A facial recognition system and anesthesia cabinet using the same

Abstract

The present invention belongs to the field of facial recognition technology, and proposes a facial recognition system and an anesthesia cabinet using the system, the system comprising: an image acquisition module, used to obtain facial image data of a user to be stored through facial detection; a feature extraction module, used to obtain a plan view of the facial image of the user to be stored through the facial image data of the user to be stored; a calculation module, used to subdivide the left eye area and the right eye area of ​​the user to be stored through a loop subdivision algorithm; a facial recognition module: used to obtain the facial image data of the user to be detected through facial detection by the recognition module when the user to be detected needs to open the anesthesia cabinet; a judgment module: used to judge whether to open the anesthesia cabinet. According to the facial recognition system of the embodiment of the present invention, human face information can be accurately recognized directly through the eyes in a medical environment.

Description

Technical Field

[0001] The invention belongs to the technical field of facial recognition, and in particular relates to a facial recognition system and an anesthesia cabinet applying the system. Background Art

[0002] At present, due to the many uncertain factors in the medical environment, such as light changes, facial occlusion (masks, face shields), and the diversity of facial expressions of people; traditional recognition often cannot perform accurate recognition in complex medical environments, resulting in medical staff being unable to quickly obtain anesthetics or other key supplies, and requiring unsafe operations such as taking off masks or respirators to complete face recognition, which makes medical staff susceptible to infection and has low operating efficiency. Therefore, a facial recognition method is needed that focuses on the eye area and adopts a segmentation algorithm and multiple verification mechanisms to improve the robustness of recognition and ensure that authorized personnel can still be accurately identified in complex environments. Summary of the invention

[0003] The present invention aims to solve at least one of the technical problems in the related art to a certain extent. To this end, the first object of the present invention is to provide a facial recognition system that can accurately identify human face information directly through the eyes in a medical environment;

[0004] A second object of the present invention is to provide an anesthesia cabinet using a facial recognition system.

[0005] To achieve the above object, a first embodiment of the present invention provides a facial recognition system, the facial recognition system comprising:

[0006] An image acquisition module, used to acquire facial image data of the user to be stored through facial detection;

[0007] A feature extraction module is used to obtain a plane map of the face image of the user to be stored through the face image data of the user to be stored, and circle the left eye and the right eye of the plane map through the ROI region of interest operation to obtain the left eye area and the right eye area;

[0008] The calculation module is used to subdivide the left eye area and the right eye area of ​​the user to be stored by using a loop subdivision algorithm, and calculate the average eye distance of the left eye area of ​​the user to be stored.

[0009] Facial recognition module: used to obtain facial image data of the user to be detected by performing facial detection through the recognition module when the user to be detected needs to open the anesthesia cabinet, and calculate the eye mean distance of the left eye area of ​​the user to be detected based on the facial image data of the user to be detected;

[0010] Judgment module: used to judge whether to open the anesthesia cabinet based on the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

[0011] The facial recognition system implements the following steps to implement facial recognition:

[0012] S100, acquiring facial image data of a user to be stored through facial detection;

[0013] S200, obtaining a plan view of the facial image of the user to be stored through the facial image data of the user to be stored, and circling the left eye and the right eye of the plan view through a ROI region of interest operation to obtain a left eye region and a right eye region;

[0014] S300, subdividing the left eye area and the right eye area of ​​the user to be stored by using a loop subdivision algorithm, and calculating the average eye distance of the left eye area of ​​the user to be stored;

[0015] S400, when the user to be detected needs to open the anesthesia cabinet, facial detection is performed through the recognition module to obtain facial image data of the user to be detected, and the eye mean distance of the left eye area of ​​the user to be detected is calculated according to the facial image data of the user to be detected;

[0016] S500, judging whether to open the anesthesia cabinet by comparing the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

[0017] The facial recognition system according to the present invention implements a facial recognition method that can accurately recognize facial information directly through the eyes in a medical environment.

[0018] Furthermore, in step S100, facial image data of the user to be stored is obtained through facial detection. Specifically, the image of the user is obtained through a camera, and then the image is preprocessed, wherein the preprocessing includes: scaling, graying and denoising to improve the detection accuracy; wherein the scaling includes adjusting the size of the image to fit the standard size of the viewfinder; next, the facial area is located through ROI region of interest recognition, the facial area is cropped from the image, and it is adjusted as needed. Finally, the cropped facial image data is saved or transmitted to the system to prepare for further feature extraction and analysis, wherein the user to be stored is the user who needs to store facial image data so as to open the anesthesia cabinet through the facial image data next time.

