TWI640929B - Fingerprint identification method and fingerprint identification device - Google Patents

Abstract

The invention proposes a fingerprint identification method. The fingerprint identification method includes: obtaining an object image, and storing a plurality of pixel data of the object image in a first color model format, wherein the plurality of pixel data includes a plurality of first pixel values; converting the plurality of pixel data; Pixel data is in a second color model format, and a plurality of second pixel values are obtained based on the converted plurality of pixel data and a first gain value; according to the plurality of first pixel values and the plurality of second Calculating a plurality of third pixel values by a pixel value; calculating a first standard deviation based on the plurality of third pixel values; and determining whether the first standard deviation is higher than a first preset threshold value, and if the first standard deviation is Above the first preset critical value, the object image is identified as a fingerprint image of a real finger.

Description

Fingerprint recognition method and fingerprint recognition device

The present invention relates to a recognition technology, and more particularly to a fingerprint recognition method and a fingerprint recognition device.

Types of biometrics include face, voice, iris, retina, vein, and fingerprint recognition. Because each person's fingerprint is unique, and the fingerprint is not easy to change with age or physical health, the fingerprint identification device has become the most popular type of biometric identification system. According to different sensing methods, fingerprint recognition devices can also be classified into optical, capacitive, ultrasonic, and thermal sensing types.

However, traditional fingerprint recognition devices cannot effectively distinguish the difference between real and fake fingers, which leads to fake fingers made of silicone by ordinary people, and fingerprints and sweat holes appear on fake fingers made of silicone. In this way, after pressing the fingerprint recognition device with the characteristics of silicon rubber and a fake finger with fingerprints and sweat holes, the fake finger can also deceive the fingerprint recognition device with the deformation characteristics of the finger and the characteristics of the fingerprints and sweat holes. The recognition device cannot correctly recognize whether it is pressed by a real finger, thereby causing a loophole in recognition. In view of this, the present invention will propose solutions of several embodiments below.

The invention provides a fingerprint recognition device and a fingerprint recognition method, which can provide a good fingerprint recognition function, and can effectively recognize whether an object image is a fingerprint image of a real finger, so as to effectively prevent fake fingers from passing through the recognition.

The fingerprint identification method of the present invention is applicable to a fingerprint identification device. The fingerprint identification method includes the following steps: obtaining an object image, and storing a plurality of pixel data of the object image in a first color model format, wherein the plurality of pixel data includes a plurality of first pixel values; The plurality of pixel data are in a second color model format, and a plurality of second pixel values are obtained based on the converted plurality of pixel data and a first gain value; according to the plurality of first pixel values and the plurality of Calculating a plurality of third pixel values with the second pixel value; calculating a first standard deviation based on the plurality of third pixel values; and determining whether the first standard deviation is higher than a first preset threshold value, and if the first If the standard deviation is higher than the first preset threshold, the object image is identified as a fingerprint image of a real finger.

In an embodiment of the present invention, the first color model format is a YUV color model format.

In an embodiment of the present invention, the plurality of first pixel value data are a plurality of brightness values.

In an embodiment of the present invention, the second color model format is an RGB color model format.

In an embodiment of the present invention, the plurality of second pixel values and the plurality of third pixel values are a plurality of red pixel values. The first standard deviation is a standard deviation of a red pixel value.

In an embodiment of the present invention, the step of capturing the object image and storing the plurality of pixel data of the object image in the first color model format includes: capturing a complete object image Image, and sample a partial block of the complete object image as the object image.

In an embodiment of the present invention, the step of calculating the plurality of third pixel values based on the plurality of first pixel values and the plurality of second pixel values includes: separately The pixel values are subtracted from the plurality of second pixel values to obtain the plurality of third pixel values.

In an embodiment of the present invention, the fingerprint identification method further includes the following steps: obtaining a plurality of fourth pixel values based on the converted plurality of pixel data and a second gain value; and according to the plurality of first pixels Calculate a plurality of fifth pixel values and the plurality of fourth pixel values; and calculate a second standard deviation based on the plurality of fifth pixel values.

