特征值之间相似度进行计算,所得到的计算结果即为两者之间的第二相似度,通过计算的相似度来查找新的第二匹配单位包含的各关键数据点中与新的第一匹配单位的各注册数据点作为相似的各关键数据点。与第一相似度相同的是,将每一注册数据点的特征值作为第一元素与由各关键数据点的特征值形成的第二元素,生成多个第二相似度值,从多个第二相似度值中确定数值最大的第二相似度值,生成该最大第二相似度值的第二元素与该第一元素之间的相似程度最高。因此,可将该第二元素对应的关键数据点和该第一元素对应的注册数据点,形成为新的第一匹配单位中注册数据点与新的第二匹配单位中关键数据点之间的第二匹配数据对。在由各注册数据点的特征值所形成的第一元素均依据各自的第二相似度值,查找得到与各自相似程度最高的第二元素后,则形成新的第一匹配单位中各注册数据点与新的第二匹配单位中各关键数据点之间的多个第二匹配数据对。
步骤b3,根据各所述第一匹配数据对和各所述第二匹配数据对,确定所述分块掌纹图像和所述目标分块注册图像之间的目标数据对;
可理解地,各第一匹配数据对,表征了以分块掌纹图像为基础,与该分块掌纹图像对应的分块注册图像及其相邻分块注册图像之间,各关键数据点与各注册数据点之间的匹配关系;各关键数据点均具有唯一的注册数据点与之匹配。各第二匹配数据对,表征了以分 块注册图像为基础,与该分块注册图像对应的分块掌纹图像及其临期分块掌纹图像之间,各注册数据点与各关键数据点之间的匹配关系;各注册数据点均具有唯一的关键数据点与之匹配。通过各第一匹配数据对和各第二匹配数据对,可判定以不同方式所生成的关键数据点与注册数据点之间的匹配关系是否发生变化。若未发生变化,则说明关键数据点与注册数据点之间的匹配为有效匹配,两者之间具有较高的相似性。如在第一匹配数据对中,关键数据点a1匹配的注册数据点为b1,在第二匹配数据对中,关键数据点a1匹配的注册数据点仍然为b1,则判定a1与b1之间的相似性较高,两者为有效匹配。若发生变化,则说明关键数据点与注册数据点之间的匹配为无效匹配,两者之间的相似性交底。如在第一匹配数据对中,关键数据点a1匹配的注册数据点为b1,在第二匹配数据对中,关键数据点a1匹配的注册数据点为b2,则判定a1与b1之间,a1与b2之间的相似性均较低,两者为无效匹配。
进一步地,根据各第一匹配数据对和各第二匹配数据对,确定有效匹配的关键数据点和注册数据点,并对各有效匹配的关键数据点和注册数据点作为分块掌纹图像和分块注册图像之间的目标数据对。
步骤b4,在所述待处理掌纹图像中的各所述分块掌纹图像均生成所述目标数据对之后,完成所述已注册掌纹图像和所述待处理掌纹图像之间的分块匹配,并将各所述目标数据对生成为匹配数据组。
更进一步地,在待处理掌纹图像中的各分块掌纹图像均和已注册掌纹图像中对应的各分块注册图像分块匹配,生成各自的目标数据对后,则完成已注册掌纹图像和待处理掌纹图像之间的分块匹配,并且将各个目标匹配对生成为匹配数据组。以此,通过各个目标匹配对表征的待处理掌纹图像中各分块掌纹图像与对应已注册掌纹图像中各分块注册图像之间分块图像的相似性,来表征已注册掌纹图像和待处理掌纹图像之间整体相似性,即由各目标数据对所形成的匹配数据组,来表征已注册掌纹图像和待处理掌纹图像之间的整体相似性。
本实施例在将已注册掌纹图像和待处理掌纹图像进行匹配的过程中,设置分块匹配机制,并且在分块匹配过程中将一个较小的分块和一个较大的分块进行匹配,以弥补掌纹图像裁剪过程中的位移误差,避免相应匹配点落在匹配区域之外。同时在将较小分块和较大分块进行匹配之后,将原先较小的分块形成为较大分块,而将原先较大的分块形成为最小的分块,再次进行匹配,避免单次匹配误差,提高了匹配的准确性,使得所确定的已注册掌纹图像和待处理掌纹图像之间的相似性更为准确。
进一步的,基于本申请掌纹图像的识别方法第一实施例或第二实施例,提出本申请掌纹图像的识别方法第三实施例,在第三实施例中,所述根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像的步骤包括:
步骤S31,统计每一所述匹配数据组中包含目标数据对的数量,并确定各所述数量中数值最大的目标数量;
