Intelligent access control system with alarm function

Abstract

The invention belongs to the technical field of access control systems, and particularly relates to an intelligent access control system with an alarm function. Comprising an access control unit, an alarm unit and a face recognition unit, the access control unit is used for allowing an authorizer to access when the authorizer is recognized; the alarm unit is used for generating an alarm when an illegal person is recognized, and sending an alarm sound to the scene when a person without facial recognition approaches the access control system; the face recognition unit comprises a face storage unit, a depth image shooting unit, a depth image correction unit, a depth image rotation unit, a depth image stretching unit, a face detection unit, a face feature extraction unit, a face feature comparison unit and a personnel matching determination unit. According to the intelligent access control system with the alarm function, the perspective distortion phenomenon of a face recognition image is corrected, the reliability of the face recognition access control system is improved, and technical support is provided for efficient and accurate operation of the intelligent access control system with the alarm function.

Description

Technical Field

[0001] This invention belongs to the field of access control system technology, and particularly relates to a smart access control system with alarm. Background Technology

[0002] Facial recognition technology, as an important component of biometrics, is based on the core principle of machines using the unique feature information contained in each person's face to achieve automatic identification and authentication of individuals. This technology has significant advantages in image acquisition; facial images can be easily and naturally obtained from various video media. Through technical processing, faces can be separated from complex backgrounds, and key features such as the eyes, nose, and mouth can be accurately located. This allows for the creation of facial templates, which are then stored in a database using statistical methods, providing a foundation for subsequent registration, recognition, and authentication. In the field of biometric systems, fingerprint, vein, and iris recognition technologies are widely used and have been commercialized. However, these biometric systems have significant drawbacks: they often require human contact, making data collection difficult and lacking in intuitiveness. In contrast, facial recognition technology uses a contactless photographing method, and the facial image data it records and uses is highly intuitive, making it more promising for applications in many scenarios. Facial recognition technology typically uses image processing techniques to compare the image to be recognized with pre-stored reference images to perform pattern recognition. This pattern recognition method has typical applications in many fields, such as PCB defect determination in optical inspection systems on printed circuit board production lines, automatic license plate recognition in intelligent transportation systems, and pattern matching technology in Internet of Things (IoT) technologies. However, in practical applications, when using traditional cameras to capture images, the camera position can cause geometric distortion due to perspective, which poses a challenge to the accuracy of pattern recognition. Summary of the Invention

[0003] The purpose of this invention is to address the aforementioned technical problems by providing an alarm-based smart access control system that corrects perspective distortion in facial recognition images, improves the reliability of facial recognition access control systems, and provides technical support for the efficient and accurate operation of smart access control systems with alarms.

[0004] In view of this, the present invention provides a smart access control system with alarm, including an access control unit, an alarm unit and a face recognition unit; Access control units are used to grant access when an authorized person is identified; The alarm unit is used to generate an alarm when an unauthorized person is detected. When a person who has not undergone facial recognition approaches the access control system, it sends an alarm sound to both the site and the central control room, and displays an emergency warning message on the control monitor. The face recognition unit includes a face storage unit, a depth image capture unit, a depth image correction unit, a depth image rotation unit, a depth image stretching unit, a face detection unit, a face feature extraction unit, a face feature comparison unit, and a person matching and determination unit; The system includes: a face storage unit for storing facial feature depth information; a depth image capture unit for capturing facial depth images; a depth image correction unit for correcting depth value errors; a depth image rotation unit for performing image rotation transformation and face alignment; a depth image stretching unit for expanding and shrinking facial images according to the image capture distance; a face detection unit for extracting facial portions from the depth image; a face feature extraction unit for extracting facial features from the depth image; a face feature comparison unit for comparing the data stored in the face storage unit; and a personnel matching determination unit for determining the degree of personnel matching.

[0005] In the above technical solution, the depth image rotation unit further includes a depth information calculation unit, a coordinate transformation unit, a local normal vector calculation unit, a plane normal vector calculation unit, a rotation matrix calculation unit, and a perspective distortion correction unit. The depth information calculation unit is used to calculate depth information from the planar image captured by the depth image capture unit; The coordinate transformation unit is used to calculate the position of the depth image capturing unit in the coordinate system using the depth information calculated by the depth information calculation unit; The local normal vector calculation unit is used to calculate the local normal vector of a pixel using the surrounding information of each pixel calculated by the coordinate transformation unit; The plane normal vector calculation unit is used to obtain the normal vector of the entire plane using the local normal vector obtained by the local normal vector calculation unit; The rotation matrix calculation unit is used to calculate the angle between the rotation axis and the depth image to obtain the rotation matrix; The perspective distortion correction unit is used to correct image distortion by applying a rotation transformation based on the position of the depth image capture unit.

