TW201835813A - Method and electrical device for merging fingerprint images - Google Patents

Abstract

A method for merging fingerprint images are provided and includes: setting up a hierarchy structure, in which the hierarchy structure includes more than 2 levels, each slice in a lowest level of the levels is a single fingerprint image generated by a fingerprint sensor, a slice in a second level of the levels includes at most M slices in a first level of the levels, the second level is one level higher than the first level, and M is a positive integer greater than 1; obtaining a new fingerprint image, adding the new fingerprint image into the lowest level, and updating the hierarchy structure; and outputting an enroll fingerprint image according to a slice in a highest level.

Description

 Fingerprint image merging method and electronic device  

The invention relates to a method for merging fingerprint images, and in particular to a method for merging fingerprint images by using a hierarchical structure.

Human fingerprints are characterized by many details, are almost unique, are not easily changeable, and are durable in personal life, making fingerprints well suited for long-term identification as human identity. There are at least two stages in the application of fingerprints. The first stage is enrolling, and the second stage is to identify two people's fingerprints or identity verification. During the registration phase, the user must press a finger on the fingerprint sensor to obtain sufficient fingerprint characteristics. However, in some applications, the size of the fingerprint sensor is not too large due to factors such as cost, so the user must press the fingerprint sensor multiple times during the registration phase, and then an algorithm is needed to merge. Multiple fingerprint images. A conventional algorithm first obtains the first chapter fingerprint image, and the subsequent fingerprint images are merged into the first fingerprint image. However, the disadvantage of this method is that when the first fingerprint image is unclear, it is merged. There are many errors in the fingerprint image, and there are even problems with error propagation. Therefore, how to propose a better fingerprint image merging method is a topic of concern to those skilled in the field.

Embodiments of the present invention provide a method for merging fingerprint images, which is applicable to an electronic device. The merging method includes: establishing a hierarchy, wherein the hierarchy includes a plurality of tiers, the number of the tiers being greater than 2, wherein each of the lowest tiers is a single fingerprint image generated by the fingerprint sensor, in the second tier The segment contains up to M segments in the first level, the second level is higher than the first level, and M is a positive integer greater than 1; the new fingerprint image is acquired, and the new fingerprint image is added to the lowest level, and Update the hierarchy; and output the registered fingerprint image according to the segments in the highest level in the hierarchy.

In some embodiments, the merging method further includes: setting each segment in the lowest level to be complete; and if a segment in the ith level includes M slices in the i-1th level For a complete fragment, the fragment set in the i-th level is complete, where i is a positive integer greater than one.

In some embodiments, the step of updating the hierarchy includes: setting the Lth level to the lowest level, setting the new fingerprint image as an incoming fragment, where L is a positive integer; attempting to merge the incoming fragment into the Lth +1 one of the incomplete segments to generate the first segment in the L+1th level; and the step of merging the incoming segment into the incomplete segment in the L+1th level succeeds and A segment is incomplete, searching for a complete segment in the Lth hierarchy to merge into the first segment.

In some embodiments, the step of updating the hierarchy further includes: if the attempt to merge the incoming fragment into the incomplete segment in the L+1th hierarchy is unsuccessful, attempting to merge the incoming fragment into the Lth a complete segment of the hierarchy to produce the second segment in the L+1th level; and if the attempt to merge the incoming segment into the complete segment in the Lth hierarchy is successful and the second segment is incomplete, The complete segment in the Lth hierarchy is searched for merging into the second segment.

In some embodiments, the step of updating the hierarchy further comprises: if the step of merging the incoming fragments into the incomplete segments in the L+1th level succeeds or attempts to merge the incoming fragments into the Lth The step of the complete segment in the hierarchy is successful, determining whether the segment has changed from incomplete to complete; if the third segment has changed from incomplete to complete, setting L=L+1, and treating the third segment as an incoming segment; And the step of repeating the attempt to merge the incoming segments into the incomplete segments in the L+1th level.

In some embodiments, the step of updating the hierarchy further includes: if the step of attempting to merge the incoming fragment into the complete fragment in the Lth hierarchy is unsuccessful or no fragment is changed from incomplete to complete, determining whether the output can be output Registering a fingerprint image; and if it is unable to output the registration fingerprint image, waiting to receive another new fingerprint image.

