Iris recognition control opening and closing method based on smart lock
By using a high-resolution camera and infrared light source on the smart lock for multimodal fusion processing, and using 3D-LBP algorithm for iris feature extraction, combining dynamic iris positioning and encrypted transmission technology, the iris is solved by intelligent locking under different lighting conditions. The stability issues of image acquisition and feature matching improve the accuracy and security of recognition.
Patent Information
- Application Number
- CN202411118849.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-08-15
AI Technical Summary
It is difficult for existing smart locks to stably acquire clear iris images under different lighting conditions. The accuracy of iris image feature extraction and matching algorithms is insufficient, and it is difficult to update the iris feature library in real time to adapt to user iris changes, affecting the reliability and security of identification.
The combination of high-resolution camera and infrared light source is used to extract iris image features through multimodal fusion processing and 3D-LBP algorithm, and combine dynamic iris positioning and encryption transmission technology to realize dynamic update of the iris feature library and user identity verification.
Ensure the stability and quality of iris image acquisition under different lighting conditions, improve the accuracy and robustness of iris feature extraction, ensure the security and accuracy of identification, and realize the long-term stability and user experience of smart locks.
Smart Images

Figure CN118644920B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of iris recognition, and in particular to an iris recognition-based opening and closing control method for a smart lock. Background Art
[0002] With the rapid development of smart home technology, smart locks, as an important part of home security, have been widely used. Traditional smart locks mainly rely on technologies such as fingerprint recognition, password input and wireless radio frequency identification (RFID) to realize the opening and closing control of door locks. However, these methods have some significant shortcomings in practical applications. Fingerprint recognition is easily affected by fingerprint wear or surface stains, resulting in recognition failure; password input is easy to be peeped or forgotten, posing a security risk; RFID cards are easy to be lost or copied, and the security is not high. In order to improve the security and convenience of smart locks, iris recognition technology has gradually become a hot spot for research and application due to its uniqueness, stability and difficulty in forging.
[0003] At present, smart locks based on iris recognition still face some technical difficulties in practical applications, which need to be solved urgently. First, how to stably and clearly collect the user's iris image under different lighting conditions is an important challenge. Secondly, the feature extraction and matching algorithms of iris images need to have high accuracy and high robustness to ensure the reliability of recognition. In addition, how to update the iris feature library in real time to adapt to the changes in the user's iris features, as well as to ensure the security of the unlocking process and the timeliness of data transmission are all problems that are difficult to effectively solve with existing technologies.
[0004] Therefore, it is of great practical significance to develop an iris recognition control opening and closing method based on smart locks to solve the above technical problems and improve the security, accuracy and user experience of smart locks. Summary of the invention
[0005] Based on the above objectives, the present invention provides an iris recognition controlled opening and closing method based on a smart lock.
[0006] The iris recognition control opening and closing method based on the smart lock includes the following steps:
[0007] S1: Collecting the user's iris image through the iris acquisition module on the smart lock, where the iris acquisition module includes a high-resolution camera and an infrared light source to ensure that clear images are collected under various lighting conditions;
[0008] S2: Perform multimodal fusion processing on the iris image collected by S1, and fuse the infrared spectrum image and the visible spectrum image to enhance the clarity of the iris features;
[0009] S3: Using a preset positioning algorithm to locate the iris image, extracting the iris area, and performing real-time normalization processing;
[0010] S4: Use 3D-LBP algorithm to extract features from the normalized iris image and generate a high-dimensional iris feature vector;
[0011] S5: dynamically matching the extracted iris feature vector with the pre-registered iris feature library stored in the smart lock, using a preset encryption transmission and processing technology, and regularly updating the iris feature library through a dynamic update mechanism;
[0012] S6: Perform user identity authentication based on the matching result of S5, and determine whether the user has the unlocking authority through the control module of the smart lock. If the verification is successful, the control module sends an unlocking command to the actuator to drive the motor to perform the unlocking operation. After the unlocking is successful, the unlocking record and user information are uploaded to the cloud server through the wireless communication module.
[0013] Furthermore, the S1 specifically includes:
[0014] S11: The iris acquisition module on the smart lock includes an ambient light sensor, which is used to detect the light intensity of the surrounding environment in real time and determine whether the current lighting conditions are suitable for iris acquisition;
[0015] S12: According to the ambient light intensity detected by S11, the control unit of the iris acquisition module automatically adjusts the brightness and angle of the infrared light source to provide optimal lighting conditions, ensuring that a clear iris image can be obtained even in low-light or backlit environments;
[0016] S13: a high-resolution camera in the iris acquisition module, used to capture an iris image of the user, the high-resolution camera having a resolution of at least 1080P;
[0017] S14: During the acquisition process, a high-resolution camera is used to capture infrared spectrum images and visible spectrum images at the same time. By synchronously acquiring images of the two spectrums, complete iris image information is ensured under different lighting conditions;
[0018] S15: The iris image obtained in S14 is detected by an image quality detection algorithm to ensure the clarity, contrast and noise-freeness of the image. When the image quality does not meet the preset standard, return to S12 for re-collection until an iris image that meets the quality requirements is obtained.