[0006] In any of the above technical solutions, the depth image stretching unit further includes an information acquisition unit, an average value calculation unit, and an expansion / contraction ratio calculation unit; The information acquisition unit is used to simultaneously acquire and convert color and depth information; The average value calculation unit is used to calculate the average value of the color and depth information of the current block in the current image and the average value of the color and depth information of the reference block in the reference image based on the color and depth information obtained by the information acquisition unit. The expansion / contraction ratio calculation unit is used to calculate the expansion / contraction ratio based on the average depth information of the current block and the reference block, and to perform the calculation by scaling up or down the reference screen according to the expansion ratio.

[0007] In any of the above technical solutions, the face storage unit further includes a dynamic update module and a categorized storage module; The dynamic update module is used to periodically receive the latest facial feature depth information from authorized personnel and update and replace the stored historical information. The classification and storage module is used to classify and store the facial feature depth information of authorized personnel according to their permission level, department, and other information, so as to facilitate quick retrieval and access.

[0008] In any of the above technical solutions, a linkage control unit is further included, which is connected to the access control unit, the alarm unit, and the face recognition unit respectively; when the face recognition unit recognizes an authorized person, the linkage control unit controls the access control unit to open and simultaneously sends the personnel access record to the relevant management system; When the alarm unit sounds an alarm, the linkage control unit can keep the access control unit locked and link the on-site monitoring equipment to focus on capturing and recording the alarm area, while synchronizing the relevant information to the situation room center.

[0009] In any of the above technical solutions, the access control unit further includes a turnstile, an opening and closing mechanism, a drive shaft, and a transmission mechanism; The turnstile has a symmetrically distributed body, with the facial detection unit located on the body; The opening and closing mechanism is located on the opposite surfaces of the two bodies and rotates to open when an authorized person is detected. The drive shaft is rotatably mounted on the machine body and is connected to the opening and closing mechanism; The transmission mechanism is connected to the drive shaft and drives the drive shaft and opening / closing mechanism when an authorized person is identified.

[0010] In any of the above technical solutions, the opening and closing mechanism further includes: A disc is mounted on the drive shaft and fits against the inner wall of the machine body; Ear plates are set on the disk, and there are multiple ear plates, which are evenly distributed in the circumferential direction; Through holes are formed on the ear plates; Bolts are inserted through the through holes; A nut, threaded, is attached to one end of the bolt; A connecting plate is located between the two ear plates. Bolts pass through the connecting plate, and the connecting plate is rotatably mounted on the bolts. The gate arm is mounted on the connecting plate.

[0011] In any of the above technical solutions, furthermore, the turnstiles are spaced apart to form at least one set of turnstile channels.

[0012] The beneficial effects of this invention are: 1. By processing depth images through shooting, correction, rotation and stretching, image distortion caused by shooting angle and distance is eliminated, solving problems such as perspective distortion and scaling error in traditional 2D face recognition, ensuring efficient identification and accurate authentication of authorized personnel; the depth image rotation unit corrects perspective distortion, and the depth image stretching unit eliminates size differences caused by shooting distance, ensuring accurate feature comparison and reducing recognition errors.

[0013] 2. The alarm unit triggers both on-site and remote alarms when unauthorized or unidentified personnel are detected, and the linkage control unit simultaneously locks the access control system and links with monitoring to effectively prevent illegal intrusion.

[0014] 3. Multiple turnstile channels operate in parallel to improve passage efficiency; dynamic updates and categorized storage of facial recognition data ensure information timeliness and ease of retrieval; and the linkage control unit enables automated management of access records.

[0015] 4. The opening and closing mechanism has a stable structure and the gate arm rotates flexibly. The multi-channel design is fault-tolerant, and the coordinated operation of each unit reduces the impact of failures and ensures continuous system operation.

[0016] 5. The number of turnstile channels can be set according to the needs of the scenario. The opening and closing mechanism components are detachable, which facilitates installation and maintenance and reduces costs. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the first three-dimensional structure of the present invention; Figure 2 This is a schematic diagram of the second three-dimensional structure of the present invention; Figure 3 This is a three-dimensional structural schematic diagram of the opening and closing mechanism of the present invention; Figure 4 This is a system framework diagram of the present invention; The attached diagram is labeled as follows: 1. Access control unit; 2. Alarm unit; 3. Face recognition unit; 31. Face storage unit; 311. Dynamic update module; 312. Classification storage module; 32. Depth image capture unit; 33. Depth image correction unit; 34. Depth image rotation unit; 341. Depth information calculation unit; 342. Coordinate transformation unit; 343. Local normal vector calculation unit; 344. Plane normal vector calculation unit; 345. Rotation matrix calculation unit; 346. Perspective distortion correction unit. 35. Depth Image Stretching Unit; 351. Information Acquisition Unit; 352. Average Value Calculation Unit; 353. Expansion / Contraction Comparison Unit; 36. Face Detection Unit; 37. Face Feature Extraction Unit; 38. Face Feature Comparison Unit; 39. Personnel Matching and Determination Unit; 4. Linkage Control Unit; 5. Turnstile; 51. Body; 6. Opening and Closing Mechanism; 61. Disc; 62. Ear Plate; 63. Through Hole; 64. Bolt; 65. Nut; 66. Connecting Plate; 67. Gate Arm; 7. Drive Shaft; 8. Transmission Mechanism. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0019] In the description of this application, it should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. For ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following drawings denote similar items, and therefore, once an item is defined in one drawing, it need not be further discussed in subsequent drawings.

