Intelligent property access control system

Abstract

The invention discloses an intelligent property access control system which comprises a multi-mode biological recognition module, an environment sensing module, a data processing module, a data storage module and a data management module, and the multi-mode biological recognition module is integrated with a camera module, a fingerprint module and an iris module. Biological information is recognized through the data processing module, high-precision recognition in a sheltered scene is achieved, the iris recognition mode is automatically switched to when the face is sheltered, a weighted fusion strategy is adopted, the recognition passing rate is effectively increased, meanwhile, details of a sheltered part are automatically generated in cooperation with a sheltered object detection algorithm, and the recognition efficiency is improved. A dual-identification guarantee mechanism is realized, the identification accuracy under the shielding scene is further improved, and the reliability of the access control system is ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of access control systems, and particularly relates to a smart property access control system. BACKGROUND

[0002] The property access control system is a system for managing and controlling the entrance and exit of a specific area or building based on modern information technology and security technology. It has the functions of identity recognition, access control, event recording, alarm prompt, etc., and can effectively improve the security level of the property and protect the safety of personnel and property. The property access control system verifies the identity of the access personnel through cards, passwords, biological characteristics, etc., and only allows authorized personnel to pass. The system can quickly and accurately identify the identity of the user, refuse unauthorized personnel to enter, and ensure the safety of the area.

[0003] The property access control recognition system in the prior art generally uses a camera for face recognition. When the user wears a shielding object (helmet, mask, scarf, etc.), the shielding object covers the nose, mouth and other key feature areas of the face, which will directly lead to the system being unable to effectively extract complete face feature information, and thus greatly reduces the probability of face recognition passing, making it difficult to meet the identity verification needs in different use scenarios. SUMMARY

[0004] The present application aims to at least solve one of the technical problems in the related art to some extent.

[0005] To this end, the purpose of the present application is to propose a smart property access control system, which adopts a weighted fusion strategy combined with shielding repair to double-optimize biological feature recognition, effectively solving the problem of low biological feature recognition accuracy in the shielding scenario, and improving the probability of face recognition passing.

[0006] In order to achieve the above object, the application provides a smart property access control system, comprising a multi-modal biological recognition module, an environment perception module, a data processing module, a data storage module and a data management module, wherein the multi-modal biological recognition module integrates a camera module, a fingerprint module and an iris module, is used for collecting personnel biological characteristics, including facial features, fingerprints and irises, automatically switches to iris recognition in a shielding scene, and transmits recognition information to the data processing module in real time; the environment perception module integrates an ambient light sensor and a shielding detection camera, is used for monitoring ambient light intensity and shielding object types in real time, and transmits data to the data processing module in real time; the data processing module is used for processing biological characteristic data and shielding object data, and analyzes and compares collected data by using a multi-modal feature fusion algorithm and a shielding object detection algorithm, and identifies personnel identity; the data storage module is used for storing personnel biological characteristic data, environment perception data and identification record information, and simultaneously provides data backup and recovery functions; and the data management module is used for managing system data, and inputting and querying personnel information.

[0007] In addition, the smart property access control system according to the application can have the following additional technical features: Specifically, the data processing module adopts a shielding object detection algorithm to identify shielding object types and positions, and generates a complete facial image according to an unshielded area, and the shielding object detection algorithm specifically comprises the following steps: pre-processing input data to reduce noise interference, identifying shielding object categories, first constructing a shielding object detection model, then extracting image multi-scale features, outputting category labels and position coordinates of the shielding object, merging overlapping boundary boxes, retaining the most possible target shielding object, marking a shielded facial area to provide a mask for image repair, separating an unshielded facial area as a reference for generating a complete image, generating a complete facial image based on the unshielded area, restoring details of a shielded part, optimizing a generation result, and obtaining biological facial features.

