Community security system using stranger face recognition technology

Abstract

The present application discloses a community security system using stranger face recognition technology, which comprises a main control electronic device and a plurality of cameras arranged in the community. According to the design of the present application, each camera is used to capture the image of a subject to generate a subject image. Then, the main control electronic device sequentially performs a face detection operation, a face feature extraction operation and a labeling operation on the subject image, thereby generating a labeled face feature image with a label (such as RFID). In this way, after a period of time, a plurality of labeled face feature images can be obtained. Then, the main control electronic device performs an image comparison operation on a plurality of the labeled face feature images, thereby obtaining a resident face template image containing a specific face feature, and giving the resident face template image a resident label, completing the production of a resident face template image.

Description

[0001] Technical Field: This invention relates to the technical field of community security, and more particularly to a community security system that uses stranger facial recognition technology. Background technology:

[0002] With the rapid advancement of image recognition technology, biometric identification systems are being integrated into security devices such as electronic door locks and intercoms. Biometric identification uses inherent physiological characteristics of the human body as the basis for individual identification, such as iris, face, voiceprint, and fingerprint. Taking face recognition as an example, its foundation lies in the creation of a facial template image. For instance, to unlock an Apple phone using Face ID technology, the user must first operate the Face ID application and successfully create their own facial template image.

[0003] Therefore, some upscale communities have already introduced advanced community security systems that use facial recognition technology to identify everyone entering the community, effectively preventing strangers from entering. However, before such advanced community security systems can operate efficiently, the management committee or security system supplier must first create facial template images for each resident in the community and label the recorded facial template images according to each resident's personal information (such as name or apartment number).

[0004] It's understandable that tagging facial template images directly maps X's facial template image to X's personal data, where X refers to resident A, resident B, etc. However, this approach has the following problems:

[0005] (1) There may be concerns about violating regulations;

[0006] (2) Not every resident wants their facial image to be made public; and

[0007] (3) There are many obstacles to the production of facial template images for each household in a short period of time, such as the fact that some residents live abroad.

[0008] In conclusion, it is clear that existing community security systems utilizing facial recognition technology have practical shortcomings and need improvement. Therefore, the inventors of this invention have diligently researched and developed a community security system employing unfamiliar facial recognition technology. Summary of the Invention:

[0009] The main objective of this invention is to provide a community security system that uses stranger facial recognition technology. This system can automatically generate facial template images of each resident in the community and assign a corresponding resident tag (e.g., a custom ID) to each resident's facial template image. It does not require the management committee or system supplier to record facial template images of each resident in the community, nor does it require the use or disclosure of residents' personal information (e.g., name or room number).

[0010] To achieve the above objectives, the present invention proposes an embodiment of a community security system employing stranger facial recognition technology, comprising:

[0011] Multiple cameras are installed within a community, and each camera has a field of view; and

[0012] A main control electronic device, coupled to the plurality of cameras, includes a microprocessor and a memory; wherein the memory stores an application program edited using a programming language, and the microprocessor executes the application program by accessing the memory, thereby enabling the following functions:

[0013] Each of the cameras is controlled to perform an image capturing operation on a first subject that enters its camera range, thereby generating a subject image;

[0014] Identify a face region from the main image;

[0015] A facial feature extraction operation is performed on the facial region to obtain a facial feature image;

[0016] Assign a label to the facial feature image to generate a labeled facial feature image, and store the labeled facial feature image in the memory; and

[0017] An image comparison operation is performed on a plurality of labeled facial feature images stored in the memory, so that when multiple labeled facial feature images contain the same facial feature, a resident facial template image containing the facial feature is created and stored in the memory, and a resident label is assigned to the resident facial template image.

[0018] In one embodiment, the main control electronic device is an electronic device selected from the group consisting of notebook computers, desktop computers, all-in-one computers, industrial computers, and cloud servers.