[0020] Example 1: like Figure 1 , Figure 2 and Figure 4 As shown, this embodiment provides an alarm-enabled smart access control system, including an access control unit 1, an alarm unit 2, and a face recognition unit 3; Access control unit 1 is used to allow access when an authorized person is identified; Alarm unit 2 is used to generate an alarm when an unauthorized person is detected. When a person who has not undergone facial recognition approaches the access control system, it sends an alarm sound to both the site and the central control room, and displays an emergency warning message on the control monitor. The face recognition unit 3 includes a face storage unit 31, a depth image capture unit 32, a depth image correction unit 33, a depth image rotation unit 34, a depth image stretching unit 35, a face detection unit 36, a face feature extraction unit 37, a face feature comparison unit 38, and a person matching and determination unit 39. The system includes: a face storage unit 31 for storing facial feature depth information; a depth image capturing unit 32 for capturing facial depth images; a depth image correction unit 33 for correcting depth value errors; a depth image rotation unit 34 for performing image rotation transformation and face alignment; a depth image stretching unit 35 for expanding and shrinking facial images according to the image capturing distance; a face detection unit 36 ​​for extracting facial portions from the depth image; a face feature extraction unit 37 for extracting facial features from the depth image; a face feature comparison unit 38 for comparing the data stored in the face storage unit 31; and a personnel matching determination unit 39 for determining the degree of personnel matching.

[0021] This technical solution eliminates image distortion caused by shooting angle and distance through depth image capture, correction, rotation, and stretching, solving problems such as perspective distortion and scaling errors in traditional 2D face recognition, ensuring efficient identification and accurate authentication of authorized personnel. For unauthorized personnel or those approaching without facial recognition, alarm unit 2 enables on-site warnings and remote alerts, promptly preventing unauthorized intrusion and providing real-time warning information to the monitoring center, improving security response speed. The face storage unit 31 centrally manages authorized personnel information, combined with the automated recognition and matching functions of face recognition unit 3, reducing manual intervention and achieving precise control of access permissions. It is suitable for various scenarios such as offices, residential communities, and important facilities.

[0022] Working Principle: The depth image capturing unit 32 captures a depth image of the face of the person to be identified, obtaining facial data containing three-dimensional spatial information, which better reflects the three-dimensional features of the face compared to traditional 2D images. The depth image correction unit 33 corrects depth value errors that may occur during the capturing process, ensuring the accuracy of the depth data and providing a reliable foundation for subsequent processing. The depth image rotation unit 34 performs rotation transformation and face alignment processing for image tilt or offset caused by the shooting angle. By calculating the local normal vector of pixels, the plane normal vector, and the rotation matrix, perspective distortion is corrected to ensure that the facial image is in a standard pose. The depth image stretching unit 35 enlarges or shrinks the image according to the shooting distance; by acquiring color and depth information, the average depth of the current image block and the reference image block is calculated to determine the stretching ratio, so that the size of the facial images captured at different distances is uniform, facilitating feature comparison. The face detection unit 36 ​​extracts the facial region from the processed depth image, eliminates background interference, and focuses on key recognition areas; the facial feature extraction unit 37 further extracts unique depth features from the facial region as the core basis for identity recognition. The facial feature comparison unit 38 compares the extracted facial features of the person to be identified with the pre-stored facial feature depth information of authorized personnel in the facial storage unit 31, and calculates the feature matching degree. The personnel matching determination unit 39 determines whether the person to be identified is an authorized person based on a preset matching threshold. If the matching degree reaches the threshold, the person is determined to be an authorized person; otherwise, the person is determined to be an unauthorized person. When the person is determined to be an authorized person, the facial recognition unit 3 sends a command to the access control unit 1, and the access control unit 1 performs the door opening operation, allowing the person to access. When the person is determined to be an unauthorized person, the facial recognition unit 3 triggers the alarm unit 2: the alarm unit 2 immediately emits an alarm sound to the scene to deter unauthorized personnel; at the same time, it sends an alarm signal to the central control room and displays an emergency warning message on the control monitor to remind management personnel to handle the situation in a timely manner. If a person who has not undergone facial recognition approaches the access control system, the alarm unit 2 also activates the dual alarm mechanism of the scene and remote alarm to ensure that abnormal situations are detected and handled in a timely manner.

