Quick operating system login method and system

Abstract

The invention relates to a quick operating system login method and system, which comprises the following steps of: after a computer system is started, entering an operating system login program, driving an infrared light source of a camera system to be turned on and turned off, and particularly driving the infrared light source to be turned off for the first time; an infrared light camera of the camera system is driven to shoot a user to generate continuous frames consisting of continuous dark frames and bright frames, and a first dark frame in the continuous frames is obtained, namely, a scene is judged according to brightness distribution of the first dark frame, so that exposure setting corresponding to the scene can be obtained, and the exposure setting is a product of exposure time and gain. In the operating system login program, the user is shot according to the obtained exposure setting so as to obtain a biological image of the user, and the biological image can be used for executing a biological image recognition program to log in the operating system.

Description

Technical Field

[0001] The specification discloses a method for logging into an operating system, and more particularly, a method and system for quickly logging into an operating system by predicting a scene through brightness distribution of a frame image to obtain an appropriate exposure setting. Background Art

[0002] Windows Hello is Microsoft The company released Windows 10 in 2015 ( 10) is a biometric recognition technology introduced. Windows Hello login uses biometrics (fingerprint, facial, or iris) to verify your identity, replacing traditional password login methods. It's both convenient and more secure. Facial recognition is the most widely used Windows Hello login method. In this method, the user points their face toward a camera or infrared depth camera. The operating system then scans and identifies the user's facial features, comparing them to a pre-set pattern to enable login. Compared to the other two login methods, fingerprint recognition requires an additional fingerprint reader, while iris recognition also requires an iris scanner. While iris recognition has the lowest false acceptance rate (FAR), it can be affected by wearing glasses or pupil dilation lenses. The FAR refers to the probability that the biometric system mistakenly identifies an unauthorized user as a legitimate one.

[0003] Even though Windows Hello's facial recognition technology is convenient and accurate in most cases, there are still some difficulties, especially the ambient light will affect the accuracy of facial recognition. Figure 1 Example graphs showing auto-exposure convergence curves for scenes under different ambient lighting conditions.

[0004] Figure 1The graph shows the brightness convergence of the existing automatic exposure program in consecutive frames of outdoor, backlight, front light and indoor scenes. As shown in the front light automatic exposure convergence curve 105, when the user logs into the operating system in the front light scene, sunlight or a specific light source will directly illuminate the user's face. Around the 8th frame, the exposure brightness can be converged to a certain level (based on brightness 100, brightness can be taken as Y value (Luminance) for example) and the login is completed; as shown in the backlight automatic exposure convergence curve 103, when the user is in the backlight scene, the exposure brightness can be converged to a certain level (based on brightness 100, brightness can be taken as Y value (Luminance) for example) and the login is completed. When logging into the operating system in a backlight scenario, light will enter from behind the user and directly hit the camera, causing the facial image captured by the camera to be overexposed or underexposed. In this case, most of the user's facial features will be obscured, and the camera's automatic exposure convergence must be completed. This takes a relatively long time for login, as the operating system must recognize the correct facial features before login can proceed normally. As shown in the backlight automatic exposure convergence curve 103, login is completed around the 14th frame. As shown in the outdoor automatic exposure convergence curve 101, because outdoor light has a high infrared component, the time required for automatic exposure convergence when capturing the user's image using the infrared camera is relatively affected. In this example, automatic exposure convergence is completed around the 9th frame. The indoor automatic exposure convergence curve 107 shows that because the indoor light is relatively simple and has less infrared light influence, automatic exposure convergence can be achieved quickly. In this example, a stable automatic exposure convergence curve is already present from the beginning. Summary of the Invention

[0005] The disclosure document proposes a method and system for quickly logging into an operating system. When used in a computer system using biometric recognition technology, it can effectively eliminate the problem of infrared light components in specific light sources (such as outdoor light sources outside windows, sunlight, etc.) interfering with the effectiveness of biometric image recognition, and can be applied to devices with lower computing power.

[0006] According to an embodiment of the rapid operating system login method, after the computer system is started, an operating system login program is entered, and then the light source of the camera system is driven to illuminate the user, and the camera of the camera system is driven to photograph the user to obtain continuous frames. Then, the scene is determined based on the brightness distribution of the first dark frame in the continuous frames, and the exposure setting corresponding to this scene is obtained. Therefore, the user can be photographed according to the exposure setting to obtain the user's biometric image, and then a biometric image recognition program is executed based on this biometric image to log in to the operating system.

