Eyeball recognition access control system based on neural network

Abstract

The invention discloses an eyeball recognition access control system based on a neural network. Comprising an information storage module, an identity recognition module, an image acquisition module, an eyeball information processing and analysis module, a control module, a communication and alarm module, an illumination acquisition module, a sunlight simulation module, a distance sensing module, a processing and calculation module, an infrared thermal imaging module and a mobile phone application end, the beneficial effects of the invention are that whether the eyeball information of the identified person is the same as the eyeball information of the resident granted with the door opening authority is judged by calculating the length of the connecting line of the inner canthus vertex and the outer canthus vertex and the included angle between the connecting line and the horizontal line, and the requirement on an acquisition lens is low, so that the problems of high equipment cost and low cost are solved; according to the technical scheme of the invention, the method achieves the light supplement through collecting the environment illumination intensity, achieves the light supplement, enables the light supplement angle to be adjusted according to the distance between a to-be-recognized person and a collection device, enables the image collected by a lens to be clear, and avoids the wrong recognition.

Description

Technical Field

[0001] The present invention relates to the technical field of eyeball recognition access control technology, in particular to an eyeball recognition access control system based on a neural network. Background Art

[0002] Access control refers to the prohibition authority of the "door", which is a kind of precaution against the "door". The "door" here, in a broad sense, refers to various passages that can be passed, including doors for people and cars, etc. The entrance and exit access control security management system is a new modern security management system. It integrates microcomputer automatic recognition technology and modern security management measures, involving many new technologies such as electronics, machinery, optics, computer technology, communication technology and biotechnology. It is an effective measure to solve the security prevention management of entrances and exits of important departments, and is suitable for various places. Eye recognition technology is a biometric identification technology based on the eye patterns on both sides of the user's iris for identity recognition. It is a new biometric identification technology developed after face recognition and fingerprint.

[0003] Existing eye recognition technology is based on the unique characteristics of the iris, that is, the information contained in each person's iris is different, and the possibility of having iris tissue with exactly the same shape is much lower than that of other tissues. However, iris recognition still has many shortcomings, such as the size of the equipment for acquiring images is too large and it is extremely difficult to optimize the size. The equipment for iris recognition is expensive, making it difficult to promote. In addition, the lens for collecting irises may produce image distortion, thereby affecting the reliability of iris recognition. The clarity of the image collected by the lens is also related to the light intensity. If the light intensity is not enough, incorrect recognition may occur.

[0004] Based on the above problems, it is urgent to propose an eye recognition access control system based on a neural network. By calculating the length of the line connecting the inner canthus vertex and the outer canthus vertex and the angle between the line and the horizontal line, it is determined whether the eye information of the current person being identified is the same as the eye information of the resident who is granted the right to open the door. The requirements for the lens for collecting images are low, thereby solving the problems of high equipment cost and difficulty in promotion. In addition, by collecting the light intensity in the current environment and comparing it with the light intensity threshold, corresponding light supplementation is performed, and the angle of the light supplement is adjusted according to the distance between the person to be identified and the collection equipment, so that the image collected by the lens is clear, avoiding the occurrence of incorrect recognition. Summary of the invention

[0005] The purpose of the present invention is to provide an eye recognition access control system based on a neural network to solve the problems raised in the above background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] An eyeball recognition access control system based on a neural network comprises an information storage module, an identity recognition module, an image acquisition module, an eyeball information processing and analysis module, a control module, a communication and alarm module, an illumination acquisition module, a daylight simulation module, a distance sensing module, a processing and calculation module, an infrared thermal imaging module and a mobile phone application terminal, wherein the information storage module pre-stores the eyeball information of a first resident who is granted door opening authority, the image acquisition module is used to acquire the eye image of a person who is currently undergoing eyeball information recognition, the eyeball information processing and analysis module is used to process and analyze the eye image of a person who is currently undergoing eyeball information recognition, and the identity recognition module compares the processing and analysis result of the eyeball information processing and analysis module with the eyeball information of the resident pre-stored in the information storage module to confirm the current The control module determines whether the person who is undergoing eye information recognition in front of the door has the authority to open the door, and decides whether to open the door or to alarm according to the recognition result of the identity recognition module. The control module is also used to control the daylight simulation module. The control module alarms the residents who are granted the right to open the door and are pre-stored in the information storage module through the communication and alarm module. The illumination acquisition module is used to collect the light intensity in a certain area in front of the door. The daylight simulation module is used to supplement the lighting. The distance sensing module is used to obtain the distance between the person who is undergoing eye information recognition in front of the door and the door. The processing and calculation module is used to process and calculate the data obtained by the distance sensing module. The infrared thermal imaging module is used to detect whether there is a person to be identified in front of the door. The mobile phone application end is used to receive the alarm information sent by the communication and alarm module.