[0023] like Figure 4 As shown, in this embodiment, the optimized depth image rotation unit 34 includes a depth information calculation unit 341, a coordinate transformation unit 342, a local normal vector calculation unit 343, a plane normal vector calculation unit 344, a rotation matrix calculation unit 345, and a perspective distortion correction unit 346. The depth information calculation unit 341 is used to calculate depth information from the planar image captured by the depth image capture unit; The coordinate transformation unit 342 is used to calculate the position of the depth image capturing unit 32 in the coordinate system using the depth information calculated by the depth information calculation unit 341; The local normal vector calculation unit 343 is used to calculate the local normal vector of a pixel using the surrounding information of each pixel calculated by the coordinate transformation unit 342. The plane normal vector calculation unit 344 is used to obtain the normal vector of the entire plane using the local normal vector obtained by the local normal vector calculation unit; The rotation matrix calculation unit 345 is used to calculate the angle between the rotation axis and the depth image to obtain the rotation matrix; The perspective distortion correction unit 346 is used to correct image distortion by applying a rotation transformation based on the position of the depth image capture unit.

[0024] In this technical solution, traditional camera shooting is prone to perspective distortion, such as tilting and deformation of facial images, due to factors such as the angle of the person standing and the camera installation position, affecting the accuracy of subsequent feature extraction and comparison. The depth image rotation unit 34 can effectively correct such distortions through a series of technical processing, restoring the facial image to a standard posture and providing a reliable image basis for accurate recognition. By rotating and transforming the depth image, it ensures that the key features of the face to be recognized are in a unified spatial coordinate system, reducing recognition errors caused by posture differences and improving the adaptability of the face recognition unit 3 to facial images from different angles. After correction, the depth feature information of the facial image is more consistent with the shape of the pre-stored authorized personnel feature template, which can reduce the risk of mismatch or missed match of features caused by image distortion, thereby improving the authentication reliability of the entire access control system.

[0025] Working principle: The depth information calculation unit 341 receives the planar image captured by the depth image capturing unit 32, and extracts and calculates the depth information corresponding to each pixel point from the planar image through a specific algorithm, that is, the actual distance between the point and the camera, transforming the two-dimensional planar image into depth image data containing three-dimensional spatial information. The coordinate transformation unit 342 uses the depth information obtained by the depth information calculation unit 341, combined with the intrinsic and extrinsic parameters of the depth image capturing unit 32, to convert the two-dimensional coordinates of each pixel point in the image into three-dimensional spatial coordinates in the coordinate system of the depth image capturing unit 32, clarifying the position of each pixel point in real space. The local normal vector calculation unit 343, based on the three-dimensional coordinates of each pixel point obtained by the coordinate transformation unit 342, analyzes the spatial positional relationship of the neighboring pixels around the pixel, and calculates the local normal vector of each pixel point through vector operations. The local normal vector can reflect the local tilt direction of the surface where the pixel is located, providing a basis for judging the pose of local facial regions. The planar normal vector calculation unit 344 summarizes the local normal vectors of all pixels obtained by the local normal vector calculation unit 343, and processes these local vectors using statistical analysis or fitting algorithms to extract the normal vector that represents the plane containing the entire facial image. This planar normal vector reflects the overall tilt angle and direction of the face and is a key indicator for judging the degree of perspective distortion in the image. The rotation matrix calculation unit 345 determines the axis to be rotated based on the planar normal vector obtained by the planar normal vector calculation unit 344, and calculates the angle between the rotation axis and the standard facial image plane. Based on the rotation axis and the angle, a corresponding rotation matrix is ​​generated through matrix operations. This matrix contains the transformation parameters required to rotate the distorted image to the standard pose. The perspective distortion correction unit 346 performs rotation transformation processing on the depth image based on the rotation matrix obtained by the rotation matrix calculation unit 345 and the actual position parameters of the depth image acquisition unit 32. By spatially transforming the three-dimensional coordinates of each pixel according to the rotation matrix, the originally tilted facial image is rotated to a state parallel to the standard reference plane, thereby eliminating perspective distortion and outputting a corrected depth image, providing high-quality image data for subsequent depth image stretching, feature extraction and other processing.

