TW528989B - Method and apparatus for discriminating latent fingerprint in optical fingerprint input apparatus - Google Patents

Abstract

Disclosed are a method and an apparatus for discriminating a latent fingerprint that can prevent a fingerprint recognition apparatus from mis-recognizing a latent fingerprint as a fingerprint of a biomass due to a fingerprint residual on an imaging surface of an optical fingerprint input apparatus. The method of discriminating a latent fingerprint according to the invention includes the steps of acquiring an image without illuminating the backlight onto the imaging surface, detecting an existence of a fingerprint from the acquired image, and discriminating the fingerprint, if detected, to be a latent fingerprint detected due to an external light. The apparatus for discriminating a latent fingerprint according the invention includes backlight control means for controlling switching on and off of the backlight, image acquisition means for acquiring a fingerprint image without illuminating the backlight onto the imaging surface, fingerprint detection means for detecting an existence of a fingerprint from the image acquired by the image acquisition means, and fingerprint discriminating means for discriminating that the fingerprint, if detected by the fingerprint detection means, is a latent fingerprint detected due to an external light.

Description

528989

The present invention and the device, and the residual fingerprint of the biometric fingerprint. BACKGROUND OF THE INVENTION The field of the invention is related to preventing the fingerprint identification machine from being identified by the optical finger ~ the latent finger; the image surface of the wheel-in device: The lurking finger means that the gate is locked in a Pei-style way, and the pattern of the image is explained to explain the principle of the light map. An image is aligned when there is no = the identification mechanism has been widely used in the system. Or enter ^, close, type, the second fingerprint entered by the user and the second two =, K and identification. It is used to receive fingerprints to perform fingerprints. It is necessary to classify them into two different types: one type === grain type is a non-optical type. Use an optical fingerprint wheel to enter two types of clothing. Fingers placed on a prism = finger-jackets on the top of the finger) are reflected in the pattern of fingerprints, and the input image is compared with the pre-stored image. Fingerprint input devices are mainly classified as absorption type and scattering type. The schematic diagram of 1 shows a type of absorption type fingerprint input / °; it contains-backlight 112, one hundred and three hundred and ten and one relocation mirror: iron Detector 1 1 6 and an image processor 1 2 5. The backlight i 丨 2 uses multiple LEDs. The 3110 is a rectangular triangle, and when any fingerprint is rotated, it is generated on the image surface. Total reflection. The image sensor 116 is an element that outputs an electrical signal corresponding to the total amount of input light, such as a ccd or CMOS sensor, which is well known to those skilled in the art. The inclined surface of the three-dimensional 110 is one The image plane, and the inner surface of the image plane 1 1 8 causes total reflection

528989 V. Description of the invention (2) Total reflection surface. ^ At the time when there is an input pattern, the light emitted from the backlight Π 2 is totally reflected from the inside of the image plane of the 稜鏡 1 1 〇 at the time when the image is incident on the image plane 1 18, and enters the shirt image through the lens i 丨 4. Test ^ § 1 1 6. If you place your finger on the image plane, the light shines on the indentation in the fingerprint, and a total reflection is generated from the inner surface of the image plane π 8. This reaches the scene. The image sensing is 1 1 6 because of the fingerprint. The indentation of the plane did not make contact with the: Haiying surface. In comparison, the light irradiated to the ridge of the fingerprint does not have total reflection from the inner surface of the image plane 1 1 8 because only a part of the light reaches the image sensor 116. Therefore, the amount of light incident on the image sensor π 6 will have different results depending on the position of the depression or ridge of the fingerprint. As a result, the image perception / 1 1 6 will be different due to the fingerprint pattern. Degree of electrical signal. The image processor 1 2 5 converts the output value of the image sensor π 6 into a digital signal, so the fingerprint pattern is recognized. 2A and 2B are schematic diagrams showing the operation principle of a scattering-type fingerprint input device. The fingerprint input device of FIG. 2A includes a backlight 212, a prism 210, a lens 2 1 4 and an image sensor 2 1 6. The structure of the image sensor is similar to the image sensor in FIG. However, 稜鏡 2 10 is stepped, not triangular. Unlike the absorption type shown in FIG. 1, the light is incident from the backlight 211 into the image plane 218 of the chirp 210 in a manner far smaller than a right angle or a critical angle. Therefore, a person who irradiates the light under the fingerprint without touching the image plane will penetrate the image plane 2 1 8 and will not reach the image sensor 2 1 6. At the same time, the light that hits the ridge of the fingerprint is scattered by the ridge. The scattered light is incident on