[0019] In one embodiment, the application includes a plurality of subroutines, and the plurality of subroutines include:

[0020] A camera control program containing instructions for configuring the microprocessor to control each of the cameras to perform the image capturing operation;

[0021] A face detection program contains instructions to configure the microprocessor to perform a face detection operation on the subject image, thereby finding the face region in the subject image;

[0022] A feature extraction program contains instructions to configure the microprocessor to perform the facial feature extraction operation on the face region, thereby obtaining the facial feature image;

[0023] An image tagging program contains instructions to configure the microprocessor to assign the label to the facial feature image and the resident label to the resident facial template image;

[0024] A comparison program, containing instructions to configure the microprocessor to perform the image comparison operation on a plurality of labeled facial feature images stored in the memory; and

[0025] A template image creation program contains instructions to configure the microprocessor to create a resident face template image containing the same face feature when multiple labeled face feature images simultaneously contain the same face feature, and to assign a resident label to the resident face template image.

[0026] In one embodiment, the main control electronic device further includes a first communication interface, such that an electronic device uses a second communication interface to connect with the first communication interface of the main control electronic device.

[0027] In one feasible embodiment, the plurality of subroutines further include:

[0028] A moving object monitoring program contains instructions to configure the microprocessor to control a plurality of cameras to perform a moving object detection operation on the first subject corresponding to the subject image after receiving a subject image transmitted by the camera, thereby recording a moving path of the first subject in the community in the memory.

[0029] A stranger monitoring program contains instructions to configure the microprocessor to perform a face recognition operation on a single image of the subject transmitted by the camera and multiple resident face template images stored in the memory, thereby determining whether the first subject corresponding to the main image is a stranger; and

[0030] An alert program contains instructions to configure the microprocessor to transmit information about a stranger, including the movement path, to the electronic device via a first communication interface when a stranger is detected, and simultaneously display the first entity as a stranger on a monitor screen of the main control electronic device.

[0031] In one embodiment, the main control electronic device also publishes the stranger information on multiple public display devices within the community, and marks the first subject as a stranger on the public display devices.

[0032] In one embodiment, the electronic device is selected from the group consisting of mobile electronic devices of community residents, video door access devices of community residents, mobile electronic devices of community guards, and mobile electronic devices of community managers.

[0033] In yet another feasible embodiment, the plurality of subroutines further include:

[0034] A tiered procedure, containing instructions to configure the microprocessor to simultaneously enable a tiered mechanism when executing the stranger monitoring program; and

[0035] A search program contains instructions to configure the microprocessor to compare an externally input image of a subject to be searched with a plurality of labeled facial feature images stored in the memory, thereby finding a labeled facial feature image that matches the image of the subject to be searched.

[0036] In one embodiment, the hierarchical mechanism includes:

[0037] After the first subject corresponding to a subject image transmitted by the camera is determined to be a stranger, the number of times the first subject appears in the community is confirmed.

[0038] If the occurrence count is 0, a severe level message is displayed on the monitor screen, and a majority of the community guards are notified to intercept the first subject;

[0039] When the occurrence frequency is between 0 and N times, a moderate-level message is displayed on the monitor screen; and

[0040] When the number of occurrences is between N and M, a mild level message is displayed on the monitor screen, where N and M are both positive integers.

[0041] In one embodiment, after receiving a subject image transmitted by the camera, the microprocessor simultaneously records an entry time and a departure time of the first subject corresponding to the subject image within the community. Attached image description:

[0042] Figure 1 This is a block diagram of a community security system using stranger facial recognition technology according to the present invention;

[0043] Figure 2 This is an application architecture diagram of the community security system using stranger facial recognition technology according to the present invention; and

[0044] Figure 3 for Figure 2 The block diagram of the main control electronic device 12 shown.