[0008] Specifically, the data processing module adopts a multi-modal feature fusion algorithm to analyze and compare collected biological data, and identify personnel identity, wherein the multi-modal fusion algorithm adopts a weighted fusion strategy, combines face and iris recognition confidence, and dynamically adjusts weights, and specifically comprises the following steps: respectively extracting face features and iris features, when user face features are shielded by a shielding object, processing face image data by the shielding object detection algorithm, and automatically switching to an iris recognition mode, dynamically adjusting weight proportions of two modes according to real-time face and iris recognition confidence and shielding interference factors, multiplying adjusted mode weights and corresponding confidence, summing, generating a comprehensive recognition score, and finally determining personnel identity legality by comparing with a preset threshold.

[0009] Specifically, the data storage module integrates a memory and a database, wherein the memory is used to locally save system data, including personnel biometric data, environment perception data, identification record information and access control permission setting information, and the database has a data synchronization function and is periodically synchronized with a cloud data center to realize data backup and recovery.

[0010] Specifically, the access control system further comprises an access controller configured to receive the identification result of the data processing module and control the opening or closing of the electromagnetic lock according to a preset rule.

[0011] Specifically, the access control system further comprises a card reader connected to the access controller and configured to read the access card information held by a user and transmit the read information to the access controller, and the access controller is configured to compare the access card information with the personnel information stored in the data management module after receiving the access card information, and control the electromagnetic lock to open if the information matches and meets the preset access control permission rule.

[0012] Specifically, the access control system further comprises an alarm connected to the access controller and configured to trigger the alarm to send an alarm signal when the access controller detects an abnormal situation.

[0013] Compared with the prior art, the technical scheme provided by the present application has the following beneficial effects: The present application realizes high-precision identification in a shielding scene by the cooperative work of the multi-modal biometric identification module and the environment perception module, and the identification of biometric information by the data processing module, automatically switches to the iris identification mode when the face is shielded, and effectively improves the identification pass rate by using a weighted fusion strategy. At the same time, the shielding object detection algorithm is used to automatically generate the details of the shielding part, realizing a double-identification guarantee mechanism, further improving the identification accuracy in a shielding scene, and ensuring the reliability of the access control system.

[0014] Additional aspects and advantages of the present application will be in part apparent and in part pointed out hereinafter. BRIEF DESCRIPTION OF DRAWINGS

[0015] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description, taken in conjunction with the accompanying drawings, in which: Figure 1 FIG. 1 is a structural schematic diagram of a smart property access control system according to the present application; Figure 2 FIG. 2 is a working flowchart of the smart property access control system according to the present application. DETAILED DESCRIPTION

[0016] Embodiments of the present invention are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the invention, and should not be construed as limiting the invention. Rather, embodiments of the invention include all variations, modifications, and equivalents falling within the spirit and scope of the appended claims.

[0017] The intelligent property access control system of the present invention will be described below with reference to the accompanying drawings.

[0018] like Figures 1-2 As shown in the figure, an intelligent property access control system according to an embodiment of the present invention includes a multimodal biometric module, an environmental sensing module, a data processing module, a data storage module, and a data management module. The multimodal biometric module integrates a camera module, a fingerprint module, and an iris module to collect personnel biometric features, including facial features, fingerprints, and irises. In obstructed scenarios, it automatically switches to iris recognition, and the recognition information is transmitted to the data processing module in real time.

[0019] It should be noted that the automatic switching mechanism of the multimodal biometric module relies on the real-time monitoring data of the environmental perception module. When the environmental perception module detects an obstruction covering the key area of ​​the face, it will immediately send a switching command to the multimodal biometric module to complete the seamless transition from face recognition to iris recognition. This dynamic switching mechanism not only improves the recognition accuracy, but also avoids the risk of system paralysis due to the failure of a single recognition mode.

[0020] The environmental perception module integrates an ambient light sensor and an occlusion detection camera to monitor the ambient light intensity and the type of occlusion in real time, and transmits the data to the data processing module in real time.

[0021] It should be noted that the ambient light sensor can automatically adjust the exposure parameters of the camera module and the occlusion detection camera according to the intensity of external light, ensuring that clear images of human biometric features can be captured under different lighting conditions.