[0007] Preferably, the light source is an infrared light source, and the camera is an infrared camera, and is used in a camera system, wherein the running camera system executes an automatic exposure program, a scene prediction program, a user position prediction program, and a scene weight calculation program.

[0008] Furthermore, after starting the camera system, the infrared light source is driven to be turned on and off, and the infrared light source is driven to be turned off for the first time. The infrared camera is used to capture continuous frames consisting of continuous dark frames and bright frames, and the first dark frame in the continuous frames is obtained. That is, the scene is judged based on the brightness distribution of this first dark frame.

[0009] In an embodiment of the scene determination process, a first dark frame is divided into multiple blocks. Luminance statistics are collected for each block to obtain brightness statistics for multiple central blocks and multiple corner blocks. The brightness statistics for the multiple central blocks and the multiple corner blocks are then compared with multiple thresholds measured based on actual scenes. Based on the comparison results with the thresholds for different scenes, the scene can be predicted as outdoor, backlit, front-lit, or indoor.

[0010] Furthermore, the method also includes predicting the position of the user in the first dark frame to accurately obtain the brightness value of the user's face, so that the appropriate exposure setting can be determined based on the frame brightness distribution and the brightness value of the user's face. When the scene is judged to be backlit, the position with the lowest brightness statistical value in the first dark frame is the user's position; when the scene is judged to be face-lit, the position with the highest brightness statistical value in the first dark frame is the user's position.

[0011] Furthermore, the brightness weight values ​​between different blocks are determined according to the determined scene, and the exposure setting is adjusted to obtain a biological image that is more suitable for the scene, thereby achieving the purpose of quickly logging into the operating system.

[0012] To further understand the features and technical contents of the present invention, please refer to the following detailed description and drawings of the present invention. However, the drawings provided are only for reference and illustration and are not intended to limit the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 Shows examples of the existing automatic exposure convergence curves for scenes under different ambient light conditions;

[0014] Figure 2 An example diagram showing a scenario in which a quick operating system login method is used;

[0015] Figure 3 A diagram showing an embodiment of a firmware program in a camera system that executes a fast operating system login method;

[0016] Figure 4 A flowchart showing an embodiment of a method for quickly logging into an operating system;

[0017] Figure 5 A schematic diagram showing the generation of consecutive frames in a fast operating system login method;

[0018] Figure 6 Figure showing an embodiment of performing scene prediction by frame image;

[0019] Figure 7 Figure showing an embodiment of predicting scene;

[0020] Figure 8 Figures (A) (B) (C) (D) (E) (F) showing user position offset;

[0021] Figure 9 Figure showing an example of auto exposure convergence curve using fast operating system login method; and

[0022] Figure 10 Figures (A) (B) showing user images taken before and after using fast operating system login method. DETAILED DESCRIPTION

[0023] The present application will now be described in greater detail by way of specific embodiments only which should not be construed as restricting the scope of the present application. The present application can be variously embodied and implemented, and thus the following embodiments are merely provided to more fully describe the present application to those skilled in the art. Various modifications made to the embodiments of the present application based on the principles of the present application can be derived by those skilled in the art and the embodiments should not be construed as limiting the scope of the present application. In addition, the drawings accompanying the present application are simple schematic illustrations and not actual depictions, and it is hereby declared in advance. The following embodiments will further describe the technical contents of the present application in detail, but the disclosed contents are not intended to limit the scope of the present application.

[0024] It should be understood that although the terms "first", "second", "third", etc. can be used herein to describe various elements or signals, these elements or signals should not be limited by these terms. These terms are mainly used to distinguish one element from another element, or one signal from another signal. In addition, the term "or" used herein can include any one or more combinations of the associated listed items.