In an embodiment of the present invention, the above-mentioned determination of whether the first standard deviation is higher than the first preset threshold, and if the first standard deviation is higher than the first preset threshold, identifying The step of the object image being the fingerprint image of the real finger includes: further determining whether the second standard deviation is lower than a second preset threshold, and if the second standard deviation is lower than the first Two preset thresholds identify the object image as the fingerprint image of the real finger.

In an embodiment of the present invention, the plurality of fourth pixel values and the plurality of fifth pixel values are a plurality of green pixel values, and the second standard deviation is a standard deviation of green pixel values.

A fingerprint identification device of the present invention includes a storage device, a fingerprint sensor, and a processor. The fingerprint sensor is used for capturing an image of an object. The processor is coupled to the fingerprint sensor and the storage device. The processor is configured to receive the object image and store a plurality of pixel data of the object image to the storage device in a first color model format. The processor converts the plurality of pixel data into a second color model format, and the processor obtains a plurality of second pixel values based on the converted plurality of pixel data and a first gain value. The processor calculates a plurality of third pixel values according to the plurality of first pixel values and the plurality of second pixel values, and the processor calculates a first standard deviation according to the plurality of third pixel values. The processor determines whether the first standard deviation is higher than a first preset critical value, and if the first standard deviation is higher than the first preset critical value, the processor recognizes the object image A fingerprint image of a real finger.

In an embodiment of the present invention, the first color model format is a YUV color model format.

In an embodiment of the present invention, the plurality of first pixel value data are a plurality of brightness values.

In an embodiment of the present invention, the second color model format is an RGB color model format.

In an embodiment of the present invention, the plurality of second pixel values and the plurality of third pixel values are a plurality of red pixel values, and the first standard deviation is a standard deviation of red pixel values.

In an embodiment of the present invention, the processor captures a complete object image, and the processor samples a partial block of the complete object image as the object image.

In an embodiment of the present invention, the above-mentioned processor respectively subtracts the plurality of first pixel values from the plurality of second pixel values to obtain the plurality of third pixel values.

In an embodiment of the present invention, the processor obtains a plurality of fourth pixel values based on the converted plurality of pixel data and a second gain value, and the processor is configured according to the plurality of first pixel values. The pixel value and the plurality of fourth pixel values calculate a plurality of fifth pixel values. The processor calculates a second standard deviation according to the plurality of fifth pixel values.

In an embodiment of the present invention, the processor further determines whether the second standard deviation is lower than a second preset threshold. If the second standard deviation is lower than the second preset threshold, the processor recognizes the object image as the fingerprint image of the real finger.

In an embodiment of the present invention, the plurality of fourth pixel values and the plurality of fifth pixel values are a plurality of green pixel values, and the second standard deviation is a standard deviation of green pixel values.

Based on the above, the fingerprint recognition device and the fingerprint recognition method of the present invention can analyze and calculate at least a part of the object image of the object image to obtain a calculated number of specific pixel values for the part of the object image. Standard deviation. In addition, the fingerprint recognition device of the present invention can effectively determine whether the object image belongs to a fingerprint image of a real finger by judging the size of the standard deviation by a preset threshold value, so as to prevent fake fingers from passing through the recognition.

In order to make the above features and advantages of the present invention more comprehensible, embodiments are hereinafter described in detail with reference to the accompanying drawings.

In order to make the content of the present invention easier to understand, the following presents multiple embodiments to describe the present invention, but the present invention is not limited to the illustrated multiple embodiments. Appropriate combinations are also allowed between embodiments. In addition, wherever possible, the same reference numbers are used in the drawings and embodiments to refer to the same or similar components.