本实施例在得到每个已注册掌纹图像与待处理掌纹图像之间的多个匹配数据组之后,则可依据各匹配数据组所表征的各项已注册掌纹图像和待处理掌纹图像之间的整体相似性,来确定与待处理掌纹图像相似性最大的已注册掌纹图像。具体地,对每个匹配数据组中所包含的目标数据对的数量进行统计,匹配数据组包含目标数据对的数量越多,则说明待处理掌纹图像中各分块掌纹图像与已注册掌纹图像中各分块注册图像之间的相似性越高,该已注册掌纹图像和待处理掌纹图像之间的整体相似性越高。从而在统计得到每个匹配数据组中包含目标数据对的数量后,对各个数据进行对比,确定其中数值最大的目标数量,由目标数据对来确定与待处理掌纹图像相似性最高的已注册掌纹图像。
步骤S32,查找多个所述匹配数据组中与所述目标数量对应的目标匹配数据组,并将与目标匹配数据组对应的已注册掌纹图像确定为所述目标掌纹图像。
进一步地,依据目标数量,对多个匹配数据组进行查找,查找其中包含目标数据对的数量为目标数量的匹配数据组,将该匹配数据组作为目标匹配数据组,进而查找生成该目标匹配数据组的已注册掌纹图像,作为与目标匹配数据组对应的已注册掌纹图像,该对应的已注册掌纹图像即为与待处理掌纹图像匹配的目标掌纹图像,实现对待处理掌纹图像的识别。
本实施例通过包含目标数据对数量最多的匹配数据组,确定对应的已注册掌纹图像,并将其识别为与待处理掌纹图像匹配的目标掌纹图像。因包含目标数据对数量最多的匹配数据组表征了已注册掌纹图像和待处理掌纹图像之间最大的相似性,故而使得所确定的目标掌纹图像与待处理掌纹图像之间具有最大的相似性,确保了待处理掌纹图像识别的准确性。
进一步地,本申请还提供一种掌纹图像的识别装置。
参照图3,图3为本申请掌纹图像的识别装置第一实施例的功能模块示意图。所述掌纹图像的识别装置包括:
读取模块10,用于当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
匹配模块20,用于将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
识别模块30,用于根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
本实施例的掌纹图像的识别装置,在获取到待处理掌纹图像,表征具有掌纹识别需求时,先由读取模块10从预设数据库中读取出已注册掌纹图像,并由匹配模块20将各个已注册掌纹图像分别和待处理掌纹图像进行分块匹配,得到多个匹配数据组;每一已注册掌纹图像与待处理掌纹图像分块匹配,生成一个匹配数据组,表征每一已注册掌纹图像与待处理掌纹图像之间的匹配程度;进而由识别模块30依据多个匹配数据组,从各个以注册掌纹图像中确定出与待处理掌纹图像最为匹配的目标掌纹图像,该目标掌纹图像即为对待处理掌纹图像识别的掌纹图像,实现待处理掌纹图像的识别。其中,分块匹配为将掌纹图像划分为多个分块图像进行匹配的机制,通过将各个已注册掌纹图像分别和待处理掌纹图像进行分块匹配,可避免以整张掌纹图像为参照进行匹配,减少了匹配过程中参照处理的数据量,有利于掌纹图像识别效率的提高。
进一步地,每一所述已注册掌纹图像以划分为预设数量分块注册图像的形式存在于预设数据库中,所述匹配模块20包括:
划分单元,用于将所述待处理掌纹图像划分为预设数量的分块掌纹图像,并针对每一所述已注册掌纹图像均执行以下步骤:
确定单元,用于根据各所述分块掌纹图像在所述待处理掌纹图像中的排列位置,确定所述已注册掌纹图像中与各所述分块掌纹图像分别对应的分块注册图像;
匹配单元,用于根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
进一步地,所述匹配单元还用于:
调用各所述分块掌纹图像,并针对每一所述分块掌纹图像执行以下步骤:
判断所述分块掌纹图像中是否存在关键数据点,若存在关键数据点,则将所述分块掌纹图像确定为第一匹配单位,并读取所述第一匹配单位对应的第一特征值;
查找与各所述分块掌纹图像分别对应的分块注册图像中与所述第一匹配单位对应的目标分块注册图像,以及与所述目标分块注册图像相邻的其他分块注册图像,并将所述目标分块注册图像和所述其他分块注册图像一并确定为第二匹配单位;
读取与所述第二匹配单位对应的第二特征值,根据所述第一特征值和所述第二特征值, 对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
进一步地,所述匹配单元还用于:
计算所述第一特征值和所述第二特征值之间的第一相似度值,并根据所述第一相似度值,确定第二匹配单位中各注册数据点与第一匹配单位中各所述关键数据点之间的第一匹配数据对;