[0026] like Figure 4 As shown, in this embodiment, the optimized depth image stretching unit 35 includes an information acquisition unit 351, an average value calculation unit 352, and an expansion / contraction ratio calculation unit 353. The information acquisition unit 351 is used to simultaneously acquire color information and depth information and perform conversion; The average value calculation unit 352 is used to calculate the average value of the color information and depth information of the current block in the current image and the average value of the color information and depth information of the reference block in the reference image based on the color information and depth information acquired by the information acquisition unit 351. The expansion / contraction ratio calculation unit 353 is used to calculate the expansion / contraction ratio based on the average depth information of the current block and the reference block, and to perform the calculation by enlarging or shrinking the reference screen according to the expansion ratio.

[0027] In this technical solution, due to variations in the distance between the person and the depth image capturing unit 32, the captured facial images may vary in size. The depth image stretching unit 35 calculates the expansion / contraction ratio (353 / 353) and performs corresponding processing to adjust facial images captured at different distances to a uniform size, avoiding the impact of image size differences on the accuracy of subsequent feature extraction and comparison. The stretched facial image maintains the same size as the reference image pre-stored in the facial storage unit 31, ensuring that the facial features of both are at the same scale. This provides a reliable basis for accurate comparison by the facial feature comparison unit 38, reducing recognition errors caused by size mismatch. Regardless of whether the person is at a close or distant shooting position, the depth image stretching unit 35 can dynamically adjust the image size to ensure that the face recognition unit 3 can effectively recognize facial images at various distances, improving the flexibility and adaptability of the entire access control system in practical applications.

[0028] Working Principle: The information acquisition unit 351 synchronously collects color and depth information of the face of the person to be identified. Color information helps locate key features of the facial region, while depth information reflects the actual distance between each point on the face and the capturing unit. This unit converts the acquired raw information into standardized data that facilitates subsequent calculations, providing a unified input format for averaging calculations. The averaging calculation unit 352, based on the standardized color and depth information provided by the information acquisition unit 351, divides the currently captured facial image and the authorized personnel standard facial image stored in the facial storage unit 31 into blocks. For the current image, it calculates the average color and depth information of each current block; for the reference image, it similarly calculates the average color and depth information of each corresponding reference block. By calculating the average value in blocks, it can more accurately reflect the feature differences of different regions, providing detailed data support for subsequent ratio calculations. The expansion / contraction ratio calculation unit 353 calculates the ratio of the average depth information of the current block and the reference block obtained by the averaging calculation unit 352; this ratio is the expansion / contraction ratio. If the average depth of the current block is greater than the average depth of the reference block, it indicates that the current shooting distance is farther, and the image needs to be enlarged; if the average depth of the current block is less than the average depth of the reference block, it indicates that the current shooting distance is closer, and the image needs to be reduced. Subsequently, this unit enlarges or reduces the reference image or the current image according to the calculated dilation / scaling ratio of 353, so that the two are consistent in size, and outputs the adjusted image, providing standardized image data for the subsequent processing of the face detection unit 36 ​​and the face feature extraction unit 37.

[0029] Example 2: This embodiment provides an alarm-enabled smart access control system, which, in addition to the technical solutions of the above embodiments, also has the following technical features.

[0030] like Figure 4 As shown, in this embodiment, the optimized face storage unit 31 includes a dynamic update module 311 and a classification storage module 312. The dynamic update module 311 is used to periodically receive the latest facial feature depth information from authorized personnel and update and replace the stored historical information; The classification storage module 312 is used to classify and store the facial feature depth information of authorized personnel according to their permission level, department, and other information, so as to facilitate quick retrieval and access.

[0031] In this technical solution, the facial features of authorized personnel may change over time, such as changes in appearance or hairstyle due to aging. The dynamic update module 311 periodically receives and updates the latest facial feature depth information of authorized personnel, replacing historical information to ensure that the stored information always accurately reflects the current facial features of authorized personnel, avoiding recognition failure or misjudgment due to outdated information. In access control system application scenarios, the number of authorized personnel may be large, and different personnel may have different access levels, department affiliations, and other attributes. The classification storage module 312 classifies and stores facial feature depth information according to these attributes, enabling quick location of information under the relevant category during facial feature comparison, reducing the search scope, improving search speed, and facilitating unified management and maintenance of information for different categories of personnel by administrators.

[0032] Working Principle: The dynamic update module 311 proactively sends information update requests to the system administrator or authorized personnel according to a preset time period, or receives the latest facial feature depth information of authorized personnel proactively uploaded by the administrator. Upon receiving new facial feature depth information, the dynamic update module 311 verifies the information to confirm that it belongs to an authorized person and that the information format meets system requirements. After successful verification, the module replaces the stored historical facial feature depth information of the corresponding authorized person and records the update time and other relevant information, completing the dynamic update of the information. When authorized personnel information is entered into the system, the classification storage module 312 collects attribute information such as the person's permission level and department. Based on this attribute information, the module establishes corresponding classification directories or database table structures, such as categorizing by permission level into high-level, medium-level, and ordinary permissions, and by department into technical, finance, and human resources departments. When storing the facial feature depth information of authorized personnel, the classification storage module 312 stores the information in the corresponding classification directory or database table. When performing facial feature retrieval, the system will instruct the classification storage module 312 to search in the corresponding category based on the relevant attributes of the person to be identified, and quickly retrieve the relevant facial feature depth information for the facial feature comparison unit 38 to compare.