FIG. 1 is a block diagram of a fingerprint recognition device according to an embodiment of the present invention. Referring to FIG. 1, in this embodiment, the fingerprint identification device 100 includes a processor 110, a fingerprint sensor 120, and a storage device 130. The processor 110 is coupled to the fingerprint sensor 120 and the storage device 130. In this embodiment, the fingerprint sensor 120 is used to capture an object image and provide the object image to the processor 110 so that the processor 110 analyzes the object image. The storage device 130 stores a plurality of program modules. The processor 110 can read these program modules of the storage device 130 to implement the fingerprint identification method according to the embodiments of the present invention. Also. In this embodiment, the fingerprint sensor 120 may include a light source, a light receiver, and the like, such as an optical fingerprint sensing element, but the present invention is not limited thereto. In an embodiment, the fingerprint sensor 120 may also include a capacitive or other type of fingerprint sensing element.

In this embodiment, the processor 110 is, for example, a central processing unit (CPU), a system on chip (SOC), or other programmable general-purpose or special-purpose microprocessors. ), Digital Signal Processor (DSP), Programmable Controller, Application Specific Integrated Circuits (ASIC), Programmable Logic Device (PLD), other similar processing Device or a combination of these devices.

In this embodiment, the storage device 130 is, for example, any type of fixed or removable random access memory (RAM), read-only memory (ROM), and flash memory. Flash memory or similar elements or a combination of the aforementioned elements. In this embodiment, the storage device 130 is used to store object image data and a plurality of program modules, etc., so that the processor 110 can read the storage device 130 and execute these data and program modules. In order to implement the fingerprint identification method according to the embodiments of the present invention.

FIG. 2 is a schematic diagram of an object image according to an embodiment of the present invention. Refer to FIG. 1 and FIG. 2. In this embodiment, the object image 200 includes a plurality of pixel data, and the object image 200 may be a complete object image. In this embodiment, the fingerprint sensor 120 provides the object image 200 to the processor 110, and the processor 110 stores a plurality of pixel data of the object image 200 in a first color model. In this embodiment, the first color model is a YUV color model format, but the present invention is not limited thereto. In an embodiment, the first color model may also be in another type of color model format.

In this embodiment, the object image 200 captured by the fingerprint sensor 120 may be a complete object image, and the object image 200 may include, for example, a fingerprint 210. However, in this embodiment, the processor 110 may capture a part of the object image 200 for analysis. That is, the processor 110 may sample a part of the object image 220 of the object image 200 and perform the following image analysis and identification operations on the part of the object image 220. For example, the object image 200 may have a number of pixels of 320 × 240, and a part of the object image 220 may have a number of pixels of 60 × 60, for example. The processor 110 may capture a part of the object image 220 in the center of the object image 200 or a position having important features, but the present invention is not limited thereto. Therefore, the processor 110 in this embodiment can reduce the calculation amount of image analysis, and can effectively determine whether the object image belongs to a fingerprint image of a real finger at the same time.

FIG. 3 is a schematic diagram of a color model format for converting pixel data of an object image according to an embodiment of the present invention. Referring to FIG. 1 to FIG. 3, in this embodiment, the processor 110 may analyze a part of the object image 220 to obtain pixel data Y (0,0) / U (0,0) of each pixel in the part of the object image 220. 0) / V (0,0) ~ Y (3,3) / U (3,3) / V (3,3). Specifically, as shown in the data matrix 310 of FIG. 3. The partial object image 220 has, for example, a number of pixels of 4 × 4. In this embodiment, the pixel data of each pixel in the partial object image 220 is Y (0,0) / U (0,0) / V (0,0) ~ Y (3,3) / U (3 , 3) / V (3,3) is the data in YUV color model format. In addition, the processor 110 may convert these pixel data Y (0,0) / U (0,0) / V (0,0) ~ Y (3,3) / U according to the following formulas (1) to (3). (3,3) / V (3,3) is a second color model format, where the second color model format is, for example, an RGB color model format, but the present invention is not limited thereto. In an embodiment, the second color model may also be in another type of color model format. Formulas (1) to (3) are as follows: ……......Formula 1) ............ Formula (2) ............ Formula (3)