[0019] Due to the many uncertainties in the medical environment, such as light changes, facial occlusion (masks, face shields), and the diversity of facial expressions of people; traditional recognition often cannot perform accurate recognition in complex medical environments, resulting in medical staff being unable to quickly obtain anesthetics or other key supplies, and requiring unsafe operations such as taking off masks or respirators to complete face recognition, which makes medical staff susceptible to infection and has low operational efficiency. Therefore, a facial recognition method is needed that focuses on the eye area and uses segmentation algorithms and multiple verification mechanisms to improve the robustness of recognition and ensure that authorized personnel can still be accurately identified in complex environments.

[0020] In order to solve the above problem, the method proposes that in step S200, a plan view of the facial image of the user to be stored is obtained through the facial image data of the user to be stored, and the left eye and the right eye of the plan view are circled through ROI region of interest recognition to obtain the left eye area and the right eye area;

[0021] Specifically, after obtaining the plan view of the facial image, the system will use ROI region of interest recognition technology to locate the eye area. ROI recognition can be based on a predefined model or algorithm, such as a feature point-based method or a deep learning model, to automatically identify and calibrate the main feature areas of the face. When the ROI region of interest is identified and calibrated, the system will further process the plan view to circle the left and right eye areas. The specific operations of circling the left and right eye areas include: drawing or marking the bounding box or area of ​​the eye on the plan view to ensure that subsequent eye analysis can be precisely confined to the eye area; after circling the left and right eye areas, the system will extract and segment these areas from the plan view for subsequent eye feature analysis or other facial recognition operations; the extraction of these areas can be based on predefined geometric shapes or based on actual eye edge detection results.

[0022] S300, subdividing the left eye area and the right eye area of ​​the user to be stored by using a loop subdivision algorithm, and calculating the average eye distance of the left eye area of ​​the user to be stored;

[0023] The left eye area and the right eye area are subdivided [30, 36] times by the loop subdivision algorithm, and the left eye area is subdivided into K1 grids, and S1 = {s1(i1)} represents the set of grids in the left eye area, s1(i1) represents the i1th grid, and i1 takes the value of [1, K1]. The right eye area is subdivided into K2 grids, and s2 = {s2(i2)} represents the set of grids in the right eye area, s2(i2) represents the i2th grid, and i2 takes the value of [1, K2]. The grid where the left eye eye center is located is s1(i3), where i3∈i1, i3 is the serial number of the grid where the left eye eye center is located, and the grid where the right eye eye center is located is s2(i4), i4∈i1, i4 is the serial number of the grid where the right eye eye center is located;

[0024] Let d(i1, i2) be the distance between the i1th grid in the left eye area and the i2th grid in the right eye area, where the distance between the grids is the length of the line segment connecting the center points of the grids. Let the median of all distances from the i1th grid in the left eye area to all grids in the right eye area be dm(i1). Calculate the mean eye distance dx(i1) of the i1th grid in the left eye area of ​​the user to be stored. The calculation method of dx(i1) is as follows:

[0025]

[0026] Where d(i3, i4) is the distance between the grid where the center of the left eye is located and the grid where the center of the right eye is located, and K2 is the number of grids in the right eye area.

[0027] The average eye distances dx(i1) of all the left eye regions of the user to be stored are stored in the facial recognition system;

[0028] The beneficial effect of this step is that by dividing the eye area into multiple small grids and calculating the mean eye distance of each grid in the user's left eye area, the system can capture the slight difference between the left eye and the right eye in more detail. With this increase in precision, the system can effectively distinguish different individuals during facial recognition, even if their facial features are very similar, and by calculating the mean eye distance of multiple grids instead of relying on a single feature point (such as the center of the eyeball), noise interference in the recognition process can be reduced. For example, if the information in a grid is distorted due to light or other factors, the system can still complete the recognition through the accurate information of other grids to avoid affecting the overall recognition result.

[0029] Furthermore, the eye mean distance is calculated for use in comparison. When comparing the facial data of the user to be detected and the user to be stored, the system can determine whether the two are consistent by comparing the eye mean distance of each grid. This method improves the reliability and accuracy of the recognition results. Although dividing the eye area into multiple grids will increase the amount of calculation, using the median and average distance method for calculation can simplify the calculation complexity to a certain extent, and the median's treatment of outliers makes the calculation more robust, rather than being dominated by a few extreme values, thereby improving the overall calculation efficiency.