[0025] To address the issue of varying ambient light in different scenarios potentially affecting the speed of automatic exposure convergence when logging into an operating system using biometric image recognition technology, the disclosure proposes a method and system for rapid operating system login. According to an embodiment of the rapid operating system login method, the method utilizes an infrared camera (IR camera) to obtain an image of a user's biometric features (e.g., facial features) and employs a rapid automatic exposure algorithm. This allows the camera system executing the method to effectively reduce the time required for automatic exposure convergence when using the infrared camera under varying light conditions (including front light and backlight), thereby increasing the speed of logging into the operating system and improving the user experience. The operating system encompasses various systems requiring user identity verification for entry, such as computer operating systems.

[0026] For example, a computer operating system that uses the fast operating system login method using facial image recognition technology can refer to Figure 2 A diagram showing an embodiment of a scenario in which a fast operating system login method is used.

[0027] Computer system 24 runs an operating system (OS). After booting up and waiting for user 20 to log in, it activates the camera system's light source 26 to assist in illuminating user 20 and activates the camera system's camera 22 to capture and detect whether anyone is in front of the computer system. Upon detecting user 20 in front of the computer system, computer system 24 immediately initiates the operating system login process and activates the camera system's camera 22 to capture the user's face. The captured images are then provided to computer system 24 for facial recognition. After facial image feature comparison, the user can successfully log into computer system 24.

[0028] The computer system 24 may refer to any system that can perform a login process through biometric image recognition. Figure 3 The diagram shows that the main processing circuitry of computer system 24 includes a processor 241 and memory 243, which are used to run an operating system 30. Computer system 24 includes the operating system, a camera system 25 consisting of a camera 22 and a light source 26, and an interface unit 245. In another embodiment, camera system 25 can be an external system, connected to computer system 24 via a specific connection. Programs running in operating system 30 primarily include a driver 301 for driving camera 22 and light source 26 of camera system 25, a biometric image recognition program 305 for acquiring biometric images and performing identity recognition, and an operating system login program 307.

[0029] Furthermore, the camera system 25 can execute an automatic exposure program 303, which may also include a scene prediction program 309, a user position prediction program 311, and a scene weight calculation program 313. The scene prediction program 309 can determine the scene currently logged into the operating system to determine an exposure setting appropriate for that scene. The user position prediction program 311 is used to determine whether the user is in a front-lit or backlit scene, so as to correctly utilize the highlight information in the user's biometric image to obtain an appropriate exposure setting. Finally, the scene weight calculation program 313 is used to obtain a weight value corresponding to each scene. The purpose of calculating the weight for each scene is to adaptively determine the product of exposure time and gain appropriate for that scene.

[0030] According to one embodiment, the camera 22 is an infrared camera (or a color camera) for capturing continuous frame images. The light source 26 is an infrared light source (e.g., an IR LED). The computer system 24, via a driver 301, drives the light source 26 of the camera system 25 to turn on or off, allowing the camera 22 to continuously capture bright and dark frames. For example, using an infrared light source, the computer system 24 captures continuous infrared bright frames and infrared dark frames. The bright frames contain energy reflected from infrared light directed at an object (e.g., a user or their face) as well as infrared light energy received from the outside, while the dark frames contain only infrared light energy received from the outside.

[0031] When the operating system login program (such as When the infrared camera's bright and dark frame pair (Hello) 307 receives the infrared camera's bright and dark frame pair, the operating system 30 also performs a subtraction process on the bright and dark frames to obtain a subtracted frame containing only the energy reflected by the infrared light directed at an object (such as a user or their face). The biometric image recognition program 305 executed by the operating system 30 then compares the feature information obtained from the subtracted frame with the biometric features pre-registered in the operating system 30 and stored in the memory 243 to determine whether the currently obtained user biometric feature (such as a face) matches the pre-registered biometric feature. Once the match is confirmed, the user logs into the computer system 24 through the operating system login program 307.

[0032] It is worth mentioning that in addition to the infrared camera (including the infrared light sensor), the camera system can also use another color camera to obtain a color image of the biometric feature and obtain color information of the biometric feature. The color information can be used for other purposes, such as implementing facial anti-spoofing detection. This detection is a process after the user's biometric features are identified using the infrared camera. That is, after completing the biometric feature recognition, facial anti-spoofing detection is performed.