In some embodiments, the step of determining whether the fingerprint image for registration can be output includes: obtaining a maximum segment in the highest level, and attempting to merge all segments in the hierarchy into the largest segment; determining that the largest segment is included Whether the number of fingerprint images is greater than a preset value; and if the number of fingerprint images included in the largest segment is greater than a preset value, the largest segment is output as a fingerprint image for registration.

In another aspect, embodiments of the present invention provide an electronic device including a fingerprint sensor and a computing circuit. The fingerprint sensor is used to acquire a new fingerprint image. The computing circuit is coupled to the fingerprint sensor for establishing a hierarchy. The hierarchy includes a plurality of levels, the number of which is greater than two, and each of the lowest levels is a single fingerprint image produced by the fingerprint sensor. The segments in the second level contain a maximum of M segments in the first hierarchy, the second hierarchy is higher than the first hierarchical level, and M is a positive integer greater than one. The calculation circuit is used to add the new fingerprint image to the lowest level, update the hierarchy, and output the registration fingerprint image according to the segment in a highest level in the hierarchy.

In some embodiments, the computing circuit is also operative to perform the merging method described above.

The above described features and advantages of the invention will be apparent from the following description.

100‧‧‧Electronic devices

110‧‧‧Finger sensor

120‧‧‧Computation circuit

200‧‧‧ Hierarchy

210(1)~210(N), 210(L), 210(L+1)‧‧‧ level

221, 222 (1) ~ 222 (M), 223, 224 ‧ ‧ fragment

230‧‧‧New fingerprint image

240‧‧‧Registered fingerprint image

310, 320, 330‧‧‧ Fingerprint images

End of 311, 321‧‧

312, 322‧‧ ‧ fork

S401~S412, S1001~S1003‧‧‧ steps

510, 520, 530, 610, 620, 630, 710, 720, 730 ‧ ‧ fragment

FIG. 1 is a schematic diagram showing an electronic device according to an embodiment.

FIG. 2 is a schematic diagram showing a hierarchical structure according to an embodiment.

FIG. 3 is a schematic diagram illustrating a merged segment according to an embodiment.

FIG. 4 is a flow chart showing an update hierarchy according to an embodiment.

[Fig. 5] to [Fig. 9] are intermediate steps of updating an hierarchy according to an embodiment.

FIG. 10 is a flow chart showing a method of merging fingerprint images according to an embodiment.

The terms "first", "second", "etc." used in this document are not intended to mean the order or the order, and are merely to distinguish between elements or operations described in the same technical terms. In addition, as used herein, "coupled" may mean that two elements are electrically connected, either directly or indirectly. That is, when the following description "the first object is coupled to the second object", other items may be disposed between the first object and the second object.

FIG. 1 is a schematic diagram of an electronic device according to an embodiment. Referring to FIG. 1 , the electronic device 100 includes a fingerprint sensor 110 and a calculation circuit 120 . The electronic device 100 may be a smart phone, a tablet computer, a music player, a notebook computer, an industrial computer, or any other form of electronic device, and the present invention is not limited thereto. The fingerprint sensor 110 can be used to obtain a fingerprint image of a human. The sensing manner of the fingerprint sensor 110 can be optical, capacitive or ultrasonic. The invention does not limit the sensing manner of the fingerprint sensor 110. The computing circuit 120 can be a central processing unit, a microprocessor, a microcontroller, a digital signal processor, a baseband processor, an image processing chip, or a special application integrated circuit.

The fingerprint sensor 110 acquires a plurality of fingerprint images of the human and transmits the obtained fingerprint images to the calculation circuit 120, and the calculation circuit 120 merges the fingerprint images. 2 is a schematic diagram showing a hierarchy in accordance with an embodiment. Referring to FIG. 2, the computing circuit 120 creates a hierarchy 200 comprising N levels 210(1)~210(N), N being a positive integer greater than two. The level 210(1) is the lowest level, and the level 210(N) is the highest level. Here, "fragment" is defined as a unit in each level. One of the levels 210(2) includes a maximum of M segments in the hierarchy 210(1), and M is a positive integer greater than one. While one of the levels 210(3) includes a maximum of M segments at level 210(2), and so on. In other words, if there is a second level and a first level in the hierarchy 200, and the second level is higher than the first level and one level, then one of the second levels contains a maximum of M pieces in the first level.

