Gate access management system based on images

Abstract

The invention discloses an image-based gate access management system, relates to the technical field of image analysis and gate management, and solves the problems that some existing gate management systems only use a facial recognition technology, have no measures for dealing with behaviors such as queue jumping, cannot effectively limit ordered queuing of people, and are inconvenient to use. And disordered entrance and exit management is easily caused. According to the invention, through the image acquisition unit, the long-shot data and the close-shot data are obtained, and through the processing and analysis of the long-shot data by the image preprocessing unit, the passing willingness of identification personnel in the identification range of the long-shot camera is effectively predicted, and the identification personnel are numbered; the face recognition unit locks the personnel through serial numbers and carries out face comparison and analysis in sequence; meanwhile, face recognition is also provided with a face secondary recognition measure and an access registration alarm step; the face recognition unit sends passing data of all people to the database management unit and the gate management unit; and the database management unit stores and records all passing records.

Description

Technical Field

[0001] The present invention relates to the technical field of image analysis and gate management, and in particular to an image-based gate access management system. Background Art

[0002] Gate management system, also known as access control system, is an automated device and system used to manage pedestrian access; this system is mainly used to control, monitor and record the entry and exit of people in specific places, such as company buildings, factories, schools, communities, scenic spots, etc. Gate management system usually consists of gate equipment, card readers, controllers, power supply systems and alarm systems;

[0003] At present, some existing gate management systems simply use facial recognition technology, and have no measures to deal with behaviors such as queue jumping. They cannot effectively restrict people from queuing in an orderly manner, which can easily cause chaos in entrance and exit management. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an image-based gate access management system, which solves the problem that some existing gate management systems simply use facial recognition technology, have no measures to deal with behaviors such as queue jumping, cannot effectively restrict the crowd to queue in an orderly manner, and easily cause chaotic entrance and exit management.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an image-based gate access management system, comprising:

[0006] The image acquisition unit is used to acquire on-site image data and registered image data; the on-site image data includes the image information acquired at the gate entrance and exit by using the camera group; the registered image data includes connecting to the face entry terminal, which is used to enter the facial information of the registrant and bind the identity information; the on-site image data and the registered image data are sent to the image preprocessing unit, wherein the registered image data is sent to the database management unit for storage;

[0007] The image preprocessing unit is used to process the on-site image data, analyze the behavior information of the people in the image for the distant view data, number the people covered in the image, and generate pre-identification information;

[0008] The face recognition unit is used to receive pre-recognition information, verify it according to the number sorting, lock the recognized person, analyze the close-up data, and mobilize the database registered face information for comparison;

[0009] The database management unit is used to store the facial information and identity information of the registered personnel, as well as the access appointment application, and receive and record the authentication information from the face recognition unit; the above information is mobilized by the face recognition unit and the gate control unit; the staff manually inputs the appointment information and the appointment application for verification, and if the verification is successful, the appointment authentication information is generated and sent to the gate control unit and recorded;

[0010] The gate control unit is used to receive the authentication information of the face recognition unit and the reservation authentication information of the database management unit, and send instructions to control the gate switch.

[0011] In the image acquisition unit, the image information acquisition process includes the following steps:

[0012] Two groups of cameras are set up at the gate entrance. One group of cameras, recorded as close-range cameras, is responsible for capturing images within two meters in front of the gate entrance and generating close-range data.

[0013] Another set of cameras, referred to as long-range cameras, are responsible for capturing images two meters in front of the gate entrance and generating long-range data;

[0014] If the long-range camera does not detect any human activity for more than c seconds, it sends a standby message to the gate control unit, and the gate enters the standby state by sending a standby command.