[0030] S400, when the user to be detected needs to open the anesthesia cabinet, facial detection is performed through the recognition module to obtain facial image data of the user to be detected, and the eye mean distance of the left eye area of ​​the user to be detected is calculated according to the facial image data of the user to be detected;

[0031] The user who needs to open the anesthesia cabinet is recorded as the user to be detected. When the user to be detected needs to open the anesthesia cabinet, the facial image data of the user to be detected is obtained by performing facial detection through the recognition module, and the eye mean distance of the i1th grid in the left eye area of ​​the user to be detected is calculated and recorded as dn(i1), wherein the calculation method of the eye mean distance dn(i1) of the i1th grid in the left eye area is the same as the calculation method of the eye mean distance dx(i1) of the i1th grid in the left eye area of ​​the user to be detected;

[0032] S500, judging whether to open the anesthesia cabinet by comparing the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

[0033] Detect whether the average eye distance in the left eye area of ​​the user to be detected is equal to the average eye distance of the grid with the corresponding sequence number in the left eye area of ​​the user to be stored. If they are all equal, the recognition is successful and the anesthesia cabinet is opened; if they are not all equal, the recognition fails and the anesthesia cabinet is not opened;

[0034] Preferably, simply checking whether the mean eye distances are all equal will make the requirements for face recognition too high, resulting in partial changes in the mean eye distance when the user wears glasses, has slight eye edema caused by sleep problems, or some subtle and unavoidable differences (such as differences caused by lighting, angles, etc.), resulting in failure of face recognition. Therefore, the embodiment of the present invention proposes a preferred method: detect whether the mean eye distance in the left eye area of ​​the user to be detected is equal to the mean eye distance of the grid with the corresponding serial number in the left eye area of ​​the user to be stored. If they are all equal, the recognition is successful and the anesthesia cabinet is opened; if they are not all equal, discrete mean eye distance detection is performed.

[0035] Preferably, the method for detecting the discrete mean interocular distance includes:

[0036] Filtering the grids whose eye mean distance in the left eye region of the user to be detected is not equal to the eye mean distance of the grids with corresponding sequence numbers in the left eye region of the user to be stored;

[0037] S501, calculating the eye distance difference and the discrete eye distance according to the average eye distance in the left eye area of ​​the user to be detected and the average eye distance in the grid with the corresponding sequence number in the left eye area of ​​the user to be stored;

[0038] Create a blank array dg[], compare the value of dn(i1) and the value of dx(i1). If dn(i1)≠dx(i1), add the current value of dn(i1) to dg[]. Traverse from i1=1 to i1=K1 to obtain an array dg[] containing all values ​​whose eye mean distances in the left eye area of ​​the user to be detected are not equal to the eye mean distances of the grids with corresponding numbers in the left eye area of ​​the user to be stored. Record the eye mean distance in dg[] as dn(j), where j∈i1, j is the grid number of the eye mean distance in dg[]. Then record the eye mean distance of the grid of the left eye area of ​​the user to be stored with the corresponding number as dn(j) as dx(j).

[0039] The average value of all dn(j) is recorded as the first reference eye distance kgm, the average value of all dx(j) is recorded as the second reference eye distance xgm, and the discrete eye distance lxd is calculated, where lxd = abs(xgm-kgm), where abs is the absolute value function, that is, lxd is the absolute value of the difference between xgm and kgm.

[0040] The arrays dn(j) and dx(j) are subtracted in turn to obtain the eye distance difference ms(j), that is, ms(j)=abs(dn(j)-dx(j)), where abs is the absolute value function, and the eye distance difference ms(j) is compared with lxd.

[0041] Furthermore, discrete eye distance is a key indicator used to judge the similarity between the user to be detected and the user to be stored in the facial recognition process. It determines an allowable error range by comprehensively analyzing two reference eye distances (the first reference eye distance and the second reference eye distance), and then sets an exact reference error for the recognition process. The discrete eye distance can be used to measure the fault tolerance in facial recognition, that is, the degree of subtle differences that the system can tolerate; the eye distance difference is the absolute difference between the average eye distance of the user to be detected and the user to be stored on the corresponding grids, which is used to measure the similarity between the user to be detected and the user to be stored on each specific grid. By calculating the eye distance difference of each mismatched grid, it can be determined whether the two users are similar enough to be considered as the same person by the system.

[0042] Furthermore, discrete eye distance sets a judgment standard by comprehensively considering the difference in eye distance between the user to be detected and the user to be stored. When the difference in eye distance between the user to be detected is within this discrete eye distance range, the system will consider the two to be a match, thereby passing the identity verification. Through discrete eye distance, the system can tolerate certain subtle and unavoidable differences (such as differences caused by factors such as lighting and angle) without misjudgment. This improves the robustness of the system in practical applications, enabling the system to distinguish between minor differences within the "normal" range and major differences within the "abnormal" range, thereby avoiding misjudgments caused by overly strict or overly loose standards. This allows the system to still accurately identify the correct user when faced with facial features in complex medical environments.