528989 Five-mirror invention description (3) 2 1 4 on 'and it is sensed by the image sensor 2 1 6. Example Figure 2B shows the operation principle of another lens input device. In FIG. 2a, the light irradiated to the concave portion of the fingerprint penetrates the image plane 3 11 and will reach the image sensor 3 1 6. The apricot irradiated on the ridge of the fingerprint is scattered out by the same principle. However, the difference lies in the use of a prism or a change in the position of the backlight 312. Using temple waist triangle type In the example of the absorption type fingerprint input device, the fingerprint receives light. Therefore, the ridge displayed on the image sensor is a dark image and the concave image is a bright image. However, in the example of a scattering-type finger = 钤 image, light is scattered off the ridges of the fingerprint. Therefore, the fingerprint image displayed on the% ^ device is a bright image on the spine, and the sensor is a light image, so the = part of the absorption type image input device is dark and completely opposite. In order to simplify the processing and comparison of fingerprint images, the input methods are different and the display on the computer monitor is generated based on avoiding dark and transitional states. Examples of scattering fingerprint input devices are displayed on the bright-view Is of the image. Inverted image. In particular, although the actual fingerprint image lacking i on the computer monitor is = j in the ridge of the fingerprint, it is not a dark image in the recess of the image fingerprint. On the other hand, the gray scale between the number of recesses 4 is an example of an absorption type fingerprint input. It is quite high, such as one. However, in the example of the optical fingerprint wheel-in device, it is in contact with the pattern recognition device, and leaves a contaminated material, which leaves a latent fingerprint on the finger. If the light is on the fingerprint recognition surface from the outside, rather than from the backlight, especially when a particular shot is shot to a reflection = clothing 4, the image feels rigid.

Page 9 528989 V. Description of the invention (4) The device is easy to detect a latent fingerprint. Therefore, if a latent fingerprint is used, the fingerprint recognition | 1 t 1 2 benefits sensing is authenticated, and the fingerprint remains after use. The right user will enter the system, instead of entering through the user, Figure 3A shows the normal reason for a creature, = ,, '_. This image is the same as the one shown in Figure 3A. It is a latent fingerprint and a latent fingerprint. ~ Is very similar. FIG. 3C shows that in order to solve the problem of misidentification, it compares the method used in the image storage technology of someone's last input with the stored image. For example, the newly entered image is determined as one. The latent finger λ solid image storage is very similar, then it is considered to be consistent with each other, or when the two-finger scripture (the location of a specific point in the text). The understandable patterns in each other overlap with each other. This method still has a latent indication that the pattern will change as the nephew is read by the image sensor and therefore cannot be changed on a fairly accurate basis. SUMMARY OF THE INVENTION Identifies a latent fingerprint. The design concept of the present invention is a latent fingerprint generated when a back-optical type 3 fingerprint irradiated onto a fingerprint is input into the device light. Secondary conduction can be detected due to external, so the purpose of the present invention is to identify the detected fingerprint, Doufang = a type of Pei identifying the latent fingerprint on the-image obtained by external = :: The backlight illuminates the image surface 丨 the latent fingerprint detected by Xianyuan. him page 10 528989 V. Description of the invention (5) In order to achieve the above object, the present invention provides a method for identifying a latent fingerprint in an optical fingerprint input device. The method includes the following steps: the backlight is not illuminated under the image plane To obtain an image; to detect whether a fingerprint exists from the obtained image; and if detected, to identify whether the fingerprint is a latent fingerprint detected by an external light source.

The invention also provides a device for recognizing a latent fingerprint in an optical fingerprint input device. The device includes: a backlight control mechanism for controlling the turning off and on of the backlight; and an imaging mechanism for not lighting the backlight. A fingerprint image is obtained under the condition of being on the mapping surface; a fingerprint detection mechanism is used to detect whether a fingerprint exists from the image obtained by the camera mechanism; and a fingerprint recognition mechanism, if detected by the fingerprint detection mechanism If it is a fingerprint, it is identified whether the fingerprint is a latent fingerprint detected by an external light source. In the method and device for identifying a latent fingerprint in the present invention, the method of obtaining the entire picture in an image can be regarded as the method of obtaining an image. In the "off" state of the backlight, the entire picture of the image is obtained and stored in a memory, so that it is possible to detect whether there is a latent fingerprint on the stored image.