[0022] The data processing module is used to process biometric data and occlusion data, and uses multimodal feature fusion algorithms and occlusion detection algorithms to analyze and compare the collected data to identify the person's identity.

[0023] The data processing module uses an occlusion detection algorithm to identify the type and location of occlusions, and generates a complete facial image based on the unoccluded areas. The occlusion detection algorithm specifically includes the following steps: The input data is preprocessed to reduce noise interference. Specifically, histogram equalization or adaptive contrast enhancement is used to improve facial details in low-light or backlit images. Bilateral filtering or non-local mean denoising is used to smooth noise while preserving edge information. The images are uniformly scaled to 256×256 pixels to adapt to the input requirements of subsequent models.

[0024] To identify the occlusion category, an occlusion detection model is first constructed. Then, a convolutional neural network is used to extract multi-scale features of the image, output the category label and location coordinates of the occlusion, merge overlapping bounding boxes, and retain the most likely target occlusion.

[0025] The occluded facial regions are marked to provide a mask for image inpainting. Specifically, the preprocessed image is input, and a pixel-level classification map is generated through an encoder-decoder structure. The probability of each pixel belonging to "occluder" or "face" is output to generate a binary mask.

[0026] Separate the unoccluded facial regions as a reference for generating the complete image. Specifically, invert the occlusion mask to obtain the unoccluded region mask (1=unoccluded, 0=occluded). Extract RGB image blocks of the unoccluded parts (such as the forehead and skin around the eyes) based on the mask.

[0027] A complete facial image is generated based on the unoccluded area, restoring the details of the occluded part, optimizing the generation result, and obtaining biological facial features. Specifically, the unoccluded area image and the mask are encoded into low-dimensional feature vectors, and the feature vectors are progressively upsampled to generate the repaired facial image. The authenticity of the generated image is judged, forcing the generator to output more natural facial details and improving the visual authenticity of the generated image.

[0028] It should be noted that the core purpose of this "generating a complete facial image" step is to infer and supplement the feature information of the occluded area based on the unoccluded part, so that a more accurate and robust overall facial feature vector can be extracted for identity matching, rather than necessarily generating a complete photo that can be viewed visually.

[0029] The data processing module uses a multimodal feature fusion algorithm to analyze and compare the collected biometric data to identify individuals. The multimodal fusion algorithm employs a weighted fusion strategy, dynamically adjusting weights based on the recognition confidence levels of face and iris scans. Specifically, it includes the following steps: Facial features and iris features are extracted separately. When a user's facial features are obscured by an occlusion, the facial image data is processed by an occlusion detection algorithm and automatically switched to iris recognition mode.

[0030] Based on the real-time recognition confidence of face and iris and occlusion interference factors, the weight ratio of the two modalities is dynamically adjusted, specifically: The face weight is set to 0.6 by default, and the iris weight is 0.4. When the confidence of either modality is lower than the threshold (e.g., face confidence < 0.7), dynamic weight adjustment is initiated.

[0031] Weighting adjustment rules: Confidence-driven: Weights are proportional to modal confidence.

[0032] For example: if the face confidence score is 0.8 and the iris confidence score is 0.9, then the adjusted weights are: Face weight = 0.8 / (0.8+0.9)≈0.47; Iris weight = 0.9 / (0.8+0.9)≈0.53; If the confidence level of a face drops to 0.4 due to mask obscuring it, while the confidence level of an iris is 0.95, then the weights are adjusted as follows: Face weight = 0.4 / (0.4+0.95)≈0.3; Iris weight = 0.95 / (0.4+0.95)≈0.7.

[0033] The adjusted modal weights are multiplied by their corresponding confidence levels and then summed to generate a comprehensive recognition score. This score is then compared with a preset threshold to determine the legitimacy of the user's identity. Specifically, if the comprehensive score is greater than or equal to the preset threshold (e.g., 0.8), the user is considered a legitimate user, and the unlocking command is triggered.