[0033] like Figure 4 As shown, in this embodiment, the optimized version also includes a linkage control unit 4, which is connected to the access control unit 1, the alarm unit 2, and the face recognition unit 3 respectively. When the face recognition unit 3 recognizes an authorized person, the linkage control unit 4 controls the access control unit 1 to open and simultaneously sends the personnel access record to the relevant management system; When alarm unit 2 issues an alarm, linkage control unit 4 can control access control unit 1 to remain locked, and link the on-site monitoring equipment to focus on capturing and recording the alarm area, while synchronizing the relevant information to the situation room center.

[0034] This technical solution breaks down the independent operation of access control, alarm, and facial recognition units. Through unified scheduling by the linkage control unit 4, each unit can respond quickly based on recognition results and alarm status, improving the overall system efficiency and consistency. When alarms are triggered, such as unauthorized intrusion, the linkage control unit 4 can promptly lock access control unit 1 to prevent the spread of illegal intrusion. Simultaneously, it links with monitoring equipment to record the scene, providing evidence for subsequent processing and enhancing the system's security capabilities. When authorized personnel pass through the identification system, the linkage control unit 4 not only controls the access control to open but also simultaneously sends access records to the management system, achieving automated management and archiving of personnel entry and exit, reducing manual intervention, and improving management convenience and accuracy.

[0035] Working Principle: The facial recognition unit 3 identifies personnel. When the personnel matching and confirmation unit 39 determines that the person is authorized, it sends an "authorization passed" signal to the linkage control unit 4. Upon receiving this signal, the linkage control unit 4 immediately issues an opening command to the access control unit 1, which then opens the door, allowing the authorized person to enter. Simultaneously, the linkage control unit 4 compiles the authorized person's identity information, access time, and other data into an access record and sends it to the relevant management system for archiving, achieving automated tracking of personnel entry and exit. When the alarm unit 2 detects an unauthorized person or a person without facial recognition approaching, it sends an "alarm triggered" signal to the linkage control unit 4. Upon receiving this signal, the linkage control unit 4 quickly issues a locking command to the access control unit 1, ensuring that the access control unit 1 remains closed to prevent unauthorized entry. At the same time, the linkage control unit 4 sends a control signal to the on-site monitoring equipment, instructing the monitoring equipment to focus on the alarm area for focused shooting and recording, capturing detailed on-site footage. In addition, the linkage control unit 4 synchronously transmits relevant information such as alarm type, occurrence time, and on-site monitoring screen to the situation room center, enabling managers to understand the on-site situation in a timely manner and take corresponding measures.

[0036] Example 3: This embodiment provides an alarm-enabled smart access control system, which, in addition to the technical solutions of the above embodiments, also has the following technical features.

[0037] like Figures 1-3 As shown, in this embodiment, the optimized access control unit 1 includes a gate 5, an opening and closing mechanism 6, a drive shaft 7, and a transmission mechanism 8. The turnstile 5 has a symmetrically distributed body 51, and the face detection unit 36 ​​is disposed on the body 51; The opening and closing mechanism 6 is located on the opposite surfaces of the two bodies 51, and rotates to open when an authorized person is detected; The drive shaft 7 is rotatably mounted on the machine body 51, and the drive shaft 7 is connected to the opening and closing mechanism 6; The transmission mechanism 8 is connected to the drive shaft 7 and drives the drive shaft 7 and the opening and closing mechanism 6 to operate when an authorized person is identified.

[0038] In this technical solution, by linking with the face recognition unit 3, the opening and closing mechanism 6 rotates and opens only when an authorized person is recognized, ensuring that only authorized personnel can enter, strictly controlling access permissions, and improving the security of the access control system. With the cooperation of the drive shaft 7 and the transmission mechanism 8, stable power and precise transmission are provided for the operation of the opening and closing mechanism 6, making the opening and closing process smooth and preventing the access control system from failing to open or close properly due to mechanical failure, thus ensuring the normal operation of the system. The face detection unit 36 ​​is set on the body 51 of the gate 5, enabling timely face detection when a person approaches, shortening the recognition response time, making the recognition and opening / closing actions more closely integrated, and improving passage efficiency.