However, using the above method wastes a lot of time to get a picture of the entire image, thereby identifying a latent fingerprint. Therefore, if a part of the latent fingerprint is obtained by applying the backlight, and a part of the image is used to identify a latent fingerprint, a considerable amount of time can be saved. Here, when the finger is quite small, or when the finger is in contact with the lower side of the image plane, it is necessary to obtain a finger of about 1/4 to 1/2 of the finger from the tip of the fingerprint image.

Page 11 528989 V. Description of the invention (6) Grain image. At the same time, a fixed window (w i n d o w i n g) function is provided to receive an image in only a part of the area in the image sensor according to the type of image. Therefore, when using this type of image sensor, it is possible to identify whether a fingerprint is detected by selecting an intermediate part of the image that clearly reflects the fingerprint as needed. The way to detect the presence of fingerprints is to use a fingerprint detection mechanism. The fingerprint detection mechanism includes calculating the sum of the differences between the gray levels corresponding to the obtained image, comparing the calculated total value of the gray levels with a predetermined reference value; and if the total value of the gray levels of the γ fruit is less than At the reference value, it is recognized whether a fingerprint has been detected. In particular, when an image has been obtained in the π-off π state of the backlight, and a fingerprint can be clearly detected, the gray scale on the X-axis (or Υ-axis) can be in a very wide range. Within the range, and based on the change, the detected fingerprint is identified, and whether the fingerprint is a latent fingerprint detected by an external light source. The method of identifying latent fingerprints is to sum up the differences in the gray levels of adjacent pixels. This technique can be easily performed by those skilled in the art. In addition to the methods described above, another method for detecting the presence of fingerprints is to calculate the sum of the differences between the gray levels corresponding to the obtained image, and compare the calculated total of the gray levels with a predetermined reference value. ; And if the total value of the gray levels is less than the reference value, identifying whether a fingerprint has been detected. Another method that can detect the presence of a fingerprint is to calculate the average and dispersi value of the gray scale corresponding to the image obtained; and if the calculated average is less than a first reference value, And

Page 12 528989 V. Description of the invention (7) If the calculated divergence value is greater than a second reference value, it is determined that a fingerprint has been detected. Other methods including this method are also well known to those skilled in the art. However, when using the above method, an image can be obtained without having to illuminate a backlight on the image surface, and the user cannot determine whether the fingerprint input device is operating. in. Moreover, it is also possible that the latent fingerprint recognition mechanism lightly recognizes that the fingerprint is not a latent / fingerprint in the absence of an incident external light source, and when the external light is irradiated into the τ when the backlight is illuminated and the fingerprint image is obtained, The fingerprint is mistaken for a biological fingerprint. To solve this problem, another method has been proposed to obtain an image by illuminating the backlight, and when a fingerprint is detected, the backlight is turned off to obtain the image. If a fingerprint is detected in this state, it is recognized as a backlight. Otherwise ', the fingerprint is considered to be a biological fingerprint. In this example, the method for identifying a latent fingerprint includes the following steps: illuminating the backlight on the image surface; when a fingerprint has been detected from the obtained image, the backlight is turned off and an image is obtained, and when the When a fingerprint is detected in the π-off state of the backlight, the detected fingerprint is identified as a latent fingerprint determined by an external light source. In this example, the device for recognizing a latent fingerprint includes: a backlight control mechanism for controlling the on and off of the backlight; and a camera mechanism for controlling the backlight without illuminating the backlight to the image plane. The fingerprint image is obtained under the circumstances; and the fingerprint detection mechanism is used to detect whether a fingerprint exists from the image obtained by the camera_structure; wherein if the fingerprint detection mechanism has detected the fingerprint, the camera mechanism obtains it again. The image,