[0033] Compared to the existing biometric image recognition technology used to log into the system, which is affected by ambient light and requires a longer automatic exposure convergence time, the fast operating system login method proposed in the disclosure uses the dark frame of the first infrared light frame (also known as the dark frame) in the continuous biometric image captured by the camera as the basis for scene detection, and predicts the exposure gain value based on the image information of this first dark frame. Figure 4 The flowchart of the embodiment of the fast operating system login method is shown, and reference is made to Figure 5 Schematic diagram of generating continuous frames in the fast operating system login method.

[0034] according to Figure 4 The process shown in FIG401 illustrates a method for rapidly logging into the operating system after the computer system is started. After the computer boots up and its peripherals are activated, the operating system login process begins. This involves a driver program driving the camera system's light source to illuminate the user and controlling the infrared light source to rapidly turn on and off (step S401).

[0035] For reference Figure 5 A schematic diagram of an embodiment of generating continuous frames is shown. When the fast operating system login method is executed, the computer system drives the infrared light source of the camera system to turn on or off according to the high and low levels of the infrared light source driving clock 503, and drives the camera of the camera system to shoot the user to obtain a biological image, forming a continuous infrared light frame pair consisting of dark and bright frames (step S403). The continuous frame pair includes a continuously changing dark frame and a bright frame, such as Figure 5 The continuous infrared light frame 501 is shown. The continuous infrared light frame 501 is composed of continuous infrared light dark frames (IR darkframe) and infrared light bright frames (IR brightframe).

[0036] In particular, when the light source is activated at the same time as the camera is triggered, the light source is first controlled to be turned off, so that the camera first takes a first dark frame 511, and then a first bright frame 512, forming a dark-bright-dark-bright frame pair, and the first dark frame 511 is used for scene prediction. The scene prediction program is used to identify the scene as a user front light or back light, and an indoor or outdoor scene, and the actual implementation is not limited to these scenes (step S405). Then, the exposure setting corresponding to each scene, i.e., the product of the exposure time and the gain, can be obtained. The scene prediction process can refer to the embodiment shown in Figure 6 and Figure 7 and equation one.

[0037] However, according to one of the embodiments, in order to more accurately expose, the process can further include predicting the user position for the front light and back light scenes by using a user position prediction program (step S407). One of the purposes is to accurately obtain the brightness value of the user's face according to the user's position, so that the camera system can simultaneously determine the appropriate exposure setting according to the brightness distribution of the first dark frame and the brightness value of the user's face. After the scene prediction and user position prediction are completed, the brightness ratio between the corner and the central part of the frame image can be further obtained, and the brightness ratio of the corner and the central part is used as the weight value corresponding to the scene (step S409) to adjust a more appropriate exposure setting. Then, the exposure setting obtained according to the scene, i.e., the product of the exposure time and the gain, or a more appropriate exposure setting can be obtained for the user position, and the exposure weight is used to adjust the exposure time and the gain suitable for this scene, and the product of the exposure time and the gain (EtGain) is calculated (step S411). The exposure time and the gain calculated from the first dark frame 511 are effective in the second bright frame 513 in the second frame pair (step S413), that is, the operating system can obtain a biometric image with correct exposure when receiving the second frame pair to perform the program for logging into the operating system (step S415).

[0038] According to the embodiment of the fast operating system login method, in addition to using scene prediction to effectively speed up the operating system login time, the user position prediction and the algorithm for calculating the scene weight are also used.

[0039] Scene prediction:

[0040] Further description of the implementation of step S405 of scene prediction in the above process can refer to the embodiment of performing scene prediction by the first dark frame shown in Figure 6 and the embodiment of the first dark frame shown in Figure 7Embodiment flowchart for displaying predicted scene.

[0041] Figure 6 A first dark frame 60 generated from a scene taken by a camera, such as an infrared light image, is shown. The first dark frame 60 is divided into a plurality of blocks (such as 5x5=25) according to image size, and the brightness statistics of each block is calculated, respectively. Thus, a plurality of brightness statistics (such as 25 brightness statistics) are obtained according to the number of blocks (step S701). A certain number (such as 7) of brightness averages can be selected from the brightness statistics to calculate the average brightness representing the whole frame.

[0042] The brightness distribution of the first dark frame 60, including the average brightness values of the corner and center blocks, is obtained by using Equation One (step S703), which is used to describe the brightness characteristics of the first dark frame 60. In Equation One, "I" represents the infrared light image, "Corner" represents the corner block, "Center" represents the center block, "UL" represents the upper left direction, "UR" represents the upper right direction, "DL" represents the down left direction, and "DR" represents the down right direction.