Each segment can be divided into complete or incomplete. First, each of the segments in the lowest level 210(1) is a single fingerprint image produced by the fingerprint sensor 110, and each segment in the lowest level 210(1) is complete. In addition, if a segment in the i-th level includes M complete segments in the i-1th hierarchy, the segment in the i-th hierarchical layer may be set to be complete (otherwise, incomplete), wherein i is a positive integer greater than one. For example, segment 221 includes M segments 222(1)~222(M), so segment 221 is complete. The segment 223 includes only one segment 224, so the segment 223 is incomplete. In another perspective, the M complete segments in the i-1th level can be merged into one complete segment in the i th level. The segment in the highest level 210 (N) can be used to output the registration fingerprint image 240. That is, the registration fingerprint image 240 has a sufficient area or feature point for registration. After the fingerprint sensor 110 obtains the new fingerprint image 230, the new fingerprint image 230 is added to the lowest level 210(1), and then the computing circuit 120 updates the hierarchy 200. The flow of the update will be described below.

First, the merging of two segments will be described. FIG. 3 is a schematic diagram illustrating a merged segment according to an embodiment. Referring to FIG. 3, it is assumed here that the fingerprint image 310 and the fingerprint image 320 are two segments, the lines in the figure refer to the ridges and the space between the ridges is the valley. The fingerprint images 310 and 320 are not fingerprint images for registration, that is, the areas of the fingerprint images 310 and 320 are not large enough. A plurality of features can be extracted from the fingerprint images 310, 320, and generally at least the ending and bifurcation of the ridges can be captured. For example, the fingerprint image 310 has an end point 311 and a bifurcation 312, and the fingerprint image 320 has an end point 321 and a bifurcation 322. Features in the fingerprint image 310 can be paired to features in the fingerprint image 320. For example, the end point 311 and the end point 321 are paired with each other, and the bifurcation 312 and the bifurcation 322 are paired with each other. After the features are paired, the fingerprint images 310, 320 can be combined into a fingerprint image 330. However, if there are not enough features in the fingerprint images 310, 320 or the paired features are not enough, it means that the fingerprint images 310, 320 cannot be merged together. However, the above merged method is merely an example, and the present invention does not limit the selection of those features from the fingerprint image, nor how to combine the two fingerprint images (slices).

In this embodiment, only two complete segments in the same hierarchy can be merged with each other, and incomplete segments in the same hierarchy are not merged with each other; when the merge is a cross-level, the lower-level segments must be complete. of. The following is a detailed description of FIG. 4, which is a flow chart showing an update hierarchy in accordance with an embodiment, the steps of FIG. 4 being performed by computing circuit 120. Referring to FIG. 4, in step S401, a new fingerprint image is acquired from the fingerprint sensor 110, and the new fingerprint image is set as an incoming segment. In step S402, the Lth hierarchical level is set to the lowest hierarchical level, and L is a positive integer. In this embodiment, L is set to 1. In step S403, the incoming segment is added to the Lth level. Referring to FIG. 4 and FIG. 5 simultaneously, the segment 510 is an incoming segment, the circle in the hierarchy is represented as a complete segment, and the square is represented as an incomplete segment. The above positive integer M is greater than 2. When the incoming segment 510 is a new fingerprint image, the level 210 (L) is the level 210 (1), and all the segments in the level 210 (1) are complete, but for a more general representation of the merge process, In Figure 5, the level 210 (L) is used, and the level 210 (L) has incomplete segments.

In step S404, an attempt is made to merge the incoming segment 510 into one of the L+1th levels to generate the first segment. For example, an attempt can be made to merge the incoming segment 510 into the incomplete segment 520 in level 210 (L+1). In some embodiments, if there is more than one incomplete segment in level 210 (L+1), then an incomplete segment that is older (ie, staying at level 210 (L+1) for a longer time) may be selected first. Come to merge. If the step of merging the incoming segment 510 into the incomplete segment 520 is successful, a new segment (referred to herein as the first segment 530) is generated.