Page 13 528989 V. Description of the invention (8) Does not illuminate the backlight on the image surface, and if the fingerprint is detected by the camera from the image obtained, the latent fingerprint recognition mechanism recognizes whether a fingerprint is the cause Latent fingerprint detected by an external light source. The method for obtaining an image in the above method and device can be performed by using the method for obtaining an image in the "off" state of the backlight. On the one hand, the method can be used to obtain an image when the backlight is illuminated or not. The entire display or part of the image. You can also use the same method to detect fingerprints from the obtained image to detect whether there is a fingerprint in the obtained image. Also, when the image is obtained, the backlight is turned on. ”Or π turn off π to receive the image, the obtained image includes a part of the image obtained under backlight illumination and a part of the image obtained under backlight η off π state. Using this method, a latent fingerprint is identified by applying each image part of the image. In other words, if the part of the image obtained under backlight illumination detects a fingerprint image, and when the backlight π is turned off π If the part of the obtained image is detected as a fingerprint image, the fingerprint is regarded as a latent fingerprint. In the opposite example, if the part of the image obtained under the backlight is detected as a fingerprint image, and the part of the image obtained under the π-on state of the backlight is detected as having no fingerprint image, then The fingerprint was identified as a biological fingerprint. SUMMARY OF THE INVENTION The features and advantages of the present invention can be further understood from the following description. Best Mode for Carrying Out the Invention

528989 V. Description of the invention (9) The best mode for carrying out the present invention will be described below, please refer to the drawings. In the following description, in order to prevent the reader from reading details not related to the present invention, well-known functions and architectures will not be described in detail. The flowchart of FIG. 4 shows a method for identifying a latent fingerprint in the present invention, and the block diagram of FIG. 12 is a function of the device for identifying a latent fingerprint according to the present invention. For convenience, the method of the present invention is used to illustrate the method. If a person who wishes to perform authentication touches an image surface with one finger ', the backlight control mechanism 503 turns off the backlight [step S 1 0 1]. Then the imaging mechanism 505 obtains the image of the fingerprint [step s103], and the fingerprint detection mechanism ^ 07 detects whether there is a fingerprint [step s105]. As mentioned above, if f is in a state where η is turned off, and a fingerprint is detected, then the identification mechanism 5 009 recognizes whether or not a one-volt fingerprint is detected due to an external light source. Step S 1 0 9]. If any fingerprint is detected by π 1 in the "off" state of the backlight, the latent fingerprint recognition mechanism 50 g determines whether the input fingerprint is a biological fingerprint. [Step S 1 0, 9]. In the procedure for detecting fingerprints in [Step S 1 05], the fingerprint is detected by adding a gray scale as described above. In particular, on a target's X-axis or Y-axis A gray level difference is added between adjacent pixels on the obtained image, as shown in FIG. 5 [step S 1 06]. In the next step, a latent fingerprint recognition mechanism recognizes whether the value to be added is greater or less than A set reference value [step S107]. If the value measured by the debt is large, it means that a fingerprint is detected under the state of "pi-off" of the backlight, and the fingerprint obtained has been identified, It is a latent heart pattern [step S 1 08]. And if the detected value is equal to or smaller