[0043] According to the legend shown in Figure 6 Figure 1, "Corner UL" in the first dark frame 60 represents the upper left corner block, such as block 0, 1, 5 and 10 shown in Figure 6 Figure 1; "Corner UR" represents the upper right corner block, such as block 3, 4, 9 and 14 shown in Figure 6 Figure 1; "Corner DL" represents the down left corner block, such as block 15 and 20 shown in Figure 6 Figure 1; and "Corner DR" represents the down right corner block, such as block 19 and 24 shown in Figure 6 Figure 1. "Center" represents the center blocks 7, 12, 17 and 22; "Center L" represents the left center blocks 6, 11, 16 and 21; and "Center R" represents the right center blocks 8, 13, 18 and 23.

[0044] Equation One:

[0045] Corner UL:(I[0]+I[1]+I[5]+I

[10] ) / 4;

[0046] Corner UR:(I[3]+I[4]+I[9]+I

[14] ) / 4;

[0047] Corner DL:(I

[15] +I

[20] ) / 2;

[0048] Corner DR:(I

[19] +I

[24] ) / 2;

[0049] Center:(I[7]+I

[12] +I

[17] +I

[22] ) / 4;

[0050] Center L:(I[6]+I

[11] +I

[16] +I

[21] ) / 4;

[0051] Center R:(I[8]+I

[13] +I

[18] +I

[23] ) / 4.

[0052] Next, the brightness statistics of a certain number (e.g., 7) of central blocks and the brightness statistics of corner blocks are used (step S705) and compared with the threshold values ​​measured using actual scenes (for outdoor, backlight, front light, and indoor scenes). Figure 6 The diagram in the figure shows a human figure in the frame image. The light distribution of the human face and its surroundings in different scenes can be used to set thresholds, such as the first threshold, second threshold, and third threshold in the process. Different thresholds are set according to the brightness distribution characteristics of the entire frame in different scenes. The threshold for each scene can be composed of the distribution values ​​of multiple blocks measured in the scene. Based on the comparison results with the thresholds for different scenes, it can be predicted whether the scene where the user logs into the operating system is outdoors, backlit, front-lit, or indoors.

[0053] exist Figure 7 During the display process, after obtaining the brightness statistics of the center and corner blocks of the first dark frame, a determination is made as to whether the brightness statistics of any corner block (Corner UL, Corner UR, Corner DL, or Corner DR) are greater than a first threshold (step S707). If no corner block's brightness statistics are greater than the first threshold (no), a determination is then made as to whether the brightness statistics of any center block (Center, Center L, or Center R) are greater than a third threshold (step S709). If any center block's brightness statistics are greater than the third threshold (yes), a surface-lit scene is predicted (step S711). If no center block's brightness statistics are greater than the third threshold (no), an indoor scene is predicted (step S713).

[0054] Returning to step S707 of the process, if it is determined that the brightness statistics of any corner block are greater than the first threshold (yes), it is then determined whether the brightness statistics of any center block (Center, Center L, or Center R) are greater than the second threshold (step S715). If the brightness statistics of any center block are greater than the second threshold (yes), an outdoor scene is predicted (step S717). If the brightness statistics of no center block are greater than the second threshold (no), a backlit scene is predicted (step S719).

[0055] For example, taking the infrared brightness statistics of each block of an entire frame as an example, if both the central block and the corner block are greater than the set threshold, it means that the entire infrared image is bright, and the user is judged to be in an outdoor scene. If only the corner block is greater than the system-set threshold, while the central block is not greater than the threshold, it is judged that the user is facing away from the light source and may be in a backlit scene.

[0056] User location prediction:

[0057] When using biometric image (such as face) recognition technology to log into an operating system, the user must be in a position where the system can recognize their biometric features, for example, the user's face must be facing the camera so that the system can capture the biometric features. However, in reality, the position of the user in the image captured by the camera may change rather than being fixed, and may be offset to the left or right. Figure 8 Schematic diagrams of user position offset shown in (A) to (F) of the disclosure. In order to more effectively utilize the highlight information in the user's biometric image to predict appropriate exposure settings, such as the brightness of the user's face, the fast operating system login method proposed in the disclosure book adopts an algorithm that can predict the user's position in backlight and frontlight scenarios, thereby determining whether the user's position is to the left, center, or right.