[0019] Furthermore, the S2 specifically includes:
[0020] S21: preprocessing the infrared spectrum image and the visible spectrum image obtained from S1, specifically including grayscale transformation, noise filtering and image enhancement to improve the quality of the image;
[0021] S22: extracting features from the preprocessed infrared spectrum image and visible spectrum image respectively, and using a preset image feature extraction algorithm to obtain feature information of each image;
[0022] S23: using a preset feature matching algorithm, matching the features of the infrared spectrum image and the visible spectrum image, aligning corresponding points in the two images, and ensuring that the two images are consistent in spatial position;
[0023] S24: Determine a fusion strategy based on the feature matching result, specifically use a preset weighted average method to assign weights to the infrared spectrum image and the visible spectrum image, and the weight assignment is based on parameters including image contrast, brightness, and feature richness; then fuse the infrared spectrum image and the visible spectrum image, and synthesize a fused image through the weighted average method to ensure that the fused image retains the features of the two spectrum images;
[0024] S25: Optimize the fused image, including contrast enhancement and detail enhancement, to provide high-quality image data for subsequent iris positioning and feature extraction.
[0025] Furthermore, the S24 specifically includes:
[0026] S241: Calculate contrast, brightness, and feature richness of the preprocessed infrared spectrum image and visible spectrum image;
[0027] The contrast ratio is calculated by calculating the standard deviation of the image gray value, and the calculation formula is:
[0028] ,
[0029] ;
[0030] in, is the contrast of infrared spectrum image; Represents the infrared spectrum image The gray value of pixels; is the average gray value of the infrared spectrum image; is the total number of pixels of the infrared spectrum image; is the contrast of the visible spectrum image; Represents the visible spectrum image The gray value of pixels; is the average gray value of the visible spectrum image; is the total number of pixels of the visible spectrum image;
[0031] The brightness is calculated by the method of calculating the average gray value of the image, and the formula is:
[0032] ,
[0033] ;
[0034] in, is the brightness of the infrared spectrum image; The infrared spectrum image The gray value of pixels; is the brightness of the visible spectrum image; The visible spectrum image The gray value of pixels;
[0035] The feature richness uses an edge detection algorithm to calculate the number of edges in the image, and the formula is:
[0036] ,
[0037] ,
[0038] in, The richness of the external spectral image; Represents the infrared spectrum image A binary flag indicating whether a pixel is an edge pixel; is the richness of the visible spectrum image; Represents the visible spectrum image A binary flag indicating whether a pixel is an edge pixel;
[0039] S242: According to the parameters calculated in S241, weights are assigned to the infrared spectrum image and the visible spectrum image. The weight of the infrared spectrum image is , the weight of the visible spectrum image is , then the weight calculation formulas are:
[0040] Contrast Weight: , ;
[0041] Luminance Weight: , ;
[0042] Feature richness weight: , ;
[0043] S243: Comprehensively combine various weight parameters to obtain the final fusion weight, and set the comprehensive weight to and , then the calculation formula of the comprehensive weight is:
[0044] ,
[0045] ;
[0046] in, , , are the weight coefficients of contrast, brightness and feature richness, which need to be satisfied ;
[0047] S244: According to the comprehensive weight calculated in S243, the infrared spectrum image and the visible spectrum image are weighted average fused, and the fused image The calculation formula is:
[0048] ,
[0049] in, is the gray value matrix of the infrared spectrum image; is the gray value matrix of the visible spectrum image.
[0050] Furthermore, the S3 specifically includes:
[0051] S31: Use the improved Canny edge detection algorithm to perform edge detection on the iris image and extract the edges of the iris and pupil. The specific formula is:
[0052] ,
[0053] in, is the gradient amplitude; For images in Directional gradient; For images in Directional gradient;
[0054] S32: Perform circular Hough transform on the edge image extracted in S31 to identify the inner and outer boundaries of the iris. The formula of circular Hough transform is: ,in, is the coordinate of the center of the circle; is the radius; is the coordinate of the edge point;
[0055] S33: extracting the iris region according to the inner and outer boundaries of the iris identified in S32, generating an image containing only the