对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新,计算更新后的所述第一特征值和所述第二特征值之间的第二相似度值,并根据所述第二相似度值,确定更新后第一匹配单位中各注册数据点与更新后第二匹配单位中各关键数据点之间的第二匹配数据对;
根据各所述第一匹配数据对和各所述第二匹配数据对,确定所述分块掌纹图像和所述目标分块注册图像之间的目标数据对;
在所述待处理掌纹图像中的各所述分块掌纹图像均生成所述目标数据对之后,完成所述已注册掌纹图像和所述待处理掌纹图像之间的分块匹配,并将各所述目标数据对生成为匹配数据组。
进一步地,所述匹配单元还用于:
查找与第一匹配单位相邻的其他分块掌纹图像,并将所述第一匹配单位和所述其他分块掌纹图像一并作为新的第二匹配单位;
根据与新的所述第二匹配单位对应的特征值,对所述第二特征值进行更新;
将所述目标分块注册图像作为新的第一匹配单位,并根据与新的所述第一匹配单位对应的特征值,对所述第一特征值进行更新。
进一步地,所述识别模块30还包括:
统计单元,用于统计每一所述匹配数据组中包含目标数据对的数量,并确定各所述数量中数值最大的目标数量;
查找单元,用于查找多个所述匹配数据组中与所述目标数量对应的目标匹配数据组,并将与目标匹配数据组对应的已注册掌纹图像确定为所述目标掌纹图像。
进一步地,所述掌纹图像的识别装置还包括:
确定模块,用于当接收到掌纹图像时,根据预设网络模型,对所述掌纹图像进行左右手识别,确定所述掌纹图像的左右手属性;
所述识别模块还用于根据所述左右手属性,识别所述掌纹图像上的纹路裁剪点;
生成模块,用于根据所述纹路裁剪点,对所述掌纹图像进行裁剪,生成待处理掌纹图像。
本申请掌纹图像的识别装置具体实施方式与上述掌纹图像的识别方法各实施例基本相同,在此不再赘述。
此外,本申请实施例还提出一种计算机可读存储介质,计算机可读存储介质可以是易失性的,也可以是非易失性的。
计算机可读存储介质上存储有掌纹图像的识别程序,掌纹图像的识别程序被处理器执行时实现如下步骤:
当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
本申请计算机可读存储介质的具体实施方式与上述掌纹图像的识别方法各实施例基本相同,在此不再赘述。
在另一个实施例中,本申请所提供的掌纹图像的识别方法,为进一步保证上述所有出 现的数据的私密和安全性,上述所有数据还可以存储于一区块链的节点中。例如待处理掌纹图像及目标掌纹图像等,这些数据均可存储在区块链节点中。
需要说明的是,本申请所指区块链是分布式数据存储、点对点传输、共识机制、加密算法等计算机技术的新型应用模式。区块链(Blockchain),本质上是一个去中心化的数据库,是一串使用密码学方法相关联产生的数据块,每一个数据块中包含了一批次网络交易的信息,用于验证其信息的有效性(防伪)和生成下一个区块。区块链可以包括区块链底层平台、平台产品服务层以及应用服务层等。
需要说明的是,在本文中,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的过程、方法、物品或者系统不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种过程、方法、物品或者系统所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括该要素的过程、方法、物品或者系统中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述,不代表实施例的优劣。
通过以上的实施方式的描述,本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现,当然也可以通过硬件,但很多情况下前者是更佳的实施方式。基于这样的理解,本申请的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来,该计算机软件产品存储在如上所述的一个计算机可读存储介质(如ROM/RAM、磁碟、光盘)中,包括若干指令用以使得一台终端设备(可以是手机,计算机,服务器,空调器,或者网络设备等)执行本申请各个实施例所述的方法。
以上仅为本申请的优选实施例,并非因此限制本申请的专利范围,凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本申请的专利保护范围内。