[0008] Furthermore, the information storage module pre-stores the eye information of the first resident who is granted the right to open the door, including the length of the eye canthus distance of the resident's eyes, the angle between the line connecting the inner canthus vertices and the outer canthus vertices and the horizontal line. The length of the inner canthus distance of each person is different, and combined with the difference in eye shape, the different eye shapes will cause each person's eyes to be tilted to different degrees relative to the horizontal line, that is, the angle between the line connecting the inner canthus vertices and the outer canthus vertices and the horizontal line is slightly different. If only the length of the inner canthus distance or the tilt angle is used, misidentification may occur. However, combining them together to determine whether the eye information of the identified person is consistent with the pre-stored eye information of the resident who is granted the right to open the door can avoid misidentification and have low requirements for the collection equipment, thereby solving the problems of high equipment cost and difficult promotion.

[0009] Furthermore, the eye information processing and analysis module connects the inner canthus vertex and the outer canthus vertex of the acquired eye image, where the connecting line is a first straight line, and establishes a first rectangular coordinate system with the inner canthus vertex as the origin of the coordinate axis. The horizontal axis direction of the first rectangular coordinate system is the same as the direction from the inner canthus vertex to the outer canthus vertex. The first angle θ between the first straight line and the horizontal axis and the length L of the eye angle distance are calculated, where the length of the eye angle distance is the length of the first straight line. The calculated first angle θ and the length L of the eye angle distance are compared and analyzed with the eye information of the first resident who is granted the door opening authority pre-stored in the information storage module. Through the double comparison of the first straight line and the first angle θ, it is determined whether the eye information of the currently identified person is consistent with the eye information of the resident who is granted the door opening authority, thereby realizing the access control authority management.

[0010] Furthermore, during the comparison and analysis, the eye information processing and analysis module first compares the difference between the eye corner lengths of the identified person and the first resident. When the difference is less than or equal to the threshold, the eye information processing and analysis module performs the next step of comparison and analysis on whether the first angle θ between the first straight line of the identified person and the first resident and the horizontal axis of the first right-angle coordinate system are equal. The eye information processing and analysis module outputs the comparison and analysis result to the control module. If the comparison and analysis result received by the control module is that the eye information of the current identified person matches that of the first resident, the control door is controlled to open. If the comparison and analysis result is that the eye information of the current identified person does not match that of the first resident, an alarm message is sent to the mobile phone application of the first resident through the communication and alarm module. A connection is established between the communication and alarm module and the resident, so that the resident can know in real time whether the eye access control system is being illegally invaded and can take precautions in advance.

[0011] Furthermore, the distance sensing module obtains the distance S between the person who is undergoing eye recognition in front of the door and the door, the daylight simulation module and the image acquisition module are both arranged on the door, the daylight simulation module is located directly above the image acquisition module, the distance between the daylight simulation module and the image acquisition module is H, the eyes of the currently identified person and the image acquisition module are located on the same horizontal line, the processing and calculation module establishes a second rectangular coordinate system with the horizontal line as the horizontal axis of the coordinate, the line connecting the image acquisition module and the eyes of the currently identified person is a first straight line, the processing and calculation module calculates a second angle ɑ between the first straight line and the horizontal axis of the second rectangular coordinate system, and the second angle The processing and calculation module outputs the calculation results to the daylight simulation module, and adjusts the irradiation angle of the supplementary light of the daylight simulation module by the second angle ɑ. In an environment with insufficient light intensity, the image captured by the image acquisition device can be clear enough, thereby avoiding erroneous recognition.