V. Description of the invention (10) A fixed value of y means that no debt is detected ^ '

Furthermore, it is necessary to discriminate the fingerprint scene and ritual image obtained by δ [step S. The flowchart of FIG. 6 is a horse-biological fingerprint according to the present invention. A procedure for identifying the existence of a fingerprint. It is determined by the best mode of Guan Yibei, and all the grayscale values of the pixels of the image are added to the fingerprint image obtained by moonlight. The added value is less than the predetermined reference value, step S 2 0 6]. If [step S 2 08], and the recognition / acquisition, ^ means that the fingerprint image has been detected. The shirt image is a latent fingerprint image. The flowchart in FIG. 7 shows an identified process. The third best mode determines whether there is a pair of guards. Using the "" i, i, ten different image pixels "smoke obtained from the fingerprint according to the invention Shooting value D [Step S 30 6]. If the ten mean values of π and ^ are determined from the sum of the hairs μ — △ Fruit discrimination < Out M is less than a read reference detected fingerprint ~ the best wedge described above The formula indicates that the latent finger 1 is identified by obtaining the entire image of a fingerprint. As mentioned above, in this example, the palladium fee required to receive the image picture is quite large, and only a part of the g pattern is obtained. , And # all fingerprints, in the "off" state of the backlight, to identify a latent fingerprint, so you can save considerable time. Figure 8 is a representative example of obtaining half of the entire image from a latent pattern. As long as about half of the entire image is applied, it is sufficient to recognize the value. 〇?], This means 丨 Zhen measured the finger, the text [y SS 30 8]. Therefore, it is identified that there is a latent element 528989 V. Description of the invention (11) Figure 9 shows the use of an image sensor (·· A part of the latent generation obtained from the latent fingerprint (= raw do; 7 = able From the one-time authentication window fixing function to receive once; silly φ & ^ Tian private sequence stress w becomes narrower, the fingerprint image received by the fingerprint recognition day and day to see myself ^ ^ _, both you Du, with the result that the time spent by the image will become narrower, the time it takes to identify a private pattern of the subordinates also decreases. The flowchart in Figure 10 is based on the respectfulness ^ ηη π & 0 Μ 古$. ^ Identify a potential report according to the best mode of the present invention. 1 ^ Guide, i 1i mechanism 505 through the control of the backlight control mechanism 50 3 / Haiyue first / step S 401] to obtain an image [step s. In the next step, the fingerprint detection mechanism 5 07 recognizes whether or not ^ Wave is detected [step S 40 3]. For the sake of certainty, the camera mechanism 5 05 is turned off under the control of the back control mechanism 50 3. The backlight [step s 404] to obtain two images [step S 40 5]. The fingerprint detection mechanism 507 then recognizes that the backlight is in the off-n state. , Whether a fingerprint has been detected [step s 4 06]. For the sake of certainty, the recognition mechanism 5 09 recognizes that the detected fingerprint is a latent fingerprint [step S 40 7]. Conversely, the recognition mechanism 5 0 9 Determine whether the fingerprint measured by the debt is a biological fingerprint [step S 4 08]. Figures 5 to 7 show the same method for fingerprint detection in the fourth best mode of the present invention, as in Example of the third best mode. In the example of the fourth best mode shown in FIG. 10, an image is obtained to detect whether there is a fingerprint in the state of "on" of the backlight, and when it has been detected, Therefore, by obtaining an image in the "off" state of the backlight to detect the existence of a fingerprint. At the same time that the image is obtained, it is included in the backlight in a similar way to receiving the image by turning on and off the back. Part of the fingerprint is obtained under the "on" condition, and in the "off" state of the backlight

Page 17 528989 V. Description of the invention (12) Part of the fingerprint. Figure 11 is a representative image of the fingerprint image obtained by turning off and on the backlight. The part n Aπ represents the image obtained in the "off" state of the backlight, and the part π Βπ represents the image obtained in the pi-on state of the backlight. As described above, the image obtained from the biological fingerprint can be identified, because the fingerprint is not detected in the part n Aπ, which is obtained under the state that the backlight is turned off, The fingerprint was detected in part 11 Βπ, where this part was obtained under the π-on π state of the backlight.

Those skilled in the art should know that the backlight is turned on and off by software, and that part of the image obtained under the π-on and π-state of the backlight can be calculated to calculate the part obtained under the π-off and π state of the backlight. I identified it in the above. The best mode for carrying out the present invention has been applied above to display and explain the present invention. Although the present invention has been described in the preferred embodiment, those skilled in the art need to understand that the above embodiments can be modified and changed Without departing from the spirit and perspective of the present invention. The viewpoint of the present invention described above is mainly that an image is obtained in the "off π" state of the backlight, and a latent fingerprint is identified from a biological fingerprint based on the fingerprint image detected from the obtained image.

Page 18 528989 Brief description of the drawing Figure 1 is a sound diagram of the operation principle of the 4th pattern input device of the absorption type; Figures 2A and 2B are schematic diagrams of the operation of the absorption type fingerprint input device. A representative pattern of a latent fingerprint; ϊ ΐ 2 ί f is a method for identifying a latent fingerprint in accordance with the present invention; FIG. ^ ^ Is a processing procedure for identifying the existence of a ί pattern according to the first best mode of the present invention; pattern recognition The flow chart of FIG. 6 is a processing procedure according to the second most existing fingerprint of the present invention; FIG. 7 is a processing procedure of identifying whether a heart pattern exists according to the third best mode of the present invention; Sexual pattern of part of the image obtained from the tip; Figure 9, a representative pattern of a part of the image obtained from a latent fingerprint using the windowing function of the image sensor; Figure 10, f light display according to the present invention A flowchart of a method for identifying a latent fingerprint in a fourth best mode; FIG. 11 is a representative pattern of a fingerprint image obtained by interactive switching between light on and off; and FIG. 1 2 square FIG display according to the present invention to identify a functional block diagram of a latent fingerprint device. Drawing number description: 110 three 稜鏡 11 2 backlight