[0043] S502, judging whether the facial image data of the user to be detected is consistent with the facial image data of the user to be stored through the eye distance difference and the discrete eye distance, and if so, opening the anesthesia cabinet;

[0044] If the eye distance difference ms(j) is less than or equal to the discrete eye distance difference lxd, the recognition is successful, the facial image data of the user to be detected is consistent with the facial image data of the user to be stored, and the anesthesia cabinet is opened; if the eye distance difference ms(j) is not less than the discrete eye distance difference lxd, the recognition fails, the facial image data of the user to be detected is not consistent with the facial image data of the user to be stored, and the anesthesia cabinet is not opened.

[0045] The beneficial effect of this step is that the system compares all the eye distance differences with the discrete eye distance. If all the eye distance differences are less than or equal to the discrete eye distance, it means that the difference between the user to be detected and the user to be stored is within the allowable range, and the recognition is successful; otherwise, if all the eye distance differences are not less than or equal to the discrete eye distance, the difference between the user to be detected and the user to be stored is not within the allowable range, and the recognition fails. This multi-level calculation and comparison method can effectively reduce the error in recognition, improve the recognition accuracy and robustness of the system in complex environments, ensure the safety of key operations in medical environments, and ensure the accuracy and reliability of facial recognition.

[0046] The beneficial effects of this method are: through detailed facial area analysis and precise eye mean distance calculation, the recognition accuracy and robustness of the system are improved, identity forgery is prevented, operational efficiency is improved, and the risk of cross-infection is reduced. Its adaptability to complex environments makes its application in medical environments safer, more reliable and more efficient.

[0047] To achieve the above-mentioned purpose, the second embodiment of the present invention further proposes an anesthesia cabinet using a facial recognition system, wherein the anesthesia cabinet is provided with the facial recognition system proposed in the first embodiment.

[0048] The facial recognition system on the anesthesia cabinet can accurately identify facial information directly through the eyes in a medical environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 Shown is a flow chart of a facial recognition method;

[0050] Figure 2 Shown is a diagram of the facial recognition system structure. DETAILED DESCRIPTION

[0051] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0052] Figure 1 Shown is a diagram of the facial recognition system structure.

[0053] Reference Figure 1 The present invention proposes a facial recognition system 20, the facial recognition system comprising:

[0054] An image acquisition module 21 is used to acquire facial image data of a user to be stored through facial detection;

[0055] The feature extraction module 22 is used to obtain a plane map of the facial image of the user to be stored through the facial image data of the user to be stored, and circle the left eye and the right eye of the plane map through the ROI region of interest operation to obtain the left eye area and the right eye area;

[0056] The calculation module 23 is used to subdivide the left eye area and the right eye area of ​​the to-be-stored user by using a loop subdivision algorithm, and calculate the average eye distance of the left eye area of ​​the to-be-stored user.

[0057] Facial recognition module 24: used for performing facial detection through the recognition module to obtain facial image data of the user to be detected when the user to be detected needs to open the anesthesia cabinet, and calculating the eye mean distance of the left eye area of ​​the user to be detected according to the facial image data of the user to be detected;

[0058] The judgment module 25 is used to judge whether to open the anesthesia cabinet according to the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

[0059] Figure 2 Shown is a flow chart of a facial recognition method.

[0060] Reference Figure 2 , the facial recognition system implements facial recognition method steps as follows:

[0061] S100, acquiring facial image data of a user to be stored through facial detection;

[0062] S200, obtaining a plan view of the facial image of the user to be stored through the facial image data of the user to be stored, and circling the left eye and the right eye of the plan view through a ROI region of interest operation to obtain a left eye region and a right eye region;

[0063] S300, subdividing the left eye area and the right eye area of ​​the user to be stored by using a loop subdivision algorithm, and calculating the average eye distance of the left eye area of ​​the user to be stored;

[0064] S400, when the user to be detected needs to open the anesthesia cabinet, facial detection is performed through the recognition module to obtain facial image data of the user to be detected, and the eye mean distance of the left eye area of ​​the user to be detected is calculated according to the facial image data of the user to be detected;

[0065] S500, judging whether to open the anesthesia cabinet by comparing the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

[0066] The facial recognition system according to the present invention implements a facial recognition method that can accurately recognize facial information directly through the eyes in a medical environment.

[0067] Furthermore, in step S100, facial image data of the user to be stored is obtained through facial detection. Specifically, the user's image is obtained through a camera, and then the image is preprocessed, wherein the preprocessing includes: scaling, graying and denoising to improve the detection accuracy; next, the facial area is located through ROI region of interest recognition, the facial area is cropped from the image, and it is adjusted as needed. Finally, the cropped facial image data is saved or transmitted to the system to prepare for further feature extraction and analysis, wherein the user to be stored is the user who needs to store facial image data in order to open the anesthesia cabinet through the facial image data next time.