Therefore, as shown in the data matrix 320 of FIG. 3, the processor 110 sets these pixel data Y (0,0) / U (0,0) / V (0,0) ~ Y (3,3) / U (3,3 ) / V (3,3) converted to multiple pixel data in RGB color model format R (0,0) / G (0,0) / B (0,0) ~ R (3,3) / G (3 , 3) / B (3,3). Moreover, in an embodiment, the processor 110 may further adjust the pixel data R (0,0) / G (0,0) / B (0,0) ~ R (3,3) by different gain values. / G (3,3) / B (3,3) to analyze and calculate the pixel data adjusted by the gain value.

FIG. 4A is a schematic diagram of adjusting an object image based on a first gain value according to an embodiment of the present invention. FIG. 4B is a schematic diagram of adjusting an object image based on a second gain value according to an embodiment of the present invention. Referring to FIG. 1 to FIG. 4B, for example, the processor 110 may adjust the pixel data R (0,0) / G () of the data matrix 320 of FIG. 3 by using the first gain value and the second gain value, respectively. 0,0) / B (0,0) ~ R (3,3) / G (3,3) / B (3,3). In this embodiment, the first gain value may be, for example, an RGB gain (Gain) of 1: 1: 1, and the second gain value may be, for example, an RGB gain of 1: 2: 2, but the present invention is not limited thereto. Therefore, the adjusted partial object image 410 may be a lighter color image, and the adjusted partial object image 420 may be a darker color image. However, the ratio of the gain value according to the present invention is not limited to this. In one embodiment, the ratio of the first gain value and the second gain value may be determined according to different usage requirements or the requirements of the fingerprint recognition device.

FIG. 5 is a schematic diagram illustrating pixel data of a calculation object image according to an embodiment of the present invention. Referring to FIG. 1 to FIG. 5, in this embodiment, the processor 110 may convert the pixel data R (0,0) / G (0,0) / B (0,0) to R (3,3). / G (3,3) / B (3,3) calculates the pixel data of each pixel according to the following formula (4) to formula (6) after the first gain value is adjusted. Formulas (4) to (6) are as follows: ............ Formula (4) ............ Formula (5) ............ Formula (6)

In the above formulas (4) to (6), i and j are positive integers greater than 0. Therefore, as shown in the data matrix 510 of FIG. 5, the processor 110 can obtain multiple pixel data ΔYR (0,0) / ΔYG (0,0) / ΔYB (0,0) ~ ΔYR (3,3) / ∆YG (3,3) / ∆YB (3,3).

It is worth noting that in this embodiment, the processor 110 is based on these pixel data ΔYR (0,0) / ΔYG (0,0) / ΔYB (0,0) ~ ΔYR (3,3) / ΔYG (3,3) / ΔYB (3,3) pixel value of at least one of the types to calculate its standard deviation. For example, first, the processor 110 may use the brightness values Y (0,0) ~ Y (3,3) of each pixel in the data matrix 310 of FIG. 3 as a plurality of first pixel values, and The red pixel values R (0,0) to R (3,3) of each pixel in the data matrix 320 of 3 are adjusted as a plurality of second pixel values after being adjusted by the first gain value. Then, the processor 110 may subtract the first pixel values from the second pixel values to obtain a plurality of pixel values ΔYR (0,0) ~ ΔYR of each pixel in the data matrix 510 of FIG. 5. (3,3). The processor 110 uses these pixel values ΔYR (0,0) ~ ΔYR (3,3) as a plurality of third pixel values, and the processor 110 calculates these third values according to the following formula (7) and formula (8) The first standard deviation of the pixel values. Formula (7) and formula (8) are as follows: ............ Formula (7) ............ Formula (8)