If step S404 is successful, then in step S405, it is determined whether the first segment 530 is incomplete. If the first segment 530 is incomplete, the complete segment in the Lth hierarchy is searched for merging into the first segment 530. For example, referring to FIG. 6, since the first segment 530 is incomplete, it may be attempted to merge the segments 610, 620 into the first segment 530. In this embodiment, there are two complete segments 610, 620 in level 210 (L), so older segments 610 can be picked first to merge into first segment 530. Step S405 will be repeated until the first segment 530 is complete or there is no complete segment in the hierarchy 210 (L). It is worth mentioning that although the segment 210 (L) has a segment 630, since the segment 630 is not complete, it does not attempt to merge the segment 630 into the first segment 530. Fragment 630 is responsible for hierarchy 210 (L-1), which may transition to integrity when processing level 210 (L-1).

In the examples of FIGS. 5 and 6, the positive integer M is greater than two. When the positive integer M is equal to 2, there is no incomplete segment in the hierarchy 210 (L+1) (the reason will be explained as follows), so the step S404 will inevitably fail, and the step S405 will not be executed.

On the other hand, if step S404 is unsuccessful, an attempt is made in step S406 to merge the incoming segments into the complete segment in the Lth hierarchy to produce the second segment in the L+1th hierarchy. For example, referring to Figures 4 and 7, an attempt can be made to merge the incoming segment 510 into a complete segment 710 in the same level 210 (L) to produce a second segment 720 in level 210 (L + 1). The positive integer M is greater than 2 in this embodiment, so the second segment 720 should be incomplete. Moreover, if there are more than one complete segment 710, 730 in the hierarchy 210 (L) in addition to the incoming segment 510, the older complete segment 710 may be selected first to attempt the merge. If the incoming segment 510 cannot be merged with the segment 710, then the attempt to merge the incoming segment 510 with the segment 730 is again attempted.

If step S406 is successful, it is determined in step S407 whether the second segment 720 is incomplete, and if the second segment 720 is incomplete, the complete segment in the Lth hierarchy may be searched for merge to the second segment 720. For example, referring to FIG. 4 and FIG. 8, since the second segment 720 is incomplete, the complete segment 730 can be found in the hierarchy 210 (L) to be merged into the second segment 720. If the segment 730 cannot be merged into the second segment 720, the second segment 720 is maintained unchanged.

It is worth mentioning that in the embodiment of Figures 7 and 8, the positive integer M is greater than 2, but if the positive integer M is 2, the segments created are complete. For example, referring to FIG. 9, the second segment 720 generated after combining the incoming segment 510 with the segment 710 is complete. Furthermore, if the positive integer M is 2, all fragments in all levels will be complete.

Referring back to FIG. 4, after step S405 or step S407, step S408 is performed to determine whether a segment has changed from incomplete to complete. In this embodiment, it is determined whether the updated first segment 530 in FIG. 6 or the newly created second segment 720 in FIG. 8 is complete. If there is a third segment (the first segment 530 or the second segment 720) is changed from incomplete to complete, in step S409, L = L + 1 is set, and the third segment is treated as an incoming segment. Next, step S404 is repeatedly performed according to the updated incoming fragment. On the other hand, if the result of step S408 is YES, this means that a complete segment is newly generated in the L+1th level, so it is possible that this new complete segment can continue to be merged to the next level. If the result of step S408 is no, it means that even if a new segment is generated in the L+1th level, since the new segment is incomplete, the new segment cannot be merged to the next level, Step S409 needs to be performed.

On the other hand, when the positive integer M is 2, step S405 is not performed, and the second segment created at step S406 is necessarily complete, so the result of step S408 is necessarily YES.