[0034] It should be noted that automatically switching to iris recognition mode essentially refers to an adjustment in the focus of the acquisition and recognition strategy. When there is no obstruction, the system simultaneously acquires face and iris information, but gives higher weight to facial features for fusion decision-making. When severe obstruction is detected, the system does not completely shut down face recognition, but attempts to repair features through algorithms, while actively shifting the focus of acquisition and recognition to the iris, and correspondingly significantly increasing the proportion of the iris in the fusion weight, thereby achieving a seamless high-precision recognition transition.

[0035] In one embodiment of the present invention, such as Figure 1 As shown, the data storage module is used to store personnel biometric data, environmental perception data, and identification record information, while also providing data backup and recovery functions.

[0036] The data storage module integrates a storage device and a database. The storage device is used to locally store system data, including personnel biometric data, environmental perception data, identification record information, and access control permission settings. The database has a data synchronization function, which performs regular synchronization with the cloud data center to achieve data backup and recovery.

[0037] It should be noted that the local storage is a hard drive, and the storage uses a redundant array design. When a single storage unit fails, the system can automatically switch to a backup unit to ensure the security of data storage.

[0038] When local storage fails or data is lost, the system can be quickly restored by backing up data in the cloud, ensuring the data security and business continuity of the access control system.

[0039] The data management module is used to manage system data, as well as to input and query personnel information. Specific functions include: when inputting personnel information, it supports batch import and single addition, and the input fields cover name, employee number, department, fingerprint template, iris feature code, face image, access control permission group, etc.; the query function supports multi-dimensional combination search by name, employee number, department, etc., and returns basic personnel information and the 10 most recent access control records.

[0040] In one embodiment of the present invention, such as Figure 1 As shown, it also includes an access control controller, which receives the identification results from the data processing module and controls the opening or closing of the electromagnetic lock according to preset rules. When the identification result is an authorized person, the access control controller sends an opening signal to the electromagnetic lock. After receiving the signal, the electromagnetic lock performs the opening action, allowing the person to pass. When the identification result is an unauthorized person or the identification fails, the access control controller does not send an opening signal, and the electromagnetic lock remains closed, preventing the person from passing.

[0041] In one embodiment of the present invention, such as Figure 1 As shown, the system also includes a card reader connected to the access control controller. The card reader reads the access control card information held by the user and transmits the read information to the access control controller. Upon receiving the access control card information, the access control controller compares it with the personnel information stored in the data management module. If the information matches and conforms to preset access control permission rules, the controller opens the electromagnetic lock. The system also includes an alarm connected to the access control controller. When the access control controller detects an abnormal situation, it triggers the alarm to emit an alarm signal.

[0042] In practical applications, when a user approaches the access control system, they can unlock via fingerprint or biometric recognition. When biometric unlocking is used, the multimodal biometric module is activated first, with the camera module and iris module working together to collect the person's biometric information. If the environmental perception module detects an obstruction on the face, such as a mask or hat, it will immediately send a switching command to the multimodal biometric module, causing it to automatically switch to iris recognition mode to ensure the accuracy of the recognition. The collected biometric data and obstruction data are transmitted to the data processing module in real time. The data processing module first uses the obstruction detection algorithm and the multimodal feature fusion algorithm to analyze and compare the data to identify the person's identity and determine the legitimacy of the person's identity.

[0043] If the person is identified as an authorized person, the data processing module sends the identification result to the access control controller. The access control controller then opens the electromagnetic lock according to preset rules, allowing the person to pass. If the person is identified as an unauthorized person or the identification fails, the access control controller does not send an opening signal, and the electromagnetic lock remains closed.

[0044] In addition, if a user has an access card, the card reader will read the access card information and transmit it to the access control controller. The access control controller will compare the information read with the personnel information stored in the data management module. If the information matches and conforms to the preset access control permission rules, it will also control the electromagnetic lock to open.