[0039] Working principle: When a person approaches the gate 5, the face detection unit 36 ​​on the machine body 51 detects the person's face and transmits the information to the face recognition unit 3 for identification. If the person is identified as authorized, the face recognition unit 3 sends a signal. Upon receiving the signal, the transmission mechanism 8 starts, driving the drive shaft 7 to rotate. The drive shaft 7 drives the connected opening and closing mechanism 6 to rotate and open, allowing the authorized person to pass. After the person passes, the transmission mechanism 8 drives the drive shaft 7 to rotate in the opposite direction, causing the opening and closing mechanism 6 to reset and close, restoring the initial state.

[0040] like Figures 1-3 As shown, in this embodiment, the optimized opening and closing mechanism 6 includes: The disc 61 is mounted on the drive shaft 7 and fits against the inner wall of the machine body 51; Ear plates 62 are disposed on the disk 61. There are multiple ear plates 62, which are evenly distributed along the circumferential direction. A through hole 63 is provided on the ear plate 62; Bolt 64 is inserted through hole 63; Nut 65, threaded, is attached to one end of bolt 64; A connecting plate 66 is located between two ear plates 62, and a bolt 64 passes through the connecting plate 66, with the connecting plate 66 rotatably mounted on the bolt 64; The gate arm 67 is mounted on the connecting plate 66.

[0041] In this technical solution, the ear plate 62 and the connecting plate 66 are tightly fixed by bolts 64 through holes 63 and threaded nuts 65, preventing loosening or displacement during the rotation of the gate arm 67, ensuring the structural stability of the opening and closing mechanism 6 in long-term use, and reducing the occurrence of mechanical failures. The connecting plate 66 is rotatably mounted on the bolts 64, allowing the gate arm 67 to rotate flexibly around the bolts 64. Operators can adjust the angle of the gate arm 67 as needed, and the gate arm 67 is secured to the adjusted angle by the cooperation of the nuts 65 and bolts 64. The components are assembled by detachable connections such as bolts 64 and nuts 65. When a component is damaged, it can be easily disassembled and replaced, reducing maintenance costs and difficulty, and ensuring the continuous normal operation of the opening and closing mechanism 6.

[0042] Working principle: When the face recognition unit 3 recognizes an authorized person and sends an open signal, the transmission mechanism 8 drives the drive shaft 7 to rotate, which in turn drives the connected disc 61 to rotate. As the disc 61 rotates, the gate arm 67 moves along with the connecting plate 66 via the ear plate 62, bolt 64, and connecting plate 66, allowing personnel to pass through the access control system smoothly. Multiple gate arms 67 are available; when one gate arm 67 rotates to allow personnel to pass, another gate arm 67 immediately takes over, completing one opening and closing cycle.

[0043] like Figures 1-3 As shown, in this embodiment, the turnstiles 5 are optimally spaced to form at least one set of turnstile 5 channels.

[0044] In this technical solution, different numbers of turnstiles (5-channel) are set up according to the traffic volume of the location. In densely populated areas, such as large office buildings and stations, multiple sets of turnstiles (5-channel) can simultaneously accommodate multiple people, avoiding congestion; in less crowded areas, one set of turnstiles (5-channel) is sufficient to meet basic passage needs, achieving rational resource allocation. Multiple sets of turnstiles (5-channel) working in parallel can simultaneously allow multiple authorized personnel to pass, reducing waiting time, especially during peak hours, significantly improving overall passage efficiency. When one set of turnstiles (5-channel) malfunctions or requires maintenance, the other turnstiles (5-channel) can continue to operate normally, ensuring the continuous operation of the access control system, not affecting normal personnel passage, and improving system reliability and fault tolerance.

[0045] The embodiments of this application have been described above with reference to the accompanying drawings. Unless otherwise specified, the embodiments and features in the embodiments of this application can be combined with each other. This application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. An intelligent access control system with alarm function, characterized in that: It includes an access control unit (1), an alarm unit (2), and a face recognition unit (3); The access control unit (1) is used to allow access when an authorized person is identified; The alarm unit (2) is used to generate an alarm when an unauthorized person is identified, and when a person who has not undergone facial recognition approaches the access control system, it sends an alarm sound to the scene and to the center of the situation room, and displays an emergency warning message on the control monitor. The face recognition unit (3) includes a face storage unit (31), a depth image capture unit (32), a depth image correction unit (33), a depth image rotation unit (34), a depth image stretching unit (35), a face detection unit (36), a face feature extraction unit (37), a face feature comparison unit (38), and a personnel matching and determination unit (39). Among them, the face storage unit (31) is used to store facial feature depth information; the depth image capturing unit (32) is used to capture facial depth images; the depth image correction unit (33) is used to correct depth value errors; the depth image rotation unit (34) is used to perform image rotation transformation and face alignment; the depth image stretching unit (35) is used to enlarge and shrink facial images according to the image capturing distance; the face detection unit (36) is used to extract the facial part of the depth image; the face feature extraction unit (37) is used to extract facial features from the depth image; the face feature comparison unit (38) is used to compare the data stored in the face storage unit (31); and the personnel matching determination unit (39) is used to determine the degree of personnel matching.