Page 19 528989

Brief description of the drawing 1 14 lens 1 1 6 image sensor 1 1 8 image plane 125 image processor 214 lens 3 1 6 image sensor 318 light penetrates the image plane 5 0 3 backlight control mechanism 5 0 5 camera mechanism 5 0 7 Fingerprint detection mechanism 5 0 9 Identification mechanism

Page 20

Claims (1)

528989 VI. Application Patent Scope 1. A method for identifying a latent fingerprint remaining on the image surface of an optical fingerprint input device. In this method, a backlight is irradiated onto a fingerprint and the light is reflected from the fingerprint. An image of a fingerprint, the method includes the following steps: obtaining an image without illuminating the backlight under the image surface; detecting whether a fingerprint exists from the obtained image; and identifying if detected Whether the fingerprint is a latent fingerprint detected by an external light source. 2. The method of claim 1 in the scope of patent application, wherein the step of obtaining an image is to receive the entire display screen of an image. 3. The method of claim 1 in which the step of obtaining an image is by receiving only a part of an image from the tip of the image. 4. For the method of claim 1 in the scope of patent application, the step of obtaining an image is to use the windowing function of the image sensor, and only receive a part of the image. 5. As in the method of any one of claims 1 to 4, the step of detecting the presence of a fingerprint further includes the following steps: Calculate the neighboring pixels of the image obtained on a target X-axis and Y-axis. The sum of the differences between the gray levels; comparing the calculated total of the gray levels with a predetermined reference value; and if the total value of the gray levels is greater than the reference value, identifying whether a fingerprint has been detected. 6. If the method of any one of claims 1 to 4 of the scope of patent application, the step of detecting the presence of fingerprints further includes the following steps: Page 21 528989 6. The sum of the differences between the gray levels of the corresponding image for the patent application range calculation; comparing the calculated total value of the gray levels with a predetermined reference value; and if the total value of the gray levels is less than At the reference value, it is recognized whether a fingerprint has been detected. 7 · The method of any one of items 1 to 4 of the scope of patent application, wherein the step of detecting the presence of fingerprints further includes the following steps: Calculate the average value and dispersion value of the gray levels corresponding to the obtained image ); If the calculated average value is less than a first reference value and the calculated divergence value is greater than a second reference value, it is determined that a fingerprint has been detected. 8. A method for identifying a latent fingerprint remaining on an image plane of an optical fingerprint input device, in which a fingerprint image is obtained by irradiating a backlight onto a fingerprint and reflecting the light from the fingerprint, the The method includes the following steps: illuminating the backlight under the image plane to obtain an image; detecting the presence of a fingerprint from the obtained image; and turning off the backlight at lf if a fingerprint has been detected π An image is obtained in the state; from the obtained image, it is detected whether a fingerprint exists, and if a fingerprint is detected, it is determined whether the fingerprint is a latent fingerprint detected by an external light source. 9. The method of claim 8 in which the step of obtaining an image is to receive the entire display screen of an image. 1 0. The method according to item 8 of the patent application scope, wherein 1_ Page 22 528989 VI. Scope of patent application The procedure is to receive only a part of an image from the tip of the image. 1 1. The method according to item 8 of the scope of patent application, wherein the step of obtaining an image is to use the windowing function of the image sensor, and only receive a part of an image. 1 2. The method according to any one of items 8 to 11 in the scope of patent application, wherein the step of detecting the existence of a fingerprint further includes the following steps: Calculate the adjacent images obtained on a target X-axis and Y-axis The sum of the differences between the gray levels of the pixels; Comparing the calculated total value of the gray levels with a predetermined reference value; and if the total value of the gray levels is greater than the reference value, identifying whether a fingerprint has been detected. 1 3. The method according to any one of claims 8 to 11 in the scope of patent application, wherein the step of detecting the existence of fingerprints further includes the following steps: calculating a total comparison of differences between gray levels corresponding to the obtained images Calculating the total value of the gray level and a predetermined reference value; and if the total value of the gray level is less than the reference value, identifying whether a fingerprint has been detected. 14. The method of any one of items 8 to 11 of the scope of patent application, wherein the step of detecting the presence of fingerprints further includes the following steps: Calculate the average value and the divergence value of the gray scale corresponding to the obtained image ( dispersion value); and If the calculated average value is less than a first reference value and the calculated divergence value is greater than a second reference value, it is determined that a fingerprint has been detected. 