[0012] Furthermore, the daylight simulation module is a light supplement device embedded in the surface of the door and can move freely up and down to adjust the angle of the illumination beam. The daylight simulation module adjusts the illumination angle of the supplementary light according to the second angle ɑ. The illumination angle of the supplementary light, i.e., the second angle ɑ, is adjusted to the optimal illumination angle so that the captured image is clear.

[0013] Furthermore, the illumination acquisition module is used to collect the light intensity in the current environment. If the light intensity is less than or equal to the light intensity threshold, the light supplement is enabled through the daylight simulation module.

[0014] Furthermore, the infrared thermal imaging module detects whether there is a person to be identified in front of the door. The judgment criteria for the person to be identified are the time the person stays in front of the door and the third angle β between the second straight line and the first plane. The two shoulder peak points of the image of the person in front of the door are selected, and the two shoulder peak points include abdominus point A and abdominus point B. The second straight line is the line connecting abdominus point A and abdominus point B. The first plane is the plane where the door is located. When the infrared thermal imaging module detects that there is a person staying in front of the door, the timing starts. When the stay time of the person in front of the door exceeds time t, the infrared thermal imaging module puts the person in the queue of people to be identified. The first judgment of the person to be identified is performed based on the stay time of the person in front of the door. This is to prevent the person passing by the door from being regarded as the person to be identified, thereby triggering the acquisition device to perform image acquisition.

[0015] Furthermore, the processing and calculation module calculates the third angle β, that is, calculates the angle between the line connecting the acromion points of the two shoulders and the first plane, draws a perpendicular line through the acromion point A perpendicular to the first plane, and the foot of the perpendicular is C, draws a perpendicular line through the acromion point D perpendicular to the first plane, and the foot of the perpendicular is D, and connects the foot of the perpendicular C and the foot of the perpendicular D. The third angle β is the angle between the second straight line and the straight line CD. When the second straight line is perpendicular to the first plane, the third angle β is 90 degrees. In daily life, non-persons to be identified may stay in front of the door. Therefore, it is possible to judge whether this person is a person to be identified based on the angle between the current person staying and the first plane where the door is located. Because when performing eye recognition, the eyes must be aligned with the acquisition device, and the angle between the person's body and the first plane must be within an angle range, so that the acquisition device can capture the person's eye image. Therefore, it is extremely reliable to judge whether the current person staying is a person to be identified based on the third angle.

[0016] Furthermore, the processing and calculation module calculates a third angle deviation, and the third angle deviation Among them, β 0is the third angle threshold. If the calculated third angle deviation p is less than or equal to the threshold, the person currently staying in front of the door is determined to be a person to be identified. The image acquisition module acquires the eye image of the person to be identified. If the calculated third angle deviation p is greater than the threshold, the person currently staying in front of the door is determined to be a person not to be identified.

[0017] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: the present invention calculates the length of the line connecting the vertex of the inner canthus angle and the vertex of the outer canthus angle and the angle between the line and the horizontal line to determine whether the eye information of the current person being identified is the same as the eye information of the resident who is granted the right to open the door. The present invention has low requirements for the lens for collecting images, thereby solving the problems of high equipment cost and difficulty in promotion. In addition, by collecting the light intensity in the current environment and comparing it with the light intensity threshold, corresponding light supplementation is performed, and the angle of the light supplement is adjusted according to the distance between the person to be identified and the collection device, so that the image collected by the lens is clear, avoiding the occurrence of incorrect recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0019] Figure 1 It is a module schematic diagram of a neural network-based eye recognition access control system of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] See also Figure 1 , the present invention provides a technical solution:

[0022] An eyeball recognition access control system based on a neural network comprises an information storage module, an identity recognition module, an image acquisition module, an eyeball information processing and analysis module, a control module, a communication and alarm module, an illumination acquisition module, a daylight simulation module, a distance sensing module, a processing and calculation module, an infrared thermal imaging module and a mobile phone application terminal, wherein the information storage module pre-stores the eyeball information of a first resident who is granted door opening authority, the image acquisition module is used to acquire the eye image of a person who is currently undergoing eyeball information recognition, the eyeball information processing and analysis module is used to process and analyze the eye image of a person who is currently undergoing eyeball information recognition, and the identity recognition module compares the processing and analysis results of the eyeball information processing and analysis module with the eyeball information of the resident pre-stored in the information storage module. The result is used to confirm whether the person currently undergoing eye information recognition has the authority to open the door. The control module decides whether to open the door or to alarm based on the recognition result of the identity recognition module. The control module is also used to control the daylight simulation module. The control module alarms the residents who are granted door opening authority and pre-stored in the information storage module through the communication and alarm module. The illumination acquisition module is used to collect the light intensity in a certain area in front of the door. The daylight simulation module is used to supplement the lighting. The distance sensing module is used to obtain the distance between the person who is undergoing eye information recognition in front of the door and the door. The processing and calculation module is used to process and calculate the data obtained by the distance sensing module. The infrared thermal imaging module is used to detect whether there is a person to be identified in front of the door. The mobile phone application end is used to receive the alarm information sent by the communication and alarm module.

[0023] The eyeball information of the first resident who is granted the door opening permission and pre-stored in the information storage module includes the length of the eye canthus of the resident's eyes, and the angle between the line connecting the inner canthus vertex and the outer canthus vertex and the horizontal line.

[0024] The eye information processing and analysis module connects the inner canthus vertex and the outer canthus vertex of the acquired eye image, and the connecting line is the first straight line. A first rectangular coordinate system is established with the inner canthus vertex as the origin of the coordinate axis. The horizontal axis direction of the first rectangular coordinate system is the same as the direction from the inner canthus vertex to the outer canthus vertex. The first angle θ between the first straight line and the horizontal axis and the length L of the eye canthus distance are calculated. The length of the eye canthus distance is the length of the first straight line. The calculated first angle θ and the length L of the eye canthus distance are compared and analyzed with the eye information of the first resident who is granted door opening authority and pre-stored in the information storage module.

[0025] During the comparison and analysis, the eye information processing and analysis module first compares the difference between the eye corner lengths of the identified person and the first resident. When the difference is less than or equal to the threshold, the eye information processing and analysis module performs the next step of comparison and analysis on whether the first angle θ between the first straight line of the identified person and the first resident and the horizontal axis of the first right-angle coordinate system are equal. The eye information processing and analysis module outputs the comparison and analysis result to the control module. If the comparison and analysis result received by the control module is that the eye information of the current identified person matches that of the first resident, the control door is controlled to open. If the comparison and analysis result is that the eye information of the current identified person does not match that of the first resident, an alarm message is sent to the mobile phone application terminal of the first resident through the communication and alarm module.

[0026] The distance sensing module obtains the distance S between the person in front of the door who is undergoing eye recognition and the door. The daylight simulation module and the image acquisition module are both set on the door. The daylight simulation module is located directly above the image acquisition module. The distance between the daylight simulation module and the image acquisition module is H. The eyes of the currently recognized person and the image acquisition module are located on the same horizontal line. The processing and calculation module establishes a second rectangular coordinate system with the horizontal line as the horizontal axis of the coordinate. The line connecting the image acquisition module and the eyes of the currently recognized person is the first straight line. The processing and calculation module calculates the second angle ɑ between the first straight line and the horizontal axis of the second rectangular coordinate system. The second angle The processing calculation module outputs the calculation results to the daylight simulation module.

[0027] The daylight simulation module is a light supplement device embedded in the surface of the door and can move freely up and down to adjust the angle of the irradiation light beam. The daylight simulation module adjusts the irradiation angle of the supplementary light according to the second angle ɑ, and the irradiation angle of the supplementary light is the second angle ɑ.

[0028] The illumination acquisition module is used to collect the light intensity in the current environment. If the light intensity is less than or equal to the light intensity threshold, the light supplement is enabled through the daylight simulation module.

[0029] The infrared thermal imaging module detects whether there is a person to be identified in front of the door. The criteria for determining the person to be identified are the time the person stays in front of the door and the third angle β between the second straight line and the first plane. The two shoulder acromion points of the image of the person in front of the door are selected. The two shoulder acromion points include acromion point A and acromion point B. The second straight line is the line connecting acromion point A and acromion point B. The first plane is the plane where the door is located. When the infrared thermal imaging module detects that there is a person staying in front of the door, the timing starts. When the person stays in front of the door for more than time t, the infrared thermal imaging module puts the person in the queue of people to be identified.