[0015] In the image preprocessing unit, the reference vector is generated by using the perspective data transmitted in real time by the perspective camera, including the following steps:

[0016] Extract reference vectors, including horizontal basis vector d, horizontal basis vector e, and trend vector g:

[0017] Extract the perspective data, identify the side direction of the person's arm swing, and generate the horizontal basis vector d;

[0018] Differentiate the image of the person to be identified, eliminate the interference of other image information, and draw a graph in the forward projection of the person to be identified. Take the midpoint of the line connecting the two shoulders of the person to be identified as the starting point, and derive a horizontal basis vector e from the back of the person to the chest;

[0019] From the time when the identified person is detected to the time when he / she walks into the two-meter range of the gate, several images are extracted, and the time interval between each image is f. The background in the image is abstracted into a pixel grid, and each identified person is a pixel point, and each pixel point is marked on the pixel grid; the software MATLAB is used to establish the xy coordinate axis, and the corresponding relationship between several pixel points is calculated, recorded as function Y(x), and the position function Y(x) is derived to obtain the first-order derivative function y(x) of Y(x). Through y(x), the trend vector of the pixel point at the last pixel point is calculated.

[0020] The analyzing the reference vector and generating the pre-identification information comprises the following steps:

[0021] Define the reference angle deviation interval D;

[0022] Calculate the horizontal basis vector d and horizontal basis vector respectively Trend Vector The offset angle θ i ; The calculation formula is:

[0023]

[0024] θ i =arccos(cosθ i );

[0025] The horizontal basis vectors Horizontal basis vectors Trend Vector Combine two by two and substitute them in turn and Calculate and find the three offset angles;

[0026] If the three offset angles are all within interval D, and it is determined that the identified person has the intention to pass through the gate, pre-identification information is generated and sent to the face recognition unit;

[0027] For those who are judged to be willing to pass through the gate, they will be sorted and numbered in sequence, and those who have no intention to pass through the gate will be fuzzy processed.

[0028] In the face recognition unit, face comparison includes the following steps:

[0029] The comparison step includes a pre-identification step and an identification step;

[0030] In the pre-identification step, firstly, the pre-identification information from the image acquisition unit is received, the identification personnel and their respective numbers are verified, and a command to start the identification step is sent;

[0031] For people who appear outside the close-up shot, including those whose face matching fails in the first identification and those who are accidentally inserted into the identification range;

[0032] If a person whose face matching fails in the first identification is detected, a secondary refresh operation is triggered to perform a second face recognition on the identified person;

[0033] If a person is detected to have accidentally entered the identification range, an alarm will be triggered and a staff member will come forward to remind the latter to queue up again or visit for registration;

[0034] The identification step includes: extracting frames for the numbered identification personnel, extracting a facial image every thirty frames, and taking a facial image that meets the preset specifications for feature point comparison; if the identification is successful, authentication information is generated and sent to the gate control unit and the database management unit;

[0035] The preset specification is: meeting the standard resolution of 350dpi.

[0036] The access registration includes the following steps:

[0037] With the assistance of the staff, the visitor enters his or her own information and appointment date at the terminal, and calls the pre-stored visit information in the database for comparison and verification. If the verification is passed, it will be renumbered and sent to the gate control unit along with the pass instruction; the renumbering rule is: take the number after the last identified person in the close-up shot.

[0038] The secondary face recognition comprises the following steps:

[0039] When the first facial recognition fails, but the facial feature comparison of the unit is not thorough and a second recognition cannot be performed quickly, the registrant only needs to move the face away from the scanning area. The sensor will detect the huge spatial position changes of the facial feature points and will immediately terminate the comparison process and enter the next round of pre-recognition steps without refreshing the number.

[0040] Beneficial Effects

[0041] The present invention provides an image-based gate access management system. Compared with the prior art, it has the following beneficial effects:

[0042] The present invention obtains long-range data and short-range data through an image acquisition unit, processes and analyzes the long-range data through an image preprocessing unit, effectively predicts the passing intention of the identified personnel within the recognition range of the long-range camera, and numbers them; the face recognition unit locks the personnel by number and performs face comparison analysis in turn; at the same time, face recognition is also provided with a face secondary recognition measure and an access registration alarm step; the face recognition unit sends all personnel passing data to a database management unit and a gate management unit; the database management unit stores and records all passing records and registration appointment application information, which can be used for access registration comparison; the gate control unit receives the switch command of the face recognition unit and controls the passage of the gate;

[0043] The present invention collects grouped data through an image acquisition unit, including near-view data and far-view data. By analyzing the near-view data and far-view data in groups, it is convenient to effectively allocate personnel in the image acquisition area, improves personnel management efficiency, and avoids the probability of recognition errors caused by the camera failing to effectively locate the face to be analyzed when there are many people.