[0058] According to one embodiment, the algorithm for predicting the user's location is as shown in Equation 2, which includes equations (1) and (2). In equation (1), "P backlight ” and “P in formula (2) frontlight ” are the predicted user positions in backlit and frontlit scenarios, respectively. The argmin function is used to obtain the minimum variable value.

[0059] Equation 2:

[0060] P backlight =argmin(Center_L,Center,Center_R);

[0061] P frontlight =argmax(Center_L,Center,Center_R).

[0062] When the scene is judged to be backlit, taking facial feature recognition as an example, in a backlit scene, the brightness statistics of multiple blocks (corners or center) distinguished by using the dark frame (preferably the first dark frame) will have a lower brightness statistics value for the user's face. Therefore, the argmin function is used to find the position with the lowest brightness statistics value in the dark frame to estimate the user's actual location. For reference, Figure 8 The diagrams of (A) (B) (C) show that the user is in a backlit scene. On the contrary, when the scene is judged to be in a front-lit scene, the argmax function is used to find the position with the highest brightness statistics in the dark frame, which is also used to estimate the user's actual position. Figure 8 Schematic diagram of the user in the surface light scene in the images shown in (D)(E)(F).

[0063] It's worth noting that the algorithm for predicting user location is suitable for scenarios where the user is in front-lit or backlit environments, but not for outdoor or indoor environments. This is because in these scenarios, infrared dark frames appear almost entirely white or black, making them ineffective for predicting the user's location. In the user location prediction step, because the brightness levels of the left, center, and right positions are similar, in one embodiment, only the center position is used as the default.

[0064] Calculate scene weights:

[0065] After determining the scene or predicting the user's position, the weight value corresponding to the scene will be calculated based on the ratio between the corner block and the central block, as shown in Equation 3, including equations (3)(4)(5)(6). backlight " is the weight value of the backlight scene. Taking the backlight scene as an example, since the brightness statistics of the image background are higher in backlight, and the brightness statistics of the user's face are lower, the brightness statistics of the brightest corner block and the predicted brightness statistics of the user's position are used to calculate the ratio of the brightness statistics of the brightest corner block to the user's position. This ratio is used as the weight value of the backlight scene to calculate a value that can reflect the intensity of the backlight scene;" W frontlight " is the weight value for the front-lit scene. When the scene is judged to be front-lit, the calculation method is opposite to that of the back-lit scene. That is, because the user's facial brightness statistics are higher in the front-lit scene, the ratio of the brightness statistics of the darkest corner block and the user's position is calculated. This ratio is used as the weight value of the front-lit scene to calculate a value that can reflect the intensity of the front-lit scene;" W outdoor" is the weight value of the outdoor scene. When the scene is judged to be outdoor, since the overall brightness statistics of outdoor scenes are higher, the weight is calculated based on the ratio of the brightness statistics of the middle part of the user's position to the desired brightness (the brightness statistics preset by the system);" W indoor " is the weight value for indoor scenes. When the scene is judged to be indoors, since most of the time when logging into the system using biometric image recognition technology, it is in indoor scenes, the product of the preset exposure time and gain is already the optimal value for indoor scenes, so the weight value is set to "1" and there is no need to change the optimal preset value.

[0066] Equation 3:

[0067] W backlight =max(Corner) / Ymean(P backlight );

[0068] W frontlight =min(Corner) / Ymean(P frontlight );

[0069] W outdoor =Center / Ymean Target;

[0070] W indoor =1.

[0071] Then, during the calculation process, according to Equation 4, the weight value "W*" calculated by Equation 3 as the scene changes is multiplied by the product of the exposure time and gain of the preset indoor scene, thereby obtaining the product of the exposure time and gain suitable for the scene, as shown in Equation 4 (7). In this way, when the operating system receives the second frame pair from the infrared camera, the product of the suitable exposure time and gain takes effect, and this product is multiplied by the weight value obtained for each scene in the above embodiment to obtain the exposure setting suitable for the scene, thereby obtaining a biometric image with an appropriate exposure value. The operating system can then use this user biometric image (such as face) for identity recognition, which can effectively improve the login speed.