If the result of step S406 is a failure or the result of step S408 is NO, then step S410 is performed to determine whether to output the registration fingerprint image. Referring back to FIG. 2, in some embodiments, a maximum segment may be found from the highest level 210 (N), which may be the largest fingerprint image or the fingerprint image that combines the most single fingerprint images. After taking the largest segment, try to merge all the segments in hierarchy 200 into this largest segment. Next, it is determined whether the number of fingerprint images included in the largest segment is greater than a preset value (for example, 8). If the number of fingerprint images included in the largest segment is greater than the preset value, the result of step S410 is YES, and the largest segment may be output as the fingerprint image for registration in step S411. However, the present invention is not limited thereto. In other embodiments, after attempting to merge all the segments in the hierarchy 200 into the largest segment, it may also be determined whether the area of the largest segment is greater than a predetermined area or the maximum segment is determined. Whether the number of features is greater than a preset number to determine whether or not to output a fingerprint image for registration. Alternatively, if the fingerprint sensor 110 has obtained more than a predetermined number of fingerprint images, it may be determined to be YES in step S410. The present invention does not limit the specific content of step S410.

If the result of step S410 is NO, the calculation circuit 120 is suspended in step S412, waiting for the next new fingerprint image to come in (step S401).

FIG. 10 is a flow chart showing a method of merging fingerprint images according to an embodiment. Referring to Fig. 10, in step S1001, a hierarchical structure is established. This hierarchy includes more than 2 levels, with each of the lowest levels being a single fingerprint image produced by the fingerprint sensor. When the second level is higher than the first level and one level, a segment in the second level contains a maximum of M pieces in the first level. In step S1002, a new fingerprint image is acquired, a new fingerprint image is added to the lowest level, and the hierarchical structure is updated. In step S1003, the registration fingerprint image is output based on the segment in the highest hierarchy. However, the steps in FIG. 10 have been described in detail above, and will not be described again here. It should be noted that the steps in FIG. 10 can be implemented as multiple codes or circuits, and the present invention is not limited thereto. In addition, the method of FIG. 10 can be used in conjunction with the above embodiments, or can be used alone. In other words, other steps can be added between the steps of FIG. Alternatively, the steps in FIG. 10 may be repeatedly performed a plurality of times (for example, step S1002).

In the merging method proposed by the embodiment of the present invention, the phenomenon of erroneous transmission does not occur, and even if an error occurs, the distribution of errors is relatively uniform. In addition, the registration fingerprint image generated by the above method can increase the accuracy of the subsequent identification stage.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the present invention, and any one of ordinary skill in the art can make some changes and refinements without departing from the spirit and scope of the present invention. The scope of the invention is defined by the scope of the appended claims.

Claims (14)