In the above formulas (7) and (8), X _k is ΔYR (0,0) ~ ΔYR (3,3). Therefore, the processor 110 can obtain a standard deviation SD (R) corresponding to the pixel values ΔYR (0,0) to ΔYR (3,3). In this embodiment, the processor 110 determines whether the standard deviation SD (R) is higher than a first preset threshold. If the standard deviation SD (R) is higher than the first preset threshold, the processor 110 recognizes the object image 200 as a fingerprint image of a real finger. That is, since the color of the fingerprint image of a real finger has a specific flesh color, it is possible to effectively distinguish that the object image belongs to a fingerprint image of a real finger or a fake finger by calculating a standard deviation about a specific pixel value.

In this embodiment, the red pixel value of the fingerprint image of the real finger should be higher than the first preset critical value after the above adjustment and the calculated standard deviation SD (R). On the contrary, the red pixel value of the fingerprint image of the fake finger will not be higher than the first preset threshold value after the adjustment and the calculated standard deviation SD (R). Therefore, the fingerprint recognition device 100 of this embodiment can recognize whether the object image belongs to the fingerprint image of a real finger according to the above-mentioned determination method.

For another example, the above-mentioned standard deviation calculation method can also be applied to calculate the second standard deviation. First, the processor 110 may use the luminance values Y (0,0) ~ Y (3,3) of each pixel in the data matrix 310 of FIG. 3 as a plurality of first pixel values, and use the data matrix of FIG. 3 The green pixel values G (0,0) ~ G (3,3) of each pixel in 320 are adjusted as the fourth gain value after being adjusted by the second gain value. Then, the processor 110 may subtract the first pixel values from the fourth pixel values to obtain a plurality of pixel values ΔYG (0,0) ~ ΔYG of each pixel in the data matrix 510 of FIG. 5. (3,3). The processor 110 uses these pixel values ΔYG (0,0) ~ ΔYG (3,3) as a plurality of fifth pixel values, and the processor 110 calculates these fifth values according to the above formula (7) and formula (8) Standard deviation SD (G) of the pixel value.

Therefore, the processor 110 can obtain a standard deviation SD (G) corresponding to the pixel values ΔYG (0,0) to ΔYG (3,3). In this embodiment, the processor 110 determines whether the standard deviation SD (G) is higher than a second preset threshold. If the standard deviation SD (G) is higher than the second preset threshold, the processor 110 recognizes the object image 200 as a fingerprint image of a real finger. In other words, since the color of the fake finger may also be flesh, in addition to the standard deviation SD (R) calculated by the third pixel values adjusted by the first gain value described above, the fingerprint identification device 100 of the present invention The standard deviation SD (G) calculated by the fifth pixel values adjusted by the second gain value can be further judged to effectively prevent a fake finger with a flesh color from passing through the identification.

In this embodiment, the green pixel value of the fingerprint image of the real finger should be higher than the second preset critical value through the above adjustment and the calculated standard deviation SD (G). On the contrary, the green pixel value of the fingerprint image of the fake finger will not be higher than the second preset critical value after the above adjustment and the calculated standard deviation SD (G). Therefore, the fingerprint recognition device 100 of this embodiment can recognize whether the object image belongs to the fingerprint image of a real finger according to the above-mentioned determination method.