Claims (20)

  1. 一种掌纹图像的识别方法,所述掌纹图像的识别方法包括以下步骤:
    当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
    将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
    根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
  2. 如权利要求1所述的掌纹图像的识别方法,其中,每一所述已注册掌纹图像以划分为预设数量分块注册图像的形式存在于预设数据库中;
    所述将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组的步骤包括:
    将所述待处理掌纹图像划分为预设数量的分块掌纹图像,并针对每一所述已注册掌纹图像均执行以下步骤:
    根据各所述分块掌纹图像在所述待处理掌纹图像中的排列位置,确定所述已注册掌纹图像中与各所述分块掌纹图像分别对应的分块注册图像;
    根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
  3. 如权利要求2所述的掌纹图像的识别方法,其中,所述根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
    调用各所述分块掌纹图像,并针对每一所述分块掌纹图像执行以下步骤:
    判断所述分块掌纹图像中是否存在关键数据点,若存在关键数据点,则将所述分块掌纹图像确定为第一匹配单位,并读取所述第一匹配单位对应的第一特征值;
    查找与各所述分块掌纹图像分别对应的分块注册图像中与所述第一匹配单位对应的目标分块注册图像,以及与所述目标分块注册图像相邻的其他分块注册图像,并将所述目标分块注册图像和所述其他分块注册图像一并确定为第二匹配单位;
    读取与所述第二匹配单位对应的第二特征值,根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
  4. 如权利要求3所述的掌纹图像的识别方法,其中,所述根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
    计算所述第一特征值和所述第二特征值之间的第一相似度值,并根据所述第一相似度值,确定第二匹配单位中各注册数据点与第一匹配单位中各所述关键数据点之间的第一匹配数据对;
    对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新,计算更新后的所述第一特征值和所述第二特征值之间的第二相似度值,并根据所述第二相似度值,确定更新后第一匹配单位中各注册数据点与更新后第二匹配单位中各关键数据点之间的第二匹配数据对;
    根据各所述第一匹配数据对和各所述第二匹配数据对,确定所述分块掌纹图像和所述目标分块注册图像之间的目标数据对;
    在所述待处理掌纹图像中的各所述分块掌纹图像均生成所述目标数据对之后,完成所述已注册掌纹图像和所述待处理掌纹图像之间的分块匹配,并将各所述目标数据对生成为匹配数据组。
  5. 如权利要求4所述的掌纹图像的识别方法,其中,所述对所述第一匹配单位、第 二匹配单位、第一特征值和第二特征值进行更新的步骤包括:
    查找与第一匹配单位相邻的其他分块掌纹图像,并将所述第一匹配单位和所述其他分块掌纹图像一并作为新的第二匹配单位;
    根据与新的所述第二匹配单位对应的特征值,对所述第二特征值进行更新;
    将所述目标分块注册图像作为新的第一匹配单位,并根据与新的所述第一匹配单位对应的特征值,对所述第一特征值进行更新。
  6. 如权利要求1-5任一项所述的掌纹图像的识别方法,其中,所述根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像的步骤包括:
    统计每一所述匹配数据组中包含目标数据对的数量,并确定各所述数量中数值最大的目标数量;
    查找多个所述匹配数据组中与所述目标数量对应的目标匹配数据组,并将与目标匹配数据组对应的已注册掌纹图像确定为所述目标掌纹图像。
  7. 如权利要求1-5任一项所述的掌纹图像的识别方法,其中,所述当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像的步骤之前包括:
    当接收到掌纹图像时,根据预设网络模型,对所述掌纹图像进行左右手识别,确定所述掌纹图像的左右手属性;
    根据所述左右手属性,识别所述掌纹图像上的纹路裁剪点;
    根据所述纹路裁剪点,对所述掌纹图像进行裁剪,生成待处理掌纹图像。
  8. 一种掌纹图像的识别装置,所述掌纹图像的识别装置包括:
    读取模块,用于当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