[0044] By analyzing the behavioral information of the identified personnel in the vision data, the behavioral intention is effectively predicted, and the identified personnel who have the intention to pass through the gate are conveniently numbered and locked, and the identified personnel who do not have the intention to pass through the gate are differentiated. When the flow of people is large, the interference of the registered personnel staying or passing by in front of the gate without intention to pass through the gate is effectively eliminated, and the accident of unregistered personnel taking the opportunity to pass through the gate during the gate opening gap is avoided;

[0045] Secondary face recognition interrupts the first recognition and refreshes the second recognition, effectively shortening the time from the first recognition failure to the second re-recognition, improving service quality, speeding up the time for people to pass through the gate, and optimizing system operation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of the principle framework of the present invention. DETAILED DESCRIPTION

[0047] 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.

[0048] Embodiment 1

[0049] See also Figure 1 , the present application provides an image-based gate access management system, including an image acquisition unit, an image preprocessing unit, a face recognition unit, a database management unit and a gate control unit;

[0050] The image acquisition unit is used to collect on-site image data and registered image data; the on-site image data includes the image information collected at the gate entrance and exit by using a camera group; the registered image data includes connecting to a face entry terminal for entering the facial information of the registrant and binding identity information; the on-site image data and the registered image data are sent to the image preprocessing unit, wherein the registered image data is sent to the database management unit for storage;

[0051] Two groups of cameras are set up at the gate entrance. One group of cameras, recorded as close-range cameras, is responsible for capturing images within two meters in front of the gate entrance and generating close-range data.

[0052] Another set of cameras, referred to as long-range cameras, are responsible for capturing images two meters in front of the gate entrance and generating long-range data;

[0053] If the vision camera does not detect human activity for more than c seconds, it sends a standby message to the gate control unit, and the gate enters the standby state by sending a standby command;

[0054] The image acquisition unit is used to collect grouped data, including near-view data and far-view data. By grouping and analyzing near-view data and far-view data, it is convenient to effectively allocate personnel in the image acquisition area, thereby improving personnel management efficiency and avoiding the possibility of recognition errors caused by the camera's failure to effectively locate the face to be analyzed when there are too many people.

[0055] Embodiment 2

[0056] The image preprocessing unit is used to process the on-site image data, analyze the behavior information of the people in the image with respect to the distant view data, number the people covered in the image, and generate pre-identification information;

[0057] Extract reference vectors, including horizontal basis vector d, horizontal basis vector e, and trend vector g:

[0058] Extract the perspective data, identify the side direction of the person's arm swing, and generate the horizontal basis vector d;

[0059] Differentiate the image of the identified person, eliminate the interference of other image information, and draw a horizontal basis vector from the back of the identified person to the chest direction, starting from the midpoint of the line connecting the two shoulders of the identified person.

[0060] From the time when the identified person is detected to the time when he / she walks into the two-meter range of the gate, several images are extracted, and the time interval between each image is f. The background in the image is abstracted into a pixel grid, and each identified person is a pixel point, and each pixel point is marked on the pixel grid; the software MATLAB is used to establish the xy coordinate axis, and the corresponding relationship between several pixel points is calculated, recorded as function Y(x), and the position function Y(x) is derived to obtain the first-order derivative function y(x) of Y(x). Through y(x), the trend vector of the pixel point at the last pixel point is calculated.

[0061] Define the reference angle deviation interval D;

[0062] Calculate the horizontal basis vector d and horizontal basis vector respectively Trend Vector The offset angle θ i ; The calculation formula is:

[0063]

[0064] θ i =arccos(cosθ i);

[0065] The horizontal basis vectors Horizontal basis vectors Trend Vector Combine two by two and substitute them in turn Calculate in and B to find the three offset angles;

[0066] If the three offset angles are all within interval D, and it is determined that the identified person has the intention to pass through the gate, pre-identification information is generated and sent to the face recognition unit;

[0067] For those who are determined to be willing to pass through the gate, they are sorted and numbered in sequence, and those who are not willing to pass through the gate are fuzzy processed to avoid affecting the subsequent analysis of the identification of personnel behavior information;

[0068] The image acquisition unit is used to collect grouped data, including near-field data and long-field data. By analyzing the behavioral information of the identified persons in the long-field data, the behavioral intention is effectively predicted, which facilitates the numbering and locking of the identified persons who are willing to pass through the gate, and the differentiation of the identified persons who do not intend to pass through the gate. When the flow of people is large, the interference of registered persons staying or passing by in front of the gate without intending to pass through the gate is effectively eliminated, and the accidents of unregistered persons taking the opportunity to pass through the gate during the opening gap are avoided.