[0045] Figure label:

[0046] 1: Community security system using facial recognition technology for strangers

[0047] 11: Camera

[0048] 12: Main control electronic device

[0049] 12P: Microprocessor

[0050] 12M: Memory

[0051] 12M1: Camera Control Program

[0052] 12M2: Face Detection Program

[0053] 12M3: Feature Extraction Program

[0054] 12M4: Image Tagging Program

[0055] 12M5: Comparison Procedure

[0056] 12M6: Template Image Creation Program

[0057] 12M7: Moving Object Monitoring Program

[0058] 12M8: Stranger Monitoring Program

[0059] 12M9: Alert Procedure

[0060] 12MA: Grading Procedure

[0061] 12MB: Search Program

[0062] 12W: First Communication Interface

[0063] 12D: Monitor screen

[0064] 2: Electronic devices Detailed implementation method:

[0065] To more clearly describe the community security system using stranger facial recognition technology proposed in this invention, the preferred embodiments of the invention will be described in detail below with reference to the accompanying drawings.

[0066] Please see Figure 1 This is a block diagram of a community security system employing stranger facial recognition technology according to the present invention. Furthermore, Figure 2 This is an application architecture diagram of the community security system using stranger facial recognition technology according to the present invention. Figure 1 and Figure 2 As shown, the community security system 1 employing stranger facial recognition technology of the present invention mainly includes, in terms of architecture, a plurality of cameras 11 and a main control electronic device 12. The plurality of cameras 11 are distributed throughout the community, and each camera 11 has a camera range. The main control electronic device 12 is coupled to the plurality of cameras 11. In one embodiment, the main control electronic device 12 is the central control computer of the community's management center; therefore, it can be a laptop, desktop computer, all-in-one computer, industrial computer, or cloud server.

[0067] Figure 3 for Figure 2 The block diagram of the main control electronic device 12 is shown. Figures 1 to 3 As shown, the main control electronic device 12 includes a microprocessor 12P and a memory 12M, wherein the memory 12M stores an application program edited using a programming language, enabling the microprocessor 12P to execute the application program by accessing the memory 12M, thereby activating the community security monitoring function. Figure 3 As shown, the application contains a plurality of subroutines, which include: a camera control program 12M1, a face detection program 12M2, a feature extraction program 12M3, an image tagging program 12M4, a comparison program 12M5, a template image creation program 12M6, a moving object monitoring program 12M7, a stranger monitoring program 12M8, an alert program 12M9, and a rating program 12MA.

[0068] The search program is 12MB.

[0069] like Figures 1 to 3As shown, the camera control program 12M1 contains instructions to configure the microprocessor 12P to control each of the cameras 11. Therefore, during normal system operation, the microprocessor 12P can control each camera 11 to perform an image capture operation on a first subject entering its field of view, thereby generating a subject image. On the other hand, the face detection program 12M2 contains instructions to configure the microprocessor 12P to perform a face detection operation on the subject image, and the feature extraction program 12M3 contains instructions to configure the microprocessor 12P to perform facial feature extraction on the face region. Therefore, after receiving a subject image transmitted by the camera 11, the microprocessor 12P first performs a face detection operation on the subject image to identify a face region (i.e., RoI region) within the subject image. Then, the microprocessor 12P performs facial feature extraction on the face region to obtain the face feature image.

[0070] To explain in more detail, the image labeling program 12M4 instructions are used to configure the microprocessor 12P to perform a labeling operation on an image, and the comparison program 12M5 configures the microprocessor 12P to perform a comparison operation on a plurality of images. Therefore, after obtaining the facial feature image, the microprocessor 12P first assigns a label (e.g., a custom ID) to the facial feature image, and then the labeled facial feature image is stored in the memory 12M. It can be understood that after the community security system 1 using stranger facial recognition technology of the present invention is installed in a community and operates for a period of time, the plurality of cameras 11 capture K resident images, and the microprocessor 12P correspondingly produces K labeled facial feature images. It should be understood that if the community has a total of J residents, then images of fewer than or equal to J residents will be included in the K resident images. At this point, it's conceivable that among the K labeled facial feature images, multiple labeled facial feature images may simultaneously contain the same facial feature. In other words, if resident A enters and exits the community multiple times, and multiple images of the resident are taken, resulting in L labeled facial feature images, naturally, these L labeled facial feature images will simultaneously contain the facial features of resident A.