[0045] In summary, the smart property access control system of this invention achieves high-precision recognition in occluded scenarios through the collaborative work of a multimodal biometric module and an environmental perception module, and by recognizing biometric information through a data processing module. It automatically switches to iris recognition mode when the face is obscured and employs a weighted fusion strategy to effectively improve the recognition pass rate. Simultaneously, in conjunction with an obstruction detection algorithm, it automatically generates details of the obscured portion, realizing a dual recognition guarantee mechanism, further improving the recognition accuracy in occluded scenarios and ensuring the reliability of the access control system.

[0046] In the description of this specification, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0047] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0048] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A smart property access control system, characterized in that, The system comprises a multi-modal biometric recognition module, an environment perception module, a data processing module, a data storage module and a data management module, wherein, The multi-modal biometric recognition module integrates a camera module, a fingerprint module and an iris module, and is used for collecting personnel biological characteristics, including facial features, fingerprints and irises, and automatically switching to iris recognition in a shielding scenario, and transmitting the recognition information to the data processing module in real time; The environment perception module integrates an ambient light sensor and a shielding detection camera, and is used for monitoring the ambient light intensity and the type of shielding object in real time, and transmitting the data to the data processing module in real time; The data processing module is used for processing biological characteristic data and shielding object data, and analyzing and comparing the collected data by using a multi-modal feature fusion algorithm and a shielding object detection algorithm to identify the personnel identity; The data storage module is used for storing personnel biological characteristic data, environment perception data and recognition record information, and simultaneously providing data backup and recovery functions; The data management module is used for managing system data, and inputting and querying personnel information.

2. The smart property access control system of claim 1, wherein, The data processing module adopts a shielding object detection algorithm to identify the type and position of the shielding object, and generates a complete facial image according to the unshielded area, and the shielding object detection algorithm specifically comprises the following steps: Pretreat the input data to reduce noise interference; Identify the shielding object category, first construct a shielding object detection model, then extract image multi-scale features, output the category label and position coordinates of the shielding object, merge overlapping bounding boxes, and retain the most possible target shielding object; Mark the shielded facial area to provide a mask for image repair; Separate the unshielded facial area as a reference for generating a complete image; Generate a complete facial image based on the unshielded area, restore the details of the shielded part, optimize the generation result, and obtain the biological facial features.

3. The smart property access control system of claim 1, wherein, The data processing module adopts a multi-modal feature fusion algorithm to analyze and compare the collected biological data to identify the personnel identity, wherein, The multi-modal fusion algorithm adopts a weighted fusion strategy, combines the recognition confidence of the face and the iris, and dynamically adjusts the weight, specifically comprising the following steps: Respectively extract the face features and the iris features, when the user face features are shielded by the shielding object, the face image data is processed by the shielding object detection algorithm, and the iris recognition mode is automatically switched; Dynamically adjust the weight proportion of the two modalities according to the real-time recognition confidence of the face and the iris and the shielding interference factors; Multiply the adjusted modal weight by the corresponding confidence, sum up, generate a comprehensive recognition score, and finally determine the legality of the personnel identity by comparing with a preset threshold.

4. The smart property access control system of claim 1, wherein, The data storage module integrates a memory and a database, wherein, The memory is used for locally saving system data, including personnel biological characteristic data, environment perception data, recognition record information and access control permission setting information; The database has a data synchronization function, and periodically synchronizes with a cloud data center to realize data backup and recovery.

5. The smart property access control system of claim 1, wherein, It also comprises an access controller, which is used for receiving the recognition result of the data processing module, and controlling the opening or closing of the electromagnetic lock according to a preset rule.

6. The smart property access control system of claim 5, wherein, It also comprises a card reader, wherein, The card reader is connected with the access control device, is used for reading the access card information held by the user, and transmits the read information to the access control device, and the access control device compares the access card information with the personnel information stored in the data management module after receiving the access card information, and if the information matches and meets the preset access permission rule, the electromagnetic lock is controlled to be opened.

7. The smart property access control system of claim 6, wherein, Further comprising an alarm connected with the access control device, when the access control device detects an abnormal situation, the alarm is triggered to send an alarm signal.

Images