    匹配模块,用于将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
    识别模块,用于根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
  9. 一种掌纹图像的识别设备,所述掌纹图像的识别设备包括存储器、处理器以及存储在所述存储器上并可在所述处理器上运行的掌纹图像的识别程序,所述掌纹图像的识别程序被所述处理器执行时实现如下步骤:
    当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
    将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
    根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
  10. 如权利要求9所述的掌纹图像的识别设备,其中,每一所述已注册掌纹图像以划分为预设数量分块注册图像的形式存在于预设数据库中;
    所述将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组的步骤包括:
    将所述待处理掌纹图像划分为预设数量的分块掌纹图像,并针对每一所述已注册掌纹图像均执行以下步骤:
    根据各所述分块掌纹图像在所述待处理掌纹图像中的排列位置,确定所述已注册掌纹图像中与各所述分块掌纹图像分别对应的分块注册图像;
    根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
  11. 如权利要求10所述的掌纹图像的识别设备,其中,所述根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待 处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
    调用各所述分块掌纹图像,并针对每一所述分块掌纹图像执行以下步骤:
    判断所述分块掌纹图像中是否存在关键数据点,若存在关键数据点,则将所述分块掌纹图像确定为第一匹配单位,并读取所述第一匹配单位对应的第一特征值;
    查找与各所述分块掌纹图像分别对应的分块注册图像中与所述第一匹配单位对应的目标分块注册图像,以及与所述目标分块注册图像相邻的其他分块注册图像,并将所述目标分块注册图像和所述其他分块注册图像一并确定为第二匹配单位;
    读取与所述第二匹配单位对应的第二特征值,根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
  12. 如权利要求11所述的掌纹图像的识别设备,其中,所述根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
    计算所述第一特征值和所述第二特征值之间的第一相似度值,并根据所述第一相似度值,确定第二匹配单位中各注册数据点与第一匹配单位中各所述关键数据点之间的第一匹配数据对;
    对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新,计算更新后的所述第一特征值和所述第二特征值之间的第二相似度值,并根据所述第二相似度值,确定更新后第一匹配单位中各注册数据点与更新后第二匹配单位中各关键数据点之间的第二匹配数据对;
    根据各所述第一匹配数据对和各所述第二匹配数据对,确定所述分块掌纹图像和所述目标分块注册图像之间的目标数据对;
    在所述待处理掌纹图像中的各所述分块掌纹图像均生成所述目标数据对之后,完成所述已注册掌纹图像和所述待处理掌纹图像之间的分块匹配,并将各所述目标数据对生成为匹配数据组。
  13. 如权利要求12所述的掌纹图像的识别设备,其中,所述对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新的步骤包括:
    查找与第一匹配单位相邻的其他分块掌纹图像,并将所述第一匹配单位和所述其他分块掌纹图像一并作为新的第二匹配单位;
    根据与新的所述第二匹配单位对应的特征值,对所述第二特征值进行更新;
    将所述目标分块注册图像作为新的第一匹配单位,并根据与新的所述第一匹配单位对应的特征值,对所述第一特征值进行更新。
  14. 如权利要求9-13任一项所述的掌纹图像的识别设备,其中,所述根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像的步骤包括:
    统计每一所述匹配数据组中包含目标数据对的数量,并确定各所述数量中数值最大的目标数量;
    查找多个所述匹配数据组中与所述目标数量对应的目标匹配数据组,并将与目标匹配数据组对应的已注册掌纹图像确定为所述目标掌纹图像。