[0071] Therefore, the present invention designs the comparison program 12M5 to contain instructions for configuring the microprocessor 12 to perform an image comparison operation on multiple images, and designs the template image creation program 12M6 to contain instructions for configuring the microprocessor 12P to perform a template image creation operation. According to this design, after the system of the present invention has run for a period of time to obtain K labeled facial feature images, the microprocessor 12P performs an image comparison operation on the K labeled facial feature images stored in the memory 12M to determine whether there are multiple (e.g., L) labeled facial feature images that simultaneously contain the same facial feature. If such a situation occurs, the microprocessor 12P creates a resident facial template image containing the facial feature and assigns a resident label to the resident facial template image. In short, after the community security system 1 using stranger facial recognition technology of the present invention is installed in a community and has been running for a period of time, the present invention can automatically generate facial template images of each resident in the community and give the facial template images of each resident a corresponding resident tag (such as a custom ID). It is not necessary to record facial template images of each resident in the community through the management committee or system supplier, nor is it necessary to use or disclose the resident's personal information (such as name or room number).

[0072] Thus, after storing facial template images of the vast majority (e.g., over 60%) of the residents in the 12M memory database, the community security system 1 of the present invention can begin to use facial recognition to monitor non-residents appearing in the community, such as cleaning staff, visitors, and strangers. Therefore, as Figures 1 to 3 As shown, the present invention designs the moving object monitoring program 12M7, which contains instructions to configure the microprocessor 12P to perform a moving object detection function; the stranger monitoring program 12M8, which contains instructions to configure the microprocessor 12P to perform a stranger monitoring function; and the alert program 12M9, which contains instructions to configure the microprocessor 12P to perform an alert function. More specifically, when a non-resident enters the community, the camera 11 captures an image of the non-resident. Then, after receiving the image from the camera 11, the microprocessor 12P controls multiple cameras 11 to perform a moving object detection operation on the subject (i.e., the non-resident) corresponding to the image, thereby recording the movement path of the first subject within the community in the memory 12M.

[0073] Simultaneously, the microprocessor 12P performs a face recognition operation on the main image transmitted by the camera 11 and a plurality of resident face template images stored in the memory 12M, thereby determining whether the subject corresponding to the main image is a stranger (i.e., a non-resident). It is worth noting that the main control electronic device 12 further includes a first communication interface 12W, allowing an electronic device 2 to connect with the first communication interface 12W of the main control electronic device 12 via its second communication interface. In one embodiment, the electronic device 2 can be a community resident's mobile electronic device, a community resident's video door station, a community security guard's mobile electronic device, and / or the community manager's mobile electronic device. Accordingly, when a stranger is detected, the microprocessor 12P transmits information about a stranger, including the movement path, to the electronic device 2 via the first communication interface 12W, and simultaneously displays the first subject as a stranger (i.e., a stranger) on a monitor screen 12D of the main control electronic device 12.

[0074] However, not all non-residents for whom the aforementioned resident facial template images have not been created are strangers; they could also be cleaning staff or visitors. Therefore, in order to classify (classify) non-residents, such as... Figures 1 to 3 As shown, the hierarchical program 12MA of this invention contains instructions to configure the microprocessor 12P so that it enables a hierarchical mechanism simultaneously when executing the stranger monitoring program 12M8. According to the design of this invention, the hierarchical mechanism includes:

[0075] After the subject (i.e., non-resident) corresponding to the subject image transmitted by the camera 11 is determined to be a stranger, the number of times the subject appears in the community is then confirmed.

[0076] If the occurrence count is 0, a severe level message is displayed on a monitor screen 12D of the main control electronic device 12, and a majority of the community guards are notified to intercept the subject.

[0077] When the occurrence frequency is between 0 and N times (e.g., 100 times), a moderate-level message is displayed on the monitor screen; and

[0078] If the number of occurrences is between N and M (e.g., 500 times), a mild level message is displayed on the monitor screen.