[0068] Due to the many uncertainties in the medical environment, such as light changes, facial occlusion (masks, face shields), and the diversity of facial expressions of people; traditional recognition often cannot perform accurate recognition in complex medical environments, resulting in medical staff being unable to quickly obtain anesthetics or other key supplies, and requiring unsafe operations such as taking off masks or respirators to complete face recognition, which makes medical staff susceptible to infection and has low operational efficiency. Therefore, a facial recognition method is needed that focuses on the eye area and uses segmentation algorithms and multiple verification mechanisms to improve the robustness of recognition and ensure that authorized personnel can still be accurately identified in complex environments.

[0069] In order to solve the above problem, the method proposes that in step S200, a plan view of the facial image of the user to be stored is obtained through the facial image data of the user to be stored, and the left eye and the right eye of the plan view are circled through ROI region of interest recognition to obtain the left eye area and the right eye area;

[0070] S300, subdividing the left eye area and the right eye area of ​​the user to be stored by using a loop subdivision algorithm, and calculating the average eye distance of the left eye area of ​​the user to be stored;

[0071] The left eye area and the right eye area are subdivided [30, 36] times by the loop subdivision algorithm, and the left eye area is subdivided into K1 grids, and S1 = {s1 (i1)} represents the set of grids in the left eye area, s1 (i1) represents the i1-th grid, and i1 takes the value of [1, K1]. The right eye area is subdivided into K2 grids, and s2 = {s2 (i2)} represents the set of grids in the right eye area, s2 (i2) represents the i2-th grid, and i2 takes the value of [1, K2]. The grid where the left eye eye center is located is s1 (i3), where i3 ∈ i1, i3 is the serial number of the grid where the left eye eye center is located, and the grid where the right eye eye center is located is s2 (i4), i4 ∈ i1, i4 is the serial number of the grid where the right eye eye center is located;

[0072] Let d(i1, i2) be the distance between the i1th grid in the left eye area and the i2th grid in the right eye area, where the distance between the grids is the length of the line segment connecting the center points of the grids. Let the median of all distances from the i1th grid in the left eye area to all grids in the right eye area be dm(i1). Calculate the mean eye distance dx(i1) of the i1th grid in the left eye area of ​​the user to be stored. The calculation method of dx(i1) is as follows:

[0073]

[0074] Where d(i3, i4) is the distance between the grid where the center of the left eye is located and the grid where the center of the right eye is located, and K2 is the number of grids in the right eye area.

[0075] The average eye distances dx(i1) of all the left eye regions of the user to be stored are stored in the facial recognition system;

[0076] The beneficial effect of this step is that by dividing the eye area into multiple small grids and calculating the mean eye distance of each grid in the user's left eye area, the system can capture the slight difference between the left eye and the right eye in more detail. With this increase in precision, the system can effectively distinguish different individuals during facial recognition, even if their facial features are very similar, and by calculating the mean eye distance of multiple grids instead of relying on a single feature point (such as the center of the eyeball), noise interference in the recognition process can be reduced. For example, if the information in a grid is distorted due to light or other factors, the system can still complete the recognition through the accurate information of other grids to avoid affecting the overall recognition result.

[0077] Furthermore, the eye mean distance is calculated for use in comparison. When comparing the facial data of the user to be detected and the user to be stored, the system can determine whether the two are consistent by comparing the eye mean distance of each grid. This method improves the reliability and accuracy of the recognition results. Although dividing the eye area into multiple grids will increase the amount of calculation, using the median and average distance method for calculation can simplify the calculation complexity to a certain extent, and the median's treatment of outliers makes the calculation more robust, rather than being dominated by a few extreme values, thereby improving the overall calculation efficiency.

[0078] S400, when the user to be detected needs to open the anesthesia cabinet, facial detection is performed through the recognition module to obtain facial image data of the user to be detected, and the eye mean distance of the left eye area of ​​the user to be detected is calculated according to the facial image data of the user to be detected;

[0079] The user who needs to open the anesthesia cabinet is recorded as the user to be detected. When the user to be detected needs to open the anesthesia cabinet, the facial image data of the user to be detected is obtained by performing facial detection through the recognition module, and the eye mean distance of the i1th grid in the left eye area of ​​the user to be detected is calculated and recorded as dn(i1), wherein the calculation method of the eye mean distance dn(i1) of the i1th grid in the left eye area is the same as the calculation method of the eye mean distance dx(i1) of the i1th grid in the left eye area of ​​the user to be detected;