  15. 如权利要求9-13任一项所述的掌纹图像的识别设备,其中,所述当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像的步骤之前,所述掌纹图像的识别程序被所述处理器执行时还实现如下步骤:
    当接收到掌纹图像时,根据预设网络模型,对所述掌纹图像进行左右手识别,确定所述掌纹图像的左右手属性;
    根据所述左右手属性,识别所述掌纹图像上的纹路裁剪点;
    根据所述纹路裁剪点,对所述掌纹图像进行裁剪,生成待处理掌纹图像。
  16. 一种计算机可读存储介质,所述计算机可读存储介质上存储有掌纹图像的识别程序,所述掌纹图像的识别程序被处理器执行时实现如下步骤:
    当获取到待处理掌纹图像时,读取预设数据库中的已注册掌纹图像;
    将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组,其中,一所述已注册掌纹图像对应生成一匹配数据组;
    根据多个所述匹配数据组,确定各所述已注册掌纹图像中与所述待处理掌纹图像匹配的目标掌纹图像。
  17. 如权利要求16所述的计算机可读存储介质,其中,每一所述已注册掌纹图像以划分为预设数量分块注册图像的形式存在于预设数据库中;
    所述将各所述已注册掌纹图像分别和所述待处理掌纹图像进行分块匹配,生成多个匹配数据组的步骤包括:
    将所述待处理掌纹图像划分为预设数量的分块掌纹图像,并针对每一所述已注册掌纹图像均执行以下步骤:
    根据各所述分块掌纹图像在所述待处理掌纹图像中的排列位置,确定所述已注册掌纹图像中与各所述分块掌纹图像分别对应的分块注册图像;
    根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
  18. 如权利要求17所述的计算机可读存储介质,其中,所述根据各所述分块掌纹图像以及与各所述分块掌纹图像分别对应的分块注册图像,将所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
    调用各所述分块掌纹图像,并针对每一所述分块掌纹图像执行以下步骤:
    判断所述分块掌纹图像中是否存在关键数据点,若存在关键数据点,则将所述分块掌纹图像确定为第一匹配单位,并读取所述第一匹配单位对应的第一特征值;
    查找与各所述分块掌纹图像分别对应的分块注册图像中与所述第一匹配单位对应的目标分块注册图像,以及与所述目标分块注册图像相邻的其他分块注册图像,并将所述目标分块注册图像和所述其他分块注册图像一并确定为第二匹配单位;
    读取与所述第二匹配单位对应的第二特征值,根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组。
  19. 如权利要求18所述的计算机可读存储介质,其中,所述根据所述第一特征值和所述第二特征值,对所述已注册掌纹图像和所述待处理掌纹图像进行分块匹配,生成匹配数据组的步骤包括:
    计算所述第一特征值和所述第二特征值之间的第一相似度值,并根据所述第一相似度值,确定第二匹配单位中各注册数据点与第一匹配单位中各所述关键数据点之间的第一匹配数据对;
    对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新,计算更新后的所述第一特征值和所述第二特征值之间的第二相似度值,并根据所述第二相似度值,确定更新后第一匹配单位中各注册数据点与更新后第二匹配单位中各关键数据点之间的第二匹配数据对;
    根据各所述第一匹配数据对和各所述第二匹配数据对,确定所述分块掌纹图像和所述目标分块注册图像之间的目标数据对;
    在所述待处理掌纹图像中的各所述分块掌纹图像均生成所述目标数据对之后,完成所述已注册掌纹图像和所述待处理掌纹图像之间的分块匹配,并将各所述目标数据对生成为匹配数据组。
  20. 如权利要求19所述的计算机可读存储介质,其中,所述对所述第一匹配单位、第二匹配单位、第一特征值和第二特征值进行更新的步骤包括:
    查找与第一匹配单位相邻的其他分块掌纹图像,并将所述第一匹配单位和所述其他分块掌纹图像一并作为新的第二匹配单位;
    根据与新的所述第二匹配单位对应的特征值,对所述第二特征值进行更新;
    将所述目标分块注册图像作为新的第一匹配单位,并根据与新的所述第一匹配单位对应的特征值,对所述第一特征值进行更新。
PCT/CN2020/135844 2020-05-20 2020-12-11 掌纹图像的识别方法、装置、设备及计算机可读存储介质 WO2021151359A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010433723.6A CN111814535B (zh) 2020-05-20 2020-05-20 掌纹图像的识别方法、装置、设备及计算机可读存储介质
CN202010433723.6 2020-05-20