 A method for merging fingerprint images is applicable to an electronic device, the merging method comprising: establishing a hierarchical structure, wherein the hierarchical structure comprises a plurality of hierarchical levels, the number of the hierarchical levels is greater than 2, and a lowest hierarchical level of the hierarchical levels Each segment is a single fingerprint image generated by a fingerprint sensor, and a segment of a second level of the layers contains a maximum of M segments in a first level, the second level being higher than The first level is a level, M is a positive integer greater than 1; a new fingerprint image is obtained, the new fingerprint image is added to the lowest level, and the hierarchy is updated; and according to a highest level in the levels The fragment outputs a registered fingerprint image.    The merging method of claim 1, further comprising: setting each of the segments in the lowest level to be complete; and if a segment in the i-th level is included in the i-1 M complete segments in the hierarchy, the segment set in the i-th hierarchy is complete, where i is a positive integer greater than one.    The merging method of claim 2, wherein the step of updating the hierarchy comprises: setting the Lth level to the lowest level, and setting the new fingerprint image as an incoming segment, where L is a positive integer; Trying to merge the incoming fragment into an incomplete fragment in the L+1th level to generate a first fragment in the L+1th level; and if attempting to merge the incoming fragment into the first The step of the incomplete segment in the L+1 levels is successful and the first segment is incomplete, searching for a complete segment in the Lth level to merge into the first segment.    The merging method of claim 3, wherein the step of updating the hierarchy further comprises: if the attempt to merge the incoming segment into the incomplete segment in the L+1th hierarchy is unsuccessful Trying to merge the incoming fragment into a complete fragment in the Lth level to generate a second fragment in the L+1th level; and if attempting to merge the incoming fragment into the Lth The steps of the complete segment in the hierarchy are successful and the second segment is incomplete, searching for the complete segment in the Lth hierarchy to merge into the second segment.    The merging method of claim 4, wherein the step of updating the hierarchy further comprises: if the step of merging the incoming segment into the incomplete segment in the L+1th level is successful or The step of attempting to merge the incoming segment into the complete segment in the Lth hierarchy is successful, determining whether the segment has changed from incomplete to complete; if a third segment has changed from incomplete to complete, set L=L +1, and treating the third segment as the incoming segment; and repeating the step of attempting to merge the incoming segment into the incomplete segment in the L+1th level.    The merging method of claim 5, wherein the step of updating the hierarchy further comprises: if the attempt to merge the incoming segment into the complete segment in the Lth hierarchy is unsuccessful or no segment is from The incomplete change is complete, and it is judged whether the registration fingerprint image can be output; and if the registration fingerprint image cannot be output, waiting for another new fingerprint image to be received.    The merging method of claim 6, wherein the step of determining whether the fingerprint image for registration can be output comprises: obtaining a maximum segment in the highest level, and attempting to have all segments in the hierarchy Merging to the maximum segment; determining whether the number of fingerprint images included in the maximum segment is greater than a preset value; and if the number of fingerprint images included in the maximum segment is greater than the preset value, outputting the maximum segment as the Register with a fingerprint image.    An electronic device includes: a fingerprint sensor for acquiring a new fingerprint image; and a computing circuit coupled to the fingerprint sensor, the computing circuit is configured to establish a hierarchical structure, wherein the hierarchical structure includes a plurality of Levels, the number of the levels being greater than 2, each of the lowest levels of the levels being a single fingerprint image generated by the fingerprint sensor, one of the second levels of the levels The segment contains a maximum of M segments in a first level, the second level is higher than the first level, and M is a positive integer greater than 1, and the calculation circuit is used to add the new fingerprint image to the lowest level. In the hierarchy, the hierarchy is updated, and a registration fingerprint image is output according to segments in one of the highest levels.    The electronic device of claim 8, wherein if a segment in the i-th hierarchy includes M segments in the i-1th hierarchy, the calculation circuit is set in the i-th hierarchy The fragment is complete, where i is a positive integer; and each of the segments set by the computing circuit in the lowest level is complete.    The electronic device of claim 9, wherein the calculating circuit sets the Lth level to the lowest level, and sets the new fingerprint image as an incoming segment; the computing circuit attempts to merge the incoming segment into An incomplete segment in the L+1th level to generate a first segment in the L+1th level; and if the attempt is to merge the incoming segment into the L+1th level The step of the incomplete segment is successful and the first segment is incomplete, and the computing circuit searches for a complete segment in the Lth hierarchy to merge into the first segment.    The electronic device of claim 10, wherein the step of merging the incoming segment into the incomplete segment in the L+1th hierarchy is unsuccessful, the computing circuit attempts to pass the incoming Framing merges into a complete segment in the Lth hierarchy to produce a second segment in the L+1th level; and if attempting to merge the incoming segment into the complete in the Lth hierarchy The step of the segment is successful and the second segment is incomplete, and the calculation circuit searches for a complete segment in the Lth hierarchy to merge into the second segment.    The electronic device of claim 11, wherein the step of merging the incoming segment into the incomplete segment in the L+1th level succeeds or attempts to merge the incoming segment into The step of the complete segment in the Lth hierarchy is successful, the calculation circuit determines whether a segment has changed from incomplete to complete; if a third segment changes from incomplete to complete, the calculation circuit sets L=L+1, And treating the third segment as the incoming segment; and the computing circuit repeats the step of attempting to merge the incoming segment into the incomplete segment in the L+1th level.    The electronic device of claim 12, wherein the calculating circuit is unsuccessful if the attempt to merge the incoming segment into the complete segment in the Lth hierarchy is unsuccessful or no segment is changed from incomplete to complete. Determining whether the fingerprint image for registration can be output; and if the fingerprint image for registration is not output, the calculation circuit waits to receive another new fingerprint image.    The electronic device of claim 13, wherein the calculation circuit obtains a maximum segment in the highest level, and attempts to merge all segments in the hierarchy into the maximum segment; the computing circuit determines Whether the number of fingerprint images included in the maximum segment is greater than a preset value; and if the number of fingerprint images included in the maximum segment is greater than the preset value, the calculation circuit outputs the maximum segment as the fingerprint image for registration .