[0069] The face recognition unit is used to receive pre-recognition information, verify it according to the number sorting, lock the recognized person, analyze the close-up data, and mobilize the database registered face information for comparison;

[0070] The comparison step includes a pre-identification step and an identification step;

[0071] In the pre-identification step, firstly, the pre-identification information from the image acquisition unit is received, the identification personnel and their respective numbers are verified, and a command to start the identification step is sent;

[0072] For people who appear outside the close-up shot, including those who failed the initial face recognition and those who accidentally inserted into the recognition range;

[0073] If a person whose face matching fails in the first identification is detected, a secondary refresh operation is triggered to perform a second face recognition on the identified person;

[0074] If a person is detected to have accidentally entered the identification range, an alarm will be triggered and a staff member will come forward to remind the latter to queue up again or visit for registration;

[0075] The identification step includes: extracting frames for the numbered identification personnel, extracting a facial image every thirty frames, and taking a facial image that meets the preset specifications for feature point comparison; if the identification is successful, authentication information is generated and sent to the gate control unit and the database management unit;

[0076] The preset specifications are: meeting the standard resolution of 350dp i

[0077] With the assistance of the staff, the visitor enters his / her personal information and appointment date at the terminal, and the visit information stored in the database is called for comparison and verification. If the verification is passed, the number is renumbered and sent to the gate control unit along with the pass instruction; the renumbering rule is: take the number after the last person identified in the close-up shot;

[0078] When the first face recognition fails, but the facial feature comparison of the unit is not thorough and the second recognition cannot be performed quickly, the registrant only needs to move the face away from the scanning area. The sensor detects the huge spatial position change of the facial feature points and immediately terminates the comparison process and enters the next round of pre-recognition steps without refreshing the number;

[0079] Secondary face recognition interrupts the first recognition and refreshes the second recognition, effectively shortening the time from the first recognition failure to the second re-recognition, improving service quality, speeding up the time for people to pass through the gate, and optimizing system operation efficiency.

[0080] The database management unit is used to store the facial information and identity information of the registered personnel, as well as the access appointment application, receive and record the authentication information from the face recognition unit; the above information is mobilized by the face recognition unit and the gate control unit; the staff manually inputs the appointment information and the appointment application for verification, and if the verification is successful, the appointment authentication information is generated and sent to the gate control unit and recorded.

[0081] Embodiment 3

[0082] The specific implementation process of this embodiment includes the entire implementation process of the above two groups of embodiments.

[0083] Some of the data in the above formulas are dimensionless and numerically calculated. Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0084] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An image-based gate access management system, characterized in that: The following steps are involved: The image acquisition unit is used to acquire on-site image data and registered image data; the on-site image data includes the image information acquired at the gate entrance and exit by using the camera group; the registered image data includes connecting to the face entry terminal, which is used to enter the facial information of the registrant and bind the identity information; the on-site image data and the registered image data are sent to the image preprocessing unit, wherein the registered image data is sent to the database management unit for storage; The image preprocessing unit is used to process the on-site image data, analyze the behavior information of the people in the image for the distant view data, number the people covered in the image, and generate pre-identification information; The face recognition unit is used to receive pre-recognition information, verify it according to the number sorting, lock the recognized person, analyze the close-up data, and mobilize the database registered face information for comparison; The database management unit is used to store the facial information and identity information of the registered personnel, as well as the access appointment application, and receive and record the authentication information from the face recognition unit; the above information is mobilized by the face recognition unit and the gate control unit; the staff manually inputs the appointment information and the appointment application for verification, and if the verification is successful, the appointment authentication information is generated and sent to the gate control unit and recorded; The gate control unit is used to receive the authentication information of the face recognition unit and the reservation authentication information of the database management unit, and send instructions to control the gate switch.