2. The intelligent access control system with alarm as described in claim 1, characterized in that, The depth image rotation unit (34) includes a depth information calculation unit (341), a coordinate transformation unit (342), a local normal vector calculation unit (343), a plane normal vector calculation unit (344), a rotation matrix calculation unit (345), and a perspective distortion correction unit (346). The depth information calculation unit (341) is used to calculate depth information from the planar image captured by the depth image capture unit; The coordinate transformation unit (342) is used to calculate the position of the depth image capturing unit (32) in the coordinate system using the depth information calculated by the depth information calculation unit (341); The local normal vector calculation unit (343) is used to calculate the local normal vector of a pixel using the surrounding information of each pixel calculated by the coordinate transformation unit (342); The plane normal vector calculation unit (344) is used to obtain the normal vector of the entire plane using the local normal vector obtained by the local normal vector calculation unit; The rotation matrix calculation unit (345) is used to calculate the angle between the rotation axis and the depth image to obtain the rotation matrix; The perspective distortion correction unit (346) is used to correct image distortion by applying rotation transformation based on the position of the depth image capture unit.

3. The intelligent access control system with alarm as described in claim 1, characterized in that, The depth image stretching unit (35) includes an information acquisition unit (351), an average value calculation unit (352), and an expansion / contraction ratio calculation unit (353). The information acquisition unit (351) is used to acquire color information and depth information simultaneously and convert them; The average value calculation unit (352) is used to calculate the average value of the color information and depth information of the current block in the current image and the average value of the color information and depth information of the reference block in the reference image based on the color information and depth information obtained by the information acquisition unit (351). The expansion / contraction ratio calculation unit (353) is used to calculate the expansion (353) / contraction ratio based on the average depth information of the current block and the reference block, and to perform the calculation by enlarging or shrinking the reference screen according to the expansion ratio.

4. The intelligent access control system with alarm as described in claim 1, characterized in that, The face storage unit (31) includes a dynamic update module (311) and a classification storage module (312). The dynamic update module (311) is used to periodically receive the latest facial feature depth information of authorized personnel and update and replace the stored historical information; The classification storage module (312) is used to classify and store the facial feature depth information of authorized personnel according to their permission level, department, and other information, so as to facilitate quick retrieval and access.

5. The intelligent access control system with alarm as described in claim 1, characterized in that, It also includes a linkage control unit (4), which is connected to the access control unit (1), the alarm unit (2) and the face recognition unit (3) respectively; when the face recognition unit (3) recognizes an authorized person, the linkage control unit (4) controls the access control unit (1) to open and simultaneously sends the personnel access record to the relevant management system; When the alarm unit (2) issues an alarm, the linkage control unit (4) can control the access control unit (1) to remain locked, and link the on-site monitoring equipment to take key pictures and record the alarm area, while synchronizing the relevant information to the situation room center.

6. The intelligent access control system with alarm function according to claim 1, characterized in that, The access control unit (1) includes a gate (5), an opening and closing mechanism (6), a drive shaft (7), and a transmission mechanism (8). The gate (5) has a symmetrically distributed body (51), and the face detection unit (36) is disposed on the body (51); The opening and closing mechanism (6) is located on the opposite surfaces of the two bodies (51) and rotates open when an authorized person is detected; The drive shaft (7) is rotatably mounted on the body (51), and the drive shaft (7) is connected to the opening and closing mechanism (6); The transmission mechanism (8) is connected to the drive shaft (7) and drives the drive shaft (7) and the opening and closing mechanism (6) to operate when an authorized person is identified.

7. The intelligent access control system with alarm as described in claim 6, characterized in that, The opening and closing mechanism (6) includes: A disc (61) is disposed on the drive shaft (7) and fits against the inner wall of the body (51); Ear plates (62) are disposed on the disk (61), and there are multiple ear plates (62) evenly distributed in the circumferential direction; A through hole (63) is provided on the ear plate (62); Bolt (64) is inserted into the through hole (63); A nut (65) is threaded onto one end of the bolt (64); A connecting plate (66) is located between the two ear plates (62), the bolt (64) passes through the connecting plate (66), and the connecting plate (66) is rotatably mounted on the bolt (64); The gate arm (67) is mounted on the connecting plate (66).

8. The intelligent access control system with alarm function according to claim 6, characterized in that, The turnstiles (5) are spaced apart to form at least one set of turnstile (5) channels.

Images