[0030] The processing and calculation module calculates the third angle β, that is, calculates the angle between the line connecting the acromion points of the two shoulders and the first plane, draws a perpendicular line through the acromion point A perpendicular to the first plane, with the foot of the perpendicular being C, draws a perpendicular line through the acromion point D perpendicular to the first plane, with the foot of the perpendicular being D, and connects the foot of the perpendicular C and the foot of the perpendicular D. The third angle β is the angle between the second straight line and the straight line CD. When the second straight line is perpendicular to the first plane, the third angle β is 90 degrees.

[0031] The processing calculation module calculates the third angle deviation, the third angle deviation Among them, β 0 is the third angle threshold. If the calculated third angle deviation p is less than or equal to the threshold, the person currently staying in front of the door is determined as a person to be identified, and the image acquisition module obtains the eye image of the person to be identified. If the calculated third angle deviation p is greater than the threshold, the person currently staying in front of the door is determined as a person not to be identified.

[0032] Working principle of the present invention:

[0033] S1: Using an infrared thermal imaging module to detect whether there is a person to be identified in front of the door. The method for detecting the person to be identified includes the following steps:

[0034] S11: Determine whether the stay time of the person staying in front of the door exceeds a preset time threshold. If the stay time is greater than or equal to the time threshold, proceed to the next step of detecting the person to be identified;

[0035] S12: Collect an image of the person staying in front of the door, extract the two shoulder acromion points of the person staying in front of the door in the image, connect the two shoulder acromion points, calculate the angle between the connecting line and the plane where the door is located, and further calculate the angle deviation, and determine whether the calculated result meets the determination condition of the person to be identified;

[0036] S2: If the person to be identified passes the detection, the light intensity in the current environment is collected through the illumination collection module. If the light intensity is lower than the preset threshold, the light supplement is turned on, and the best light supplement angle is obtained by calculation;

[0037] S3: The image acquisition module collects the eye image of the person to be identified, and determines through the identity recognition module whether the eye information of the current person to be identified is consistent with the pre-stored eye information of the resident who is granted the right to open the door, that is, determines whether the current person to be identified has the right to open the door. If it is consistent, the door is opened through the control module. If it is not consistent, an alarm message is sent to the resident's mobile phone application terminal through the communication and alarm module.