2. The image-based gate access management system according to claim 1, characterized in that: In the image acquisition unit, the image information acquisition process includes the following steps: Two groups of cameras are set up at the gate entrance. One group of cameras, recorded as close-range cameras, is responsible for capturing images within two meters in front of the gate entrance and generating close-range data. Another set of cameras, referred to as long-range cameras, are responsible for capturing images two meters in front of the gate entrance and generating long-range data; If the long-range camera does not detect any human activity for more than c seconds, it sends a standby message to the gate control unit, and the gate enters the standby state by sending a standby command.

3. The image-based gate access management system according to claim 1, characterized in that: In the image preprocessing unit, the reference vector is generated using the perspective data transmitted in real time by the perspective camera, including the following steps: Extract reference vectors, including horizontal basis vectors Horizontal basis vectors and trend vector Extract the perspective data, identify the side direction of the person's arm swing, and generate the horizontal basis vector Differentiate the image of the identified person, eliminate the interference of other image information, and draw a horizontal basis vector from the back of the identified person to the chest direction, starting from the midpoint of the line connecting the two shoulders of the identified person. From the time when the identified person is detected to the time when he / she walks into the two-meter range of the gate, several images are extracted, and the time interval between each image is f. The background in the image is abstracted into a pixel grid, and each identified person is a pixel point, and each pixel point is marked on the pixel grid; the software MATLAB is used to establish the xy coordinate axis, and the corresponding relationship between several pixel points is calculated, recorded as function Y(x), and the position function Y(x) is derived to obtain the first-order derivative function y(x) of Y(x). Through y(x), the trend vector of the pixel point at the last pixel point is calculated.

4. The image-based gate access management system according to claim 3, characterized in that: Analyzing the reference vector and generating pre-recognition information includes the following steps: Define the reference angle deviation interval D; Calculate the horizontal basis vector d and horizontal basis vector respectively Trend Vector The offset angle θ i ; The calculation formula is: i i =arccos(cosθ) i ); The horizontal basis vectors Horizontal basis vectors Trend Vector Combine two by two and substitute them in turn and Calculate and find the three offset angles; If the three offset angles are all within interval D, and it is determined that the identified person has the intention to pass through the gate, pre-identification information is generated and sent to the face recognition unit; For those who are judged to be willing to pass through the gate, they will be sorted and numbered in sequence, and those who have no intention to pass through the gate will be fuzzy processed.

5. The image-based gate access management system according to claim 1, characterized in that: In the face recognition unit, face comparison includes the following steps: The comparison step includes a pre-identification step and an identification step; In the pre-identification step, firstly, the pre-identification information from the image acquisition unit is received, the identification personnel and their respective numbers are verified, and a command to start the identification step is sent; For people who appear outside the close-up shot, including those whose face matching fails in the first identification and those who are accidentally inserted into the identification range; If a person whose face matching fails in the first identification is detected, a secondary refresh operation is triggered to perform a second face recognition on the identified person; If a person is detected to have accidentally entered the identification range, an alarm will be triggered and a staff member will come forward to remind the latter to queue up again or visit for registration; The identification step includes: extracting frames for the numbered identification personnel, extracting a facial image every thirty frames, and taking a facial image that meets the preset specifications for feature point comparison; if the identification is successful, authentication information is generated and sent to the gate control unit and the database management unit; The preset specification is: meeting the standard resolution of 350dpi.

6. The image-based gate access management system according to claim 5, characterized in that: Access registration involves the following steps: With the assistance of the staff, the visitor enters his or her own information and appointment date at the terminal, and calls the pre-stored visit information in the database for comparison and verification. If the verification is passed, it will be renumbered and sent to the gate control unit along with the pass instruction; the renumbering rule is: take the number after the last identified person in the close-up shot.

7. The image-based gate access management system according to claim 5, characterized in that: Secondary face recognition includes the following steps: When the first facial recognition fails, but the facial feature comparison of the unit is not thorough and a second recognition cannot be performed quickly, the registrant only needs to move the face away from the scanning area. The sensor will detect the huge spatial position changes of the facial feature points and will immediately terminate the comparison process and enter the next round of pre-recognition steps without refreshing the number.

Images