[0072] Equation 4:

[0073] EtGain predict =W*×EtGain indoor .

[0074] According to the above-mentioned fast operating system login method embodiment, the process of obtaining the automatic exposure value of the user's biological image can be effectively converged. Figure 9 The following is an example of an automatic exposure convergence curve diagram for outdoor scenes. The curve diagram shows the change in the average brightness value of consecutive frames of images taken in outdoor scenes, where the following are marked: Figure 1As shown, the original outdoor automatic exposure convergence curve 101 before executing the fast operating system login method proposed in the disclosure book is that the automatic exposure convergence is completed at about the 10th frame to obtain a biometric image that can be used to identify the identity; by executing the proposed fast operating system login method and obtaining the exposure value of the corresponding scene (the product of exposure time and gain), the purpose of quickly obtaining the automatic exposure value can be achieved, and the improved automatic exposure convergence curve 901 is obtained, and the login process can be completed at about the 4th frame.

[0075] Figure 10 (A) (B) shows a schematic diagram of the user image captured before and after the rapid operating system login method is used, wherein Figure 10 (A) shows the continuous frame images of the user captured before the fast operating system login method proposed in the prior art is implemented. The user image gradually becomes the clearest image at the 10th frame through the automatic exposure value convergence process from a white piece. Figure 10 (B) shows that after executing the fast operating system login method, the appropriate exposure value can be quickly obtained, that is, the automatic exposure value can be quickly converged. This example shows that the third frame can obtain a clear image that can be used for biological image recognition technology.

[0076] In summary, the fast operating system login method and system described in the above embodiments is different from the conventional operating system login process, which requires waiting for the infrared camera's automatic exposure convergence to complete before using a frame pair with more appropriate brightness to execute the login procedure, which takes a long time. The method proposed in the disclosure uses the first dark frame for scene prediction, and can make the exposure information corresponding to the predicted scene effective in the second bright frame, effectively shortening the automatic exposure convergence time, thereby achieving the purpose of fast operating system login.

[0077] The contents disclosed above are only preferred feasible embodiments of the present invention and do not limit the scope of the patent application of the present invention. Therefore, all equivalent technical changes made using the contents of the description and drawings of the present invention are included in the scope of the patent application of the present invention.

[0078]

Explanation of symbols

[0079] 101: Outdoor Auto Exposure Convergence Curve

[0080] 103: Backlight automatic exposure convergence curve

[0081] 105: Front light automatic exposure convergence curve

[0082] 107: Indoor automatic exposure convergence curve

[0083] 24: Computer System

[0084] 20: User

[0085] 22: Camera

[0086] 25: Camera System

[0087] 26: Light Source

[0088] 241:Processor

[0089] 243: Memory

[0090] 245: Interface unit

[0091] 30: Operating System

[0092] 301: Driver

[0093] 303: Automatic exposure program

[0094] 305: Biological Image Recognition Program

[0095] 307: Operating system login program

[0096] 309: Scenario Prediction Program

[0097] 311: User Location Prediction Program

[0098] 313: Scene weight calculation program

[0099] 501: continuous infrared light frames

[0100] 503: Infrared light source driving clock

[0101] 511: First dark frame

[0102] 512: First bright frame

[0103] 513: Second bright frame

[0104] 60: First dark frame

[0105] 901: Improved auto exposure convergence curve

[0106] 0,2,…24: blocks

[0107] Steps S401 to S415: Quick login system process

[0108] Steps S701 to S719 are scene prediction procedures.

Claims

1. A method for quickly logging into an operating system, comprising: After starting a computer system, enter an operating system login program; photographing a user to obtain continuous frames; Determine a scene according to the brightness distribution of the first dark frame in the continuous frames; Determine an exposure setting corresponding to the scene; photographing the user according to the exposure setting to obtain a biological image of the user; and A biometric image recognition program is executed using the biometric image, and the operating system is logged in.

2. The fast operating system login method according to claim 1, wherein: After starting a camera system, an infrared light source is driven to turn on and off, and a continuous frame consisting of continuous dark frames and bright frames is captured by an infrared camera. The first dark frame in the continuous frame is obtained, and the scene is determined based on the brightness distribution of the first dark frame.