[0079] For example, the aforementioned severe level message is a red warning message, indicating that the non-resident is a stranger who has never appeared before. In this case, the microprocessor 12P will transmit information about the stranger, including the movement path, to the electronic device 2 via the first communication interface 12W, such as the mobile electronic device of a community resident and / or the mobile electronic device of a community guard. Furthermore, the main control electronic device 12 will also publish the stranger information on multiple public display devices within the community, and mark the first entity as a stranger on the public display devices. On the other hand, the aforementioned moderate level message is a yellow warning message, indicating that this non-resident may be a frequent visitor to a resident in the community, such as a relative, friend, or tutor. Moreover, the aforementioned mild level message is a green warning message, indicating that this non-resident appears in the community quite frequently, and may be a cleaner or community manager.

[0080] To further clarify, regardless of whether the person is a resident or not, when any entity enters the community, the camera 11 generates a corresponding image of that entity. The microprocessor 12P can then sequentially perform face detection, facial feature extraction, and face recognition operations on this image to confirm whether the entity is a resident of the community. Simultaneously, the microprocessor 12P also records the entry and exit times of the entity corresponding to the image within the community. Furthermore, if the entity is a non-resident, as mentioned above, the microprocessor 12P then performs a classification (grading) operation to determine whether the entity is a cleaning staff member (community manager), a visitor, or a stranger appearing for the first time. Alternatively, for one-time visitors, such as delivery drivers or relatives, residents can submit basic visitor information in advance and apply for a short-term QR code. This QR code is then provided to the visitor and has a short validity period (e.g., valid for one hour, after which it expires). When the visitor enters the community, they can scan the code for basic freedom of movement and passage. The system will automatically identify the visitor as a visitor rather than a resident.

[0081] Furthermore, the present invention also includes instructions in the 12MB search program to configure the microprocessor 12P to perform a search function, comparing an externally input image of a subject to be searched with a plurality of labeled facial feature images stored in the memory 12MB, thereby finding a labeled facial feature image that matches the image of the subject to be searched. Based on this, the movement path of the subject image in the present invention is not limited to recording strangers, but can be fully recorded for all subjects entering and leaving the community and visible to the camera 11. Moreover, in elevator environments, a floor display device can be added to display floor images or floor information in the digital files captured by the camera 11, thereby assisting in the recording of floor information and further refining the movement path of the subject image. For example, when a theft or missing person case occurs, a specific image can be input into the main control electronic device 12, causing the microprocessor 12P to sequentially perform a face detection operation, a facial feature extraction operation, and a face recognition operation on the specific image, thereby confirming whether the person in the specific image has entered / exited the community or entered / exited certain floors. If so, relevant data including movement path, entry time, and exit time can be retrieved (i.e., the aforementioned stranger information). In addition, the images captured by this invention can be further processed and post-processed to assist in face recognition, movement trajectory generation, floor trajectory recording, and the classification and searching of other related images.

[0082] Thus, the above fully and clearly describes a community security system employing stranger facial recognition technology according to the present invention. Specifically, the community security system employing stranger facial recognition technology of the present invention can automatically generate a facial template image of each resident in the community and assign a corresponding resident tag (e.g., a custom ID) to the resident's facial template image. It does not require the management committee or system supplier to record facial template images of each resident in the community, nor does it require the use or disclosure of residents' personal information (e.g., name or room number).

[0083] It must be emphasized that the above detailed description is a specific description of feasible embodiments of the present invention. However, the embodiments are not intended to limit the scope of the claims of the present invention. All equivalent implementations or modifications that do not depart from the spirit of the present invention should be included in the scope of the claims of this case.