Publications (1)

Publication Number Publication Date
WO2021151359A1 true WO2021151359A1 (zh) 2021-08-05

Family

ID=72847706

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/135844 WO2021151359A1 (zh) 2020-05-20 2020-12-11 掌纹图像的识别方法、装置、设备及计算机可读存储介质

Country Status (2)

Country Link
CN (1) CN111814535B (zh)
WO (1) WO2021151359A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111814535B (zh) * 2020-05-20 2024-05-07 平安科技(深圳)有限公司 掌纹图像的识别方法、装置、设备及计算机可读存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103150561A (zh) * 2013-03-19 2013-06-12 华为技术有限公司 人脸识别方法和设备
US20180075282A1 (en) * 2016-09-09 2018-03-15 MorphoTrak, LLC Palm print image matching techniques
CN108921212A (zh) * 2018-06-27 2018-11-30 努比亚技术有限公司 一种图像匹配方法、移动终端及计算机可读存储介质
CN110287775A (zh) * 2019-05-15 2019-09-27 平安科技(深圳)有限公司 手掌图像裁剪方法、装置、计算机设备及存储介质
CN111814535A (zh) * 2020-05-20 2020-10-23 平安科技(深圳)有限公司 掌纹图像的识别方法、装置、设备及计算机可读存储介质

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101251889B (zh) * 2007-12-25 2010-10-13 哈尔滨工业大学 基于手掌静脉与掌纹的近红外成像设备与身份识别方法
WO2018072099A1 (zh) * 2016-10-18 2018-04-26 厦门中控智慧信息技术有限公司 一种手掌检测方法和设备
CN107122700A (zh) * 2017-03-02 2017-09-01 华南理工大学 一种基于视频的掌纹掌脉联合注册和识别方法
CN110852216A (zh) * 2019-10-30 2020-02-28 平安科技(深圳)有限公司 一种掌纹验证方法、装置、计算机设备及可读存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103150561A (zh) * 2013-03-19 2013-06-12 华为技术有限公司 人脸识别方法和设备
US20180075282A1 (en) * 2016-09-09 2018-03-15 MorphoTrak, LLC Palm print image matching techniques
CN108921212A (zh) * 2018-06-27 2018-11-30 努比亚技术有限公司 一种图像匹配方法、移动终端及计算机可读存储介质
CN110287775A (zh) * 2019-05-15 2019-09-27 平安科技(深圳)有限公司 手掌图像裁剪方法、装置、计算机设备及存储介质
CN111814535A (zh) * 2020-05-20 2020-10-23 平安科技(深圳)有限公司 掌纹图像的识别方法、装置、设备及计算机可读存储介质

Also Published As

Publication number Publication date
CN111814535B (zh) 2024-05-07
CN111814535A (zh) 2020-10-23

Similar Documents

Publication Publication Date Title
US8977861B2 (en) Method and system for biometric authentication
US8520903B2 (en) Method and system of accounting for positional variability of biometric features
KR101938033B1 (ko) 카메라 장착 디바이스들과 관련된 생체인식 인증
US7046829B2 (en) Fingerprint verification
CN109284675B (zh) 一种用户的识别方法、装置及设备
JP5801601B2 (ja) 画像認識装置、画像認識装置の制御方法、およびプログラム
US20100080425A1 (en) Minutiae-based template synthesis and matching
US7474769B1 (en) Bioindex mechanism for increasing the relative speed of biometric identification against large population samples
CN109829071B (zh) 人脸图像搜索方法、服务器、计算机设备及存储介质
Uz et al. Minutiae-based template synthesis and matching for fingerprint authentication
WO2023273058A1 (zh) 身份识别方法、系统、装置、计算机设备及存储介质
CN111783882B (zh) 关键点检测方法、装置、电子设备及存储介质
AU2011252761B2 (en) Automatic identity enrolment
CN114359553B (zh) 一种基于物联网的签章定位方法、系统及存储介质
JP2016151831A (ja) 生体情報登録装置、生体情報登録方法、及び生体情報登録プログラム
US10997609B1 (en) Biometric based user identity verification
CN110689046A (zh) 图像识别方法、装置、计算机装置及存储介质
CN111553241B (zh) 掌纹的误匹配点剔除方法、装置、设备及存储介质
AU2011252761A1 (en) Automatic identity enrolment
CN112883980A (zh) 一种数据处理方法及系统
WO2021151359A1 (zh) 掌纹图像的识别方法、装置、设备及计算机可读存储介质
CN110263726B (zh) 一种基于深度相关性特征学习的指静脉识别方法及装置
CN112597978B (zh) 指纹匹配方法、装置、电子设备及存储介质
CN113157962B (zh) 图像检索方法、电子装置和存储介质
Zhou et al. Partial fingerprint indexing: a combination of local and reconstructed global features

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20916383

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20916383

Country of ref document: EP

Kind code of ref document: A1