1 5. — A kind of identification bit remains on the image surface of an optical fingerprint input device. Page 23, 528989 6. A method of latent fingerprints with a scope of patent application. In this method, an image of a fingerprint is obtained by irradiating a backlight onto a fingerprint and reflecting the light from the fingerprint. The method includes the following steps: An image is obtained by turning off and turning on the backlight on the image plane; detecting whether there is a fingerprint for the image part obtained under the π-on state of the backlight; for the image obtained under the "off" state of the backlight The image part detects whether a fingerprint exists; and if a fingerprint is detected in the image part obtained in the π-on and π-off states of the backlight, whether the fingerprint is detected by an external light source Measured latent fingerprint. 16. The method according to item 15 of the scope of patent application, wherein the step of obtaining an image is to receive the entire display screen of an image. 17 · The method according to item 15 of the scope of patent application, wherein the step of obtaining an image is by receiving only a part of an image from the tip of the image. 18. The method according to item 15 of the scope of patent application, wherein the step of obtaining an image is to use the windowing function of the image sensor, and only receive a part of an image. 19. The method according to any one of items 15 to 18 in the scope of patent application, wherein the step of detecting the presence of a fingerprint further includes the following steps: Calculate the image phase obtained on a target X-axis and Z-axis. The sum of the differences between the gray levels of neighboring pixels; comparing the calculated total of the gray levels with a predetermined reference value; and if the total value of the gray levels is greater than the reference value, identifying whether a detection has been detected Page 24 528989 6. Scope of patent application A fingerprint. 20 · The method of any one of items 15 to 18 in the scope of patent application, wherein the step of detecting the presence of fingerprints further includes the following steps: Calculate the sum of the differences between the gray levels corresponding to the obtained images Compare the calculated total value of the gray scale with a predetermined reference value; and if the total value of the gray scale is less than the reference value, identify whether a fingerprint has been detected ^ 2 1 The method of any one of clauses, wherein the step of detecting the presence of a fingerprint further comprises the following steps: calculating an average value and a dispersion value of gray levels corresponding to the obtained image; and if the calculated average value is less than If a first reference value and the calculated divergence value are greater than a second reference value, it is determined that a fingerprint has been detected. 2 2. —A device for recognizing a latent fingerprint remaining on the image plane of a fingerprint input device. In the device, an image of a fingerprint is obtained by irradiating a backlight onto a fingerprint and reflecting the light from the fingerprint. The device includes: a backlight control mechanism for controlling the turning on and off of the backlight; a camera mechanism for obtaining an image of a fingerprint without illuminating the backlight on the image plane; a fingerprint detection mechanism for receiving the image from the camera mechanism The presence of a fingerprint is detected in the obtained image; a fingerprint recognition mechanism is used to identify the fingerprint; if the fingerprint detection mechanism detects the fingerprint, it is determined to be detected by an external light source Page 25 528989 6. Scope of patent application Latent fingerprint. 2 3. The device according to item 22 of the scope of patent application, wherein the camera mechanism receives the entire display screen of an image. 24. The device of claim 22, wherein the camera receives a part of an image from the tip of the image. 25. The device according to item 22 of the scope of patent application, wherein the fixed window function of an image sensor is used, and the camera mechanism only receives a part of an image. 26. The device according to any one of claims 22 to 25, wherein the fingerprint detection mechanism includes: a calculation mechanism for summing adjacent pixel grays of images obtained on a target X-axis and Y-axis The difference between steps; a comparison mechanism for comparing the summed value with a predetermined reference value; and a recognition mechanism for identifying whether a fingerprint has been detected if the summed value is less than the predetermined reference value . 27. The device as claimed in any one of claims 22 to 25, wherein the fingerprint detection mechanism comprises: a calculation mechanism for summing grayscale values with respect to the obtained image; a comparison mechanism for comparing the summation And a predetermined reference value; and a recognition mechanism that, if the total value is less than the predetermined reference value, determines whether a fingerprint has been detected. 2 8. The device according to the 22nd to 25th of the scope of patent application, wherein the fingerprint Page 26 528989 6. The patent application range detection mechanism includes: a calculation mechanism for calculating the average value and dispersion value of the gray scale corresponding to the obtained image; and the identification mechanism, if the calculated average value If it is less than a first reference value and the calculated divergence value is greater than a second reference value, it is determined that a fingerprint has been detected. 