3. The fast operating system login method according to claim 2, wherein: When the infrared light source is started, the infrared light source is driven to be turned off for the first time so that the infrared camera obtains the first dark frame when taking pictures.

4. The method for quickly logging into an operating system according to claim 3, wherein the process of determining the scenario comprises: Dividing the first dark frame area into a plurality of blocks, and performing brightness statistics on each block; Obtaining brightness statistics of a plurality of central blocks and brightness statistics of a plurality of corner blocks; comparing the brightness statistics of the plurality of central blocks and the brightness statistics of the plurality of corner blocks with a plurality of thresholds measured in an actual scene; as well as According to the comparison results with the thresholds of different scenes, it is predicted that the scene is outdoor, backlit, front-lit or indoor.

5. The fast operating system login method according to claim 4, further comprising predicting the position of the user in the first dark frame to accurately obtain the brightness value of the user's face, so as to determine the appropriate exposure setting based on both the frame brightness distribution and the brightness value of the user's face.

6. The fast operating system login method according to claim 5, wherein: In the step of predicting the user's position in the frame, when the scene is determined to be backlit, the position with the lowest brightness statistical value in the first dark frame is used as the user's position; when the scene is determined to be front-lit, the position with the highest brightness statistical value in the first dark frame is used as the user's position.

7. The fast operating system login method according to claim 5, wherein: When determining the scene based on the brightness distribution of the frame, the brightness ratio between the corners and the center of the frame is obtained as a weight value corresponding to the scene. The weight value is used to adjust the exposure setting to be more appropriate, wherein the exposure setting is the product of the exposure time and gain suitable for the scene, multiplied by the weight value of the scene to obtain the exposure setting suitable for the scene.

8. The method for quickly logging into an operating system according to claim 7, wherein: When the scene is determined to be backlit, the user's facial brightness statistics are low, and the brightness statistics of the brightest corner block and the brightness statistics of the user's position are used to calculate the ratio of the brightness statistics of the brightest corner block to the brightness statistics of the user's position as the weight value of the scene; When the scene is determined to be face-lit, the user's facial brightness statistics are higher, and the ratio of the brightness statistics of the darkest corner block to the brightness statistics of the user's position is used as the weight value of the scene; When the scene is determined to be outdoor, the ratio between the brightness statistical value at the user's location and a preset brightness statistical value is used as the weight value of the scene; as well as When the scene is judged to be indoor, the weight value of the scene is set to 1.

9. A system for executing a fast operating system login method, comprising: a computer system externally connected to or encompassing a camera system comprising a camera and a light source; The computer system runs an operating system, and the fast operating system login method is executed, including: After starting the computer system, enter an operating system login program; driving the light source of the camera system to illuminate a user, and driving the camera of the camera system to photograph the user to obtain continuous frames; Determine a scene according to the brightness distribution of a first dark frame in the continuous frames; Determine an exposure setting corresponding to the scene; photographing the user according to the exposure setting to obtain a biological image of the user; and A biometric image recognition program is executed using the biometric image, and the computer system is logged in.

10. The system of claim 9, wherein the light source is an infrared light source, the camera is an infrared camera, and the computer system runs the operating system to drive the light source and the camera of the camera system through a driver program, and the camera system further executes an automatic exposure program including a scene prediction program, a user position prediction program, and a scene weight calculation program; in, After the operating system is activated, the infrared light source of the camera system is driven to turn on and off, and the infrared light source is driven to be turned off for the first time. The infrared camera is used to capture a series of frames consisting of dark frames and bright frames, and the first dark frame in the series of frames is obtained. That is, the scene is determined based on the brightness distribution of the first dark frame. The process of determining the scene includes: Dividing the first dark frame area into a plurality of blocks, and performing brightness statistics on each block; Obtaining brightness statistics of a plurality of central blocks and brightness statistics of a plurality of corner blocks; comparing the brightness statistics of the plurality of central blocks and the brightness statistics of the plurality of corner blocks with a plurality of thresholds measured in an actual scene; and According to the comparison results with the thresholds of different scenes, it is predicted that the scene is outdoor, backlit, front-lit or indoor.

Images