Claims

1. A community security system employing stranger facial recognition technology, characterized in that, include: A plurality of cameras are installed within a community, each camera having a field of view; and a main control electronic device is coupled to the plurality of cameras and includes a microprocessor and a memory; wherein the memory stores an application program written in a programming language, which the microprocessor executes by accessing the memory, thereby enabling the following functions: controlling each camera to perform an image capture operation on a first subject entering its field of view, thereby generating a subject image; identifying a face region from the subject image; performing a facial feature extraction operation on the face region, thereby obtaining a facial feature image; assigning a label to the facial feature image to generate a labeled facial feature image, and storing the labeled facial feature image in the memory; and performing an image comparison operation on a plurality of labeled facial feature images stored in the memory, thereby creating and storing a resident facial template image containing the same facial feature in the memory when multiple labeled facial feature images simultaneously contain the same facial feature, and assigning a resident label to the resident facial template image; The application includes multiple subroutines, which include: a camera control program containing instructions to configure the microprocessor to control each camera to perform the image capturing operation; a face detection program containing instructions to configure the microprocessor to perform a face detection operation on the subject image, thereby finding the face region in the subject image; a feature extraction program containing instructions to configure the microprocessor to perform the facial feature extraction operation on the face region, thereby obtaining the face feature image; an image tagging program containing instructions to configure the microprocessor to assign a tag to the face feature image and a resident tag to the resident face template image; a comparison program containing instructions to configure the microprocessor to perform the image comparison operation on multiple labeled face feature images stored in the memory; and a template image creation program containing instructions to configure the microprocessor to create a resident face template image containing the same face feature when multiple labeled face feature images simultaneously contain the same face feature, and to assign a resident tag to the resident face template image.

2. The community security system employing stranger facial recognition technology according to claim 1, characterized in that, The main control electronic device is selected from the group consisting of laptops, desktop computers, all-in-one computers, industrial computers, and cloud servers.

3. The community security system employing stranger facial recognition technology according to claim 2, characterized in that, The main control electronic device further includes a first communication interface, enabling an electronic device to connect with the first communication interface of the main control electronic device via a second communication interface.

4. The community security system employing stranger facial recognition technology according to claim 3, characterized in that, The plurality of subroutines further include: a moving object monitoring program containing instructions to configure the microprocessor to control a plurality of cameras to perform a moving object detection operation on the first subject corresponding to the subject image after receiving a subject image transmitted by the camera, thereby recording a movement path of the first subject in the community in the memory; a stranger monitoring program containing instructions to configure the microprocessor to perform a face recognition operation on a subject image transmitted by the camera and a plurality of resident face template images stored in the memory, thereby determining whether the first subject corresponding to the subject image is a stranger; and an alert program containing instructions to configure the microprocessor to transmit information about a stranger, including the movement path, to the electronic device through a first communication interface when a stranger is detected, and simultaneously display the first subject as a stranger on a monitor screen of the main control electronic device.

5. The community security system employing stranger facial recognition technology according to claim 4, characterized in that, The main control electronic device also publishes the stranger information on multiple public display devices within the community, and marks the first subject as a stranger on the public display devices.

6. The community security system employing stranger facial recognition technology according to claim 5, characterized in that, The electronic device is selected from the group consisting of mobile electronic devices of community residents, video doorbell systems of community residents, mobile electronic devices of community security guards, and mobile electronic devices of community executives.

7. The community security system employing stranger facial recognition technology according to claim 5, characterized in that, The plurality of subroutines further include: a hierarchical procedure containing instructions to configure the microprocessor to enable a hierarchical mechanism while executing the stranger monitoring procedure; and a search procedure containing instructions to configure the microprocessor to compare an externally input image of a subject to be searched with a plurality of labeled facial feature images stored in the memory, thereby finding a labeled facial feature image that matches the image of the subject to be searched.

8. The community security system employing stranger facial recognition technology according to claim 7, characterized in that, The classification mechanism includes: after the first subject corresponding to a subject image transmitted by the camera is determined to be a stranger, confirming the number of times the first subject appears in the community; if the number of appearances is 0, displaying a severe level message on the monitor screen and notifying a majority of the community guards to intercept the first subject; if the number of appearances is between 0 and N, displaying a moderate level message on the monitor screen; and if the number of appearances is between N and M, displaying a mild level message on the monitor screen, where N and M are both positive integers.

9. The community security system employing stranger facial recognition technology according to claim 8, characterized in that, After receiving a subject image transmitted by the camera, the microprocessor simultaneously records an entry time and a departure time of the first subject corresponding to the subject image within the community.

Images