The following further presents the experimental results of multiple samples in Table 1 to assist in explaining the foregoing implementation example.         <TABLE border = "1" borderColor = "# 000000" width = "85%"> <TBODY> <tr> <td> Sample </ td> <td> GAIN A </ td> <td> GAIN B </ td> <td> result </ td> </ tr> <tr> <td> SD (R) </ td> <td> SD (G) </ td> <td> SD (B) </ td> <td> SD (R) </ td> <td> SD (G) </ td> <td> SD (B) </ td> </ tr> <tr> <td> FAKE1 </ td> <td > 3 </ td> <td> 13 </ td> <td> 38 </ td> <td> 5 </ td> <td> 61 </ td> <td> 37 </ td> <td> FAIL </ td> </ tr> <tr> <td> FAKE2 </ td> <td> 3 </ td> <td> 15 </ td> <td> 37 </ td> <td> 5 </ td > <td> 61 </ td> <td> 37 </ td> <td> FAIL </ td> </ tr> <tr> <td> FAKE3 </ td> <td> 83 </ td> <td > 5 </ td> <td> 120 </ td> <td> 2 </ td> <td> 47 </ td> <td> 46 </ td> <td> FAIL </ td> </ tr> <tr> <td> FAKE4 </ td> <td> 124 </ td> <td> 1 </ td> <td> 97 </ td> <td> 2 </ td> <td> 123 </ td > <td> 42 </ td> <td> FAIL </ td> </ tr> <tr> <td> FAKE5 </ td> <td> 2 </ td> <td> 92 </ td> <td > 43 </ td> <td> 122 </ td> <td> 1 </ td> <td> 62 </ td> <td> FAIL </ td> </ tr> <tr> <td> FAKE6 < / td> <td> 1 </ td> <td> 81 </ td> <td> 44 </ td> <td> 95 </ td> <td> 1 </ td> <td> 96 </ td > <td> FAIL </ td> </ tr> <tr> <td> FAKE7 </ td> <td> 76 </ td> <td> 1 </ td> <td> 43 </ td> < td> 3 </ td> <td> 124 </ td> <td> 41 </ td> <td> FAIL </ td> </ tr> <tr> <td> TRUE </ td> <td> 30 </ td> <td> 1 </ td> <td> 41 </ td> <td> 3 </ td> <td> 8 </ td> <td> 40 </ td> <td> PASS </ td> </ tr> </ TBODY> </ TABLE> Table 1       

Take the identification results of multiple samples as an example. According to the above Table 1, these samples include samples FAKE1 to FAKE7, and TRUE. In this embodiment, the processor 110 obtains the partial object images of these samples through the first deviation value (GAIN A) and the second gain value (GAIN B) through the standard deviation calculation method described in the above embodiment. Calculated standard deviation of the red, green, and blue pixel values of each pixel and the corresponding brightness values.

For example, in this embodiment, the processor 110 may separately determine whether the standard deviation SD (R) of the first gain value (GAIN A) of these samples is higher than a preset critical value of 30. And the processor 110 further judges whether the standard deviation SD (G) of the second gain value (GAIN B) of these samples is lower than a preset critical value of 30, respectively. In other words, the processor 110 can recognize whether the image is captured from a real finger by using images with different degrees of color cast. Therefore, in Table 1 above, because only the sample TRUE meets the above two standard deviation conditions, the processor 110 may determine that the sample TRUE is a fingerprint image of a real finger. However, in an embodiment, the processor 110 may also set a plurality of preset thresholds to judge the standard deviations of other pixel values adjusted through different gain values respectively. Alternatively, the processor 110 may set one or more preset thresholds to determine a standard deviation of at least one pixel value adjusted by a single gain value, and the present invention is not limited thereto.

It is worth noting that after the object image 200 passes the above identification operation, the fingerprint identification device 100 may further perform a fingerprint authentication operation on the object image 200 to determine whether the fingerprint characteristics in the object image 200 are the same as the fingerprint identification device. 100 pre-registered fingerprint characteristics match. However, the fingerprint authentication operations described in the embodiments of the present invention are those skilled in the art can obtain sufficient teaching, suggestions and implementation instructions based on the common knowledge in the technical field, so they will not be repeated here.