[0080] S500, judging whether to open the anesthesia cabinet by comparing the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

[0081] Detect whether the average eye distance in the left eye area of ​​the user to be detected is equal to the average eye distance of the grid with the corresponding sequence number in the left eye area of ​​the user to be stored. If they are all equal, the recognition is successful and the anesthesia cabinet is opened; if they are not all equal, the recognition fails and the anesthesia cabinet is not opened;

[0082] Preferably, simply checking whether the mean eye distances are all equal will make the requirements for face recognition too high, resulting in partial changes in the mean eye distance when the user wears glasses, has slight eye edema caused by sleep problems, or some subtle and unavoidable differences (such as differences caused by lighting, angles, etc.), resulting in failure of face recognition. Therefore, the embodiment of the present invention proposes a preferred method: detect whether the mean eye distance in the left eye area of ​​the user to be detected is equal to the mean eye distance of the grid with the corresponding serial number in the left eye area of ​​the user to be stored. If they are all equal, the recognition is successful and the anesthesia cabinet is opened; if they are not all equal, discrete mean eye distance detection is performed.

[0083] Preferably, the method for detecting the discrete mean interocular distance includes:

[0084] Filtering the grids whose eye mean distance in the left eye region of the user to be detected is not equal to the eye mean distance of the grids with corresponding sequence numbers in the left eye region of the user to be stored;

[0085] S501, calculating the eye distance difference and the discrete eye distance according to the average eye distance in the left eye area of ​​the user to be detected and the average eye distance in the grid with the corresponding sequence number in the left eye area of ​​the user to be stored;

[0086] Create a blank array dg[], compare the value of dn(i1) and the value of dx(i1). If dn(i1)≠dx(i1), add the current value of dn(i1) to dg[]. Traverse from i1=1 to i1=K1 to obtain an array dg[] containing all values ​​whose eye mean distances in the left eye area of ​​the user to be detected are not equal to the eye mean distances of the grids with corresponding numbers in the left eye area of ​​the user to be stored. Record the eye mean distance in dg[] as dn(j), where j∈i1, j is the grid number of the eye mean distance in dg[]. Then record the eye mean distance of the grid of the left eye area of ​​the user to be stored with the corresponding number as dn(j) as dx(j).

[0087] The average value of all dn(j) is recorded as the first reference eye distance kgm, the average value of all dx(j) is recorded as the second reference eye distance xgm, and the discrete eye distance lxd is calculated, where lxd = abs(xgm-kgm), where abs is the absolute value function, that is, lxd is the absolute value of the difference between xgm and kgm.

[0088] The arrays dn(j) and dx(j) are subtracted in turn to obtain the eye distance difference ms(j), that is, ms(j)=abs(dn(j)-dx(j)), where abs is the absolute value function, and the eye distance difference ms(j) is compared with lxd.

[0089] Furthermore, discrete eye distance is a key indicator used to judge the similarity between the user to be detected and the user to be stored in the facial recognition process. It determines an allowable error range by comprehensively analyzing two reference eye distances (the first reference eye distance and the second reference eye distance), and then sets an exact reference error for the recognition process. The discrete eye distance can be used to measure the fault tolerance in facial recognition, that is, the degree of subtle differences that the system can tolerate; the eye distance difference is the absolute difference between the average eye distance of the user to be detected and the user to be stored on the corresponding grids, which is used to measure the similarity between the user to be detected and the user to be stored on each specific grid. By calculating the eye distance difference of each mismatched grid, it can be determined whether the two users are similar enough to be considered as the same person by the system.

[0090] Furthermore, discrete eye distance sets a judgment standard by comprehensively considering the difference in eye distance between the user to be detected and the user to be stored. When the difference in eye distance between the user to be detected is within this discrete eye distance range, the system will consider the two to be a match, thereby passing the identity verification. Through discrete eye distance, the system can tolerate certain subtle and unavoidable differences (such as differences caused by factors such as lighting and angle) without misjudgment. This improves the robustness of the system in practical applications, enabling the system to distinguish between minor differences within the "normal" range and major differences within the "abnormal" range, thereby avoiding misjudgments caused by overly strict or overly loose standards. This allows the system to still accurately identify the correct user when faced with facial features in complex medical environments.

[0091] S502, judging whether the facial image data of the user to be detected is consistent with the facial image data of the user to be stored through the eye distance difference and the discrete eye distance, and if so, opening the anesthesia cabinet;

[0092] If the eye distance difference ms(j) is less than or equal to the discrete eye distance difference lxd, the recognition is successful, the facial image data of the user to be detected is consistent with the facial image data of the user to be stored, and the anesthesia cabinet is opened; if the eye distance difference ms(j) is not less than the discrete eye distance difference lxd, the recognition fails, the facial image data of the user to be detected is not consistent with the facial image data of the user to be stored, and the anesthesia cabinet is not opened.