2 9 · —A device for recognizing a latent fingerprint remaining on the image plane of an optical fingerprint input device. In the device, an image of a fingerprint is obtained by irradiating a backlight onto a fingerprint and reflecting the light from the fingerprint. The device includes: a backlight control mechanism for controlling the on and off of the backlight; a camera mechanism for obtaining an image of a fingerprint under the control of the backlight control mechanism without illuminating the backlight to the image plane; and a fingerprint A detection mechanism is used to detect whether a fingerprint exists from the image obtained by the camera mechanism; if the fingerprint detection mechanism has detected the fingerprint, the camera mechanism obtains the image without illuminating the backlight on the image If a fingerprint is detected by the camera mechanism from the obtained image, the latent fingerprint recognition mechanism identifies whether a fingerprint is a latent fingerprint detected by an external light source. 30. The device as claimed in claim 29, wherein the camera mechanism receives the entire display screen of an image. 31. The device according to item 29 of the patent application, wherein the camera mechanism Page 27 528989 6. Scope of patent application Receive the end and end of the image. 32. The device according to item 29 of the patent application, wherein the windowing function of an image sensor is used, and the camera mechanism only receives a part of an image. 33. The device according to any one of claims 29 to 32 in the patent application scope, wherein the fingerprint detection mechanism includes: a calculation mechanism for summing adjacent pixel grays of images obtained on a target X-axis and Y-axis The difference between steps; a comparison mechanism for comparing the summed value with a predetermined reference value; and a recognition mechanism for identifying whether a fingerprint has been detected if the summed value is less than the predetermined reference value . 3 4 · The device according to any one of claims 29 to 32, wherein the fingerprint detection mechanism includes: a calculation mechanism that adds up the grayscale value with respect to the obtained image; a comparison mechanism for comparison The summed value and a predetermined reference value; and a recognition mechanism, if the summed value is less than the predetermined reference value, whether a fingerprint has been detected. 3 5 · If the device in the range of patent application Nos. 29 to 32, the fingerprint detection mechanism includes: a calculation mechanism for calculating an average value and a dispersion value of gray levels corresponding to the obtained image And identification agencies, if the calculated average is less than a first reference value, Page 28 528989 VI. Patent application and the calculated divergence value is greater than a second reference value, it is deemed that a fingerprint has been detected. 3 6. —A device for recognizing a latent fingerprint remaining on the image plane of an optical fingerprint input device. In the device, an image of a fingerprint is obtained by irradiating a backlight onto a fingerprint and reflecting the light from the fingerprint. The device includes: a backlight control mechanism for controlling the turning on and off of the backlight; and a fingerprint detection mechanism that detects whether the backlight is in the "on n" state and the backlight is in the "pi off" state. There is a fingerprint; and a latent fingerprint recognition mechanism. If fingerprints are detected in the image portion obtained from the “on” state and the π off state of the backlight, the fingerprint obtained is determined to be due to external Latent fingerprint detected by a light source. 3 7 · As in the device under the scope of patent application No. 36, wherein the camera mechanism receives the entire display screen of an image. 3 8. As in the device under the scope of patent application No. 36, where The camera mechanism receives a part of an image from the tip of the image. 3 9 · If the device in the scope of patent application No. 36, uses the window fixing function of an image sensor, The camera mechanism only receives a part of an image. 40 · For the device of any one of the items 36 to 39 of the scope of patent application, the fingerprint detection mechanism includes: a calculation mechanism for summing a target The difference between the gray levels of adjacent pixels of the image obtained on the X-axis and the Y-axis; Page 29 528989 6. The patent application scope comparison mechanism is used to compare the summed value with a predetermined reference value; and the identification mechanism, if the summed value is less than the predetermined reference value, whether the detection has been detected A fingerprint. 41. The device according to any one of claims 36 to 39, wherein the fingerprint detection mechanism comprises: a calculation mechanism for adding a grayscale value to the obtained image; and a comparison mechanism for comparing the summation. And a predetermined reference value; and The identification mechanism determines whether a fingerprint has been detected if the total value is less than a predetermined reference value. 4 2. The device as claimed in claims 36 to 39, wherein the fingerprint detection mechanism comprises: a calculation mechanism for calculating an average value and a dispersion value of gray levels corresponding to the obtained image; and The identification mechanism determines that a fingerprint has been detected if the calculated average value is less than a first reference value and the calculated divergence value is greater than a second reference value. Page 30