FIG. 6 is a flowchart of a fingerprint identification method according to an embodiment of the present invention. Referring to FIG. 1 and FIG. 6, the fingerprint identification method of FIG. 6 can be applied to at least the fingerprint identification device 100 of FIG. 1. In step S610, the processor 110 obtains the object image through the fingerprint sensor 120, and stores a plurality of pixel data of the object image to the storage device 130 in a first color model format, where the pixel data includes a plurality of first Pixel values. In step S620, the processor 110 converts the pixel data into a second color model format, and obtains a plurality of second pixel values based on the converted pixel data and the first gain value. In step S630, the processor 110 calculates a plurality of third pixel values according to the first pixel values and the second pixel values. In step S640, the processor 110 calculates a first standard deviation according to the third pixel values. In step S650, the processor 110 determines whether the first standard deviation is higher than the first preset critical value. If the first standard deviation is higher than the first preset critical value, the object image is identified as a fingerprint image of a real finger . Therefore, the fingerprint identification method of this embodiment can effectively identify whether an object image is a fingerprint image of a real finger, so as to effectively prevent fake fingers from passing through the identification.

In addition, regarding the related implementation and component characteristics of the fingerprint recognition device 100, sufficient teachings, suggestions, and implementation descriptions can be obtained from the content of the above-mentioned embodiments of FIGS. 1-5, and therefore will not be described again.

In summary, the fingerprint recognition device and fingerprint recognition method of the present invention can capture at least a part of the object image for analysis. First, the fingerprint recognition device of the present invention can adjust multiple pixel values of the part image by using different gain values. Then, the fingerprint recognition device of the present invention can further calculate the pixel values of the part object image to obtain the standard deviation corresponding to the pixel values. Finally, the fingerprint identification device of the present invention can determine the size of the standard deviation by a preset threshold value to determine whether the object image belongs to a fingerprint image of a real finger. According to this, the fingerprint recognition device and fingerprint recognition method of the present invention can effectively prevent fake fingers from passing through the recognition.

Although the present invention has been disclosed as above with the examples, it is not intended to limit the present invention. Any person with ordinary knowledge in the technical field can make some modifications and retouching without departing from the spirit and scope of the present invention. The protection scope of the present invention shall be determined by the scope of the attached patent application.

100: fingerprint identification device 110: processor

120‧‧‧Fingerprint sensor

130‧‧‧Storage device

200‧‧‧ Object image

210‧‧‧ Fingerprint

220, 410, 420‧‧‧partial fingerprint images

310, 320, 510‧‧‧ Data Matrix

S610, S620, S630, S640, S650‧‧‧ steps

FIG. 1 is a block diagram of a fingerprint recognition device according to an embodiment of the present invention. FIG. 2 is a schematic diagram of an object image according to an embodiment of the present invention. FIG. 3 is a schematic diagram of a color model format for converting pixel data of an object image according to an embodiment of the present invention. FIG. 4A is a schematic diagram of adjusting an object image based on a first gain value according to an embodiment of the present invention. FIG. 4B is a schematic diagram of adjusting an object image based on a second gain value according to an embodiment of the present invention. FIG. 5 is a schematic diagram illustrating pixel data of a calculation object image according to an embodiment of the present invention. FIG. 6 is a flowchart of a fingerprint identification method according to an embodiment of the present invention.

Claims (16)