[0093] The beneficial effect of this step is that the system compares all the eye distance differences with the discrete eye distance. If all the eye distance differences are less than or equal to the discrete eye distance, it means that the difference between the user to be detected and the user to be stored is within the allowable range, and the recognition is successful; otherwise, if all the eye distance differences are not less than or equal to the discrete eye distance, the difference between the user to be detected and the user to be stored is not within the allowable range, and the recognition fails. This multi-level calculation and comparison method can effectively reduce the error in recognition, improve the recognition accuracy and robustness of the system in complex environments, ensure the safety of key operations in medical environments, and ensure the accuracy and reliability of facial recognition.

[0094] The beneficial effects of this method are: through detailed facial area analysis and precise eye mean distance calculation, the recognition accuracy and robustness of the system are improved, identity forgery is prevented, operational efficiency is improved, and the risk of cross-infection is reduced. Its adaptability to complex environments makes its application in medical environments safer, more reliable and more efficient.

[0095] The second aspect of the present invention further provides an anesthesia cabinet using a facial recognition system, wherein the anesthesia cabinet is provided with the facial recognition system provided in the first aspect of the present invention.

[0096] The facial recognition system on the anesthesia cabinet can accurately identify facial information directly through the eyes in a medical environment.

[0097] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection portion with one or more wirings (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0098] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0099] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0100] In the description of the present invention, it is to be understood that the terms “center”, “longitudinal”, “lateral”, “length”, “width”, “thickness”, “up”, “down”, “front”, “back”, “left”, “right”, “vertical”, “horizontal”, “top”, “bottom”, “inside”, “outside”, “clockwise”, “counterclockwise”, “axial”, “radial”, “circumferential”, etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0101] In addition, the terms "first", "second", etc. used in the embodiments of the present invention are only used for descriptive purposes and should not be understood as indicating or implying relative importance, or implicitly indicating the number of technical features indicated in the present embodiment. Therefore, the features defined by the terms "first", "second", etc. in the embodiments of the present invention can explicitly or implicitly indicate that the embodiment includes at least one of the features. In the description of the present invention, the word "multiple" means at least two or two or more, such as two, three, four, etc., unless otherwise clearly and specifically defined in the embodiments.

[0102] In the present invention, unless otherwise clearly specified or limited in the embodiments, the terms "installed", "connected", "connected" and "fixed" etc. in the embodiments should be understood in a broad sense. For example, the connection can be a fixed connection, a detachable connection, or an integrated connection. It can be understood that it can also be a mechanical connection, an electrical connection, etc.; of course, it can also be a direct connection, or an indirect connection through an intermediate medium, or it can be the internal connection of two elements, or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to the specific implementation situation.

[0103] In the present invention, unless otherwise clearly specified and limited, a first feature being "above" or "below" a second feature may mean that the first and second features are in direct contact, or the first and second features are in indirect contact through an intermediate medium. Moreover, a first feature being "above", "above" or "above" a second feature may mean that the first feature is directly above or obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. A first feature being "below", "below" or "below" a second feature may mean that the first feature is directly below or obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0104] Although the embodiments of the present invention have been shown and described above, it is to be understood that the above embodiments are exemplary and are not to be construed as limitations of the present invention. A person skilled in the art may change, modify, replace and vary the above embodiments within the scope of the present invention.

Claims

1. A facial recognition system, characterized in that: The facial recognition system comprises: An image acquisition module, used to acquire facial image data of the user to be stored through facial detection; A feature extraction module is used to obtain a plane map of the face image of the user to be stored through the face image data of the user to be stored, and circle the left eye and the right eye of the plane map through the ROI region of interest operation to obtain the left eye area and the right eye area; A calculation module, used to subdivide the left eye area and the right eye area of ​​the user to be stored by a loop subdivision algorithm, and calculate the average eye distance of the left eye area of ​​the user to be stored; Facial recognition module: when the user to be detected needs to open the anesthesia cabinet, the facial recognition module is used to obtain the facial image data of the user to be detected, and the eye mean distance of the left eye area of ​​the user to be detected is calculated according to the facial image data of the user to be detected; wherein, the method for calculating the eye mean distance of the left eye area of ​​the user to be detected is: The left eye area and the right eye area are subdivided [30, 36] times by the loop subdivision algorithm, and the left eye area is subdivided into K1 grids, and S1={s1(i1)} represents the set of grids in the left eye area, s1(i1) represents the i1th grid, and i1 takes the value of [1, K1]. The right eye area is subdivided into K2 grids, and s2={s2(i2)} represents the set of grids in the right eye area, s2(i2) represents the i2th grid, and i2 takes the value of [1, K2]. The grid where the left eye eyeball center is located is s1(i3), where i3∈i1, i3 is the serial number of the grid where the left eye eyeball center is located, and the grid where the right eye eyeball center is located is s2(i4), i4∈i1, i4 is the serial number of the grid where the right eye eyeball center is located; Let d(i1, i2) be the distance between the i1th grid in the left eye area and the i2th grid in the right eye area, where the distance between the grids is the length of the line segment connecting the center points of the grids. Let the median of all distances from the i1th grid in the left eye area to all grids in the right eye area be dm(i1). Calculate the mean eye distance dx(i1) of the i1th grid in the left eye area of ​​the user to be stored. The calculation method of dx(i1) is as follows: ； Where d(i3, i4) is the distance between the grid where the center of the left eye is located and the grid where the center of the right eye is located, and K2 is the number of grids in the right eye area; The average eye distances dx(i1) of all the left eye regions of the user to be stored are stored in the facial recognition system; Judgment module: used to judge whether to open the anesthesia cabinet based on the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