[0038] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0039] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or replace some of the technical features therein by equivalents. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An eye recognition access control system based on neural network, Features: The invention comprises an information storage module, an identity recognition module, an image acquisition module, an eye information processing and analysis module, a control module, a communication and alarm module, an illumination acquisition module, a daylight simulation module, a distance sensing module, a processing and calculation module, an infrared thermal imaging module and a mobile phone application terminal, wherein the information storage module pre-stores the eye information of the first resident who is granted the door opening authority, the image acquisition module is used to acquire the eye image of the person who is currently undergoing eye information recognition, the eye information processing and analysis module is used to process and analyze the eye image of the person who is currently undergoing eye information recognition, and the identity recognition module confirms the person who is currently undergoing eye information recognition according to the comparison result of the processing and analysis result of the eye information processing and analysis module with the eye information of the resident pre-stored in the information storage module. Whether the identified person has the right to open the door, the control module decides to open the door or alarm according to the identification result of the identity recognition module, the control module is also used to control the daylight simulation module, the control module alarms the residents who are granted the right to open the door and are pre-stored in the information storage module through the communication and alarm module, the illumination acquisition module is used to collect the light intensity in a certain area in front of the door, the daylight simulation module is used to supplement the light, the distance sensing module is used to obtain the distance between the person who is undergoing eye recognition in front of the door and the door, the processing and calculation module is used to process and calculate the data obtained by the distance sensing module, the infrared thermal imaging module is used to detect whether there is a person to be identified in front of the door, and the mobile phone application end is used to receive the alarm information sent by the communication and alarm module; The eyeball information processing and analysis module connects the inner canthus vertex and the outer canthus vertex of the acquired eye image, the connecting line is a first straight line, and a first rectangular coordinate system is established with the inner canthus vertex as the origin of the coordinate axis. The horizontal axis direction of the first rectangular coordinate system is the same as the direction from the inner canthus vertex to the outer canthus vertex. The first angle θ between the first straight line and the horizontal axis and the length L of the eye angle distance are calculated. The length of the eye angle distance is the length of the first straight line, and the calculated first angle θ and the length L of the eye angle distance are compared and analyzed with the eyeball information of the first resident who is granted the door opening authority and pre-stored in the information storage module; The infrared thermal imaging module detects whether there is a person to be identified in front of the door. The criteria for determining the person to be identified are the time the person stays in front of the door and the third angle β between the second straight line and the first plane. The two shoulder acromion points of the image of the person in front of the door are selected. The two shoulder acromion points include acromion point A and acromion point B. The second straight line is the line connecting acromion point A and acromion point B. The first plane is the plane where the door is located. When the infrared thermal imaging module detects that there is a person staying in front of the door, the timing starts. When the person stays in front of the door for more than time t, the infrared thermal imaging module puts the person into the queue of people to be identified. The distance sensing module obtains the distance S between the person who is undergoing eye recognition in front of the door and the door. The daylight simulation module and the image acquisition module are both set on the door. The daylight simulation module is located directly above the image acquisition module. The distance between the daylight simulation module and the image acquisition module is H. The eyes of the currently recognized person and the image acquisition module are located on the same horizontal line. The processing and calculation module establishes a second rectangular coordinate system with the horizontal line as the horizontal axis of the coordinate. The line connecting the image acquisition module and the eyes of the currently recognized person is a first straight line. The processing and calculation module calculates a second angle ɑ between the first straight line and the horizontal axis of the second rectangular coordinate system. The second angle The processing and calculation module outputs the calculation results to the daylight simulation module; The processing and calculation module calculates the third angle β, that is, calculates the angle between the line connecting the acromion points of the two shoulders and the first plane, draws a perpendicular line through the acromion point A perpendicular to the first plane, with the foot of the perpendicular being C, draws a perpendicular line through the acromion point D perpendicular to the first plane, with the foot of the perpendicular being D, and connects the foot of the perpendicular C and the foot of the perpendicular D. The third angle β is the angle between the second straight line and the straight line CD. When the second straight line is perpendicular to the first plane, the third angle β is 90 degrees.

2. According to claim 1, an eye recognition access control system based on a neural network, Features: The information storage module pre-stores the eyeball information of the first resident who is granted the door opening permission, including the length of the eye canthus of the resident's eyes, and the angle between the line connecting the inner canthus vertex and the outer canthus vertex and the horizontal line.

3. According to claim 2, an eye recognition access control system based on a neural network, Features: During the comparison and analysis, the eye information processing and analysis module first compares the difference between the eye corner lengths of the identified person and the first resident. When the difference is less than or equal to the threshold, the eye information processing and analysis module performs the next step of comparison and analysis on whether the first angle θ between the first straight line of the identified person and the first resident and the horizontal axis of the first right-angle coordinate system are equal. The eye information processing and analysis module outputs the comparison and analysis result to the control module. If the comparison and analysis result received by the control module is that the eye information of the current identified person matches that of the first resident, the control door is controlled to open. If the comparison and analysis result is that the eye information of the current identified person does not match that of the first resident, an alarm message is sent to the mobile phone application terminal of the first resident through the communication and alarm module.

4. According to claim 3, an eye recognition access control system based on a neural network, Features: The daylight simulation module is a light supplement device embedded in the surface of the door and can move freely up and down to adjust the angle of the illumination beam. The daylight simulation module adjusts the illumination angle of the supplementary light according to the second angle ɑ, and the illumination angle of the supplementary light is the second angle ɑ.

5. According to claim 4, an eye recognition access control system based on a neural network, Features: The illumination acquisition module is used to collect the light intensity in the current environment. If the light intensity is less than or equal to the light intensity threshold, the light supplement is enabled through the daylight simulation module.

6. The neural network-based eye recognition access control system according to claim 5, Features: The processing and calculation module calculates a third angle deviation, wherein the third angle deviation Among them, β 0 is the third angle threshold. If the calculated third angle deviation p is less than or equal to the threshold, the person currently staying in front of the door is determined to be a person to be identified. The image acquisition module acquires the eye image of the person to be identified. If the calculated third angle deviation p is greater than the threshold, the person currently staying in front of the door is determined to be a person not to be identified.

Images