A fingerprint identification method includes: obtaining an object image, and storing a plurality of pixel data of the object image in a first color model format, wherein the plurality of pixel data includes a plurality of first pixel values; The plurality of pixel data is a second color model format, and a plurality of second pixel values are obtained based on the converted plurality of pixel data and a first gain value, wherein the first color model format is a YUV mode, And the second color model format is RGB mode; the plurality of first pixel values are subtracted from the plurality of second pixel values respectively to obtain a plurality of third pixel values; according to the plurality of third Calculate a first standard deviation of the pixel value; and determine whether the first standard deviation is higher than a first preset threshold; if the first standard deviation is higher than the first preset threshold, identify the The object image is a fingerprint image of a real finger. The fingerprint identification method according to item 1 of the scope of patent application, wherein the plurality of first pixel values are a plurality of brightness values. The fingerprint recognition method according to item 1 of the scope of patent application, wherein the plurality of second pixel values and the plurality of third pixel values are a plurality of red pixel values, and the first standard deviation is a red pixel Value standard deviation. The fingerprint recognition method according to item 1 of the scope of patent application, wherein the step of capturing the object image and storing the plurality of pixel data of the object image in the first color model format includes: capturing Take a complete object image, and sample a part of the complete object image as the object image. The fingerprint identification method according to item 1 of the scope of patent application, wherein the step of calculating the plurality of third pixel values according to the plurality of first pixel values and the plurality of second pixel values includes: The plurality of first pixel values are subtracted from the plurality of second pixel values to obtain the plurality of third pixel values. The fingerprint identification method according to item 1 of the scope of patent application, further comprising: obtaining a plurality of fourth pixel values based on the converted plurality of pixel data and a second gain value; and respectively dividing the plurality of first pixels. Subtracting values from the plurality of fourth pixel values to obtain a plurality of fifth pixel values; and calculating a second standard deviation based on the plurality of fifth pixel values. The fingerprint recognition method according to item 6 of the scope of patent application, wherein it is determined whether the first standard deviation is higher than the first preset threshold, and if the first standard deviation is higher than the first preset threshold Value, the step of identifying the object image as the fingerprint image of the real finger includes: further determining whether the second standard deviation is lower than a second preset critical value, and if the second standard deviation is Below the second preset threshold, the object image is identified as the fingerprint image of the real finger. The fingerprint recognition method according to item 6 of the scope of patent application, wherein the plurality of fourth pixel values and the plurality of fifth pixel values are multiple green pixel values, and the second standard deviation is a green pixel Value standard deviation. A fingerprint recognition device includes: a storage device; a fingerprint sensor for capturing an image of an object; and a processor coupled to the fingerprint sensor and the storage device, the processor is used for Receiving the object image, and storing a plurality of pixel data of the object image to the storage device in a first color model format, wherein the processor converts the plurality of pixel data into a second color model Format, and the processor obtains multiple second pixel values based on the converted pixel data and a first gain value, wherein the first color model format is a YUV mode and the second color model The format is RGB mode, wherein the processor subtracts the multiple first pixel values from the multiple second pixel values to obtain multiple third pixel values, and the processor is based on the multiple Calculate a first standard deviation for each third pixel value, wherein the processor determines whether the first standard deviation is higher than a first preset critical value, and if the first standard deviation is higher than the first preset threshold Critical value, then The object image recognition processor is a real fingerprint image of a finger. The fingerprint recognition device according to item 9 of the scope of patent application, wherein the plurality of first pixel values are a plurality of brightness values. The fingerprint recognition device according to item 9 of the scope of patent application, wherein the plurality of second pixel values and the plurality of third pixel values are a plurality of red pixel values, and the first standard deviation is a red pixel Value standard deviation. The fingerprint recognition device according to item 9 of the scope of patent application, wherein the processor captures a complete object image, and the processor samples a part of the complete object image as the object map image. The fingerprint recognition device according to item 9 of the scope of patent application, wherein the processor subtracts the plurality of first pixel values from the plurality of second pixel values, respectively, to obtain the plurality of third pixels. value. The fingerprint identification device according to item 9 of the scope of patent application, wherein the processor obtains a plurality of fourth pixel values based on the converted plurality of pixel data and a second gain value, and the processor respectively The plurality of first pixel values are subtracted from the plurality of fourth pixel values to obtain a plurality of fifth pixel values, wherein the processor calculates a second standard deviation according to the plurality of fifth pixel values. The fingerprint recognition device according to item 14 of the scope of patent application, wherein the processor further determines whether the second standard deviation is lower than a second preset threshold, and if the second standard deviation is lower than the first With two preset thresholds, the processor recognizes the object image as the fingerprint image of the real finger. The fingerprint recognition device according to item 14 of the scope of patent application, wherein the plurality of fourth pixel values and the plurality of fifth pixel values are a plurality of green pixel values, and the second standard deviation is a green pixel Value standard deviation.