2. A facial recognition system according to claim 1, characterized in that: The specific steps of the method for realizing facial recognition by a facial recognition system are as follows: S100, acquiring facial image data of a user to be stored through facial detection; S200, obtaining a plan view of the facial image of the user to be stored through the facial image data of the user to be stored, and circling the left eye and the right eye of the plan view through a ROI region of interest operation to obtain a left eye region and a right eye region; S300, subdividing the left eye area and the right eye area of ​​the user to be stored by using a loop subdivision algorithm, and calculating the average eye distance of the left eye area of ​​the user to be stored; S400, when the user to be detected needs to open the anesthesia cabinet, facial detection is performed through the recognition module to obtain facial image data of the user to be detected, and the eye mean distance of the left eye area of ​​the user to be detected is calculated according to the facial image data of the user to be detected; S500, judging whether to open the anesthesia cabinet by comparing the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored.

3. A facial recognition system according to claim 2, characterized in that: In step S100, the method for obtaining facial image data of the user to be stored through facial detection is: obtaining the image of the user through a camera, and preprocessing the image, wherein the preprocessing includes: scaling, graying and denoising to improve detection accuracy; locating the facial area through ROI region of interest recognition, cropping the facial area from the image, and adjusting it as needed; saving or transmitting the cropped facial image data to the system to prepare for further feature extraction and analysis, wherein the user to be stored is a user who needs to store facial image data so as to open the anesthesia cabinet through the facial image data next time.

4. A facial recognition system according to claim 1, characterized in that: In S400, when the user to be detected needs to open the anesthesia cabinet, facial detection is performed by the recognition module to obtain facial image data of the user to be detected, and the method for calculating the eye mean distance of the left eye area of ​​the user to be detected according to the facial image data of the user to be detected includes: The user who needs to open the anesthesia cabinet is referred to as the user to be detected. When the user to be detected needs to open the anesthesia cabinet, facial detection is performed through the recognition module to obtain the facial image data of the user to be detected, and the eye mean distance of the i1th grid in the left eye area of ​​the user to be detected is calculated and recorded as dn(i1), wherein the calculation method of the eye mean distance dn(i1) of the i1th grid in the left eye area is the same as the calculation method of the eye mean distance dx(i1) of the i1th grid in the left eye area of ​​the user to be detected.

5. A facial recognition system according to claim 4, characterized in that: In S500, the method for determining whether to open the anesthesia cabinet by using the average eye distance of the left eye area of ​​the user to be detected and the average eye distance of the left eye area of ​​the user to be stored is: Check whether the average eye distance in the left eye area of ​​the user to be detected is equal to the average eye distance of the grid with the corresponding serial number in the left eye area of ​​the user to be stored. If all are equal, the recognition is successful and the anesthesia cabinet is opened; if not all are equal, perform discrete average eye distance detection.

6. A facial recognition system according to claim 5, characterized in that: The specific steps of discrete mean interocular distance detection are: S501, calculating the eye distance difference and the discrete eye distance according to the average eye distance in the left eye area of ​​the user to be detected and the average eye distance in the grid with the corresponding sequence number in the left eye area of ​​the user to be stored; S502, judging whether the facial image data of the user to be detected is consistent with the facial image data of the user to be stored through the eye distance difference and the discrete eye distance, and if it is consistent, opening the anesthesia cabinet.

7. An anesthesia cabinet using a facial recognition system, characterized in that: The anesthesia cabinet is provided with a facial recognition system according to any one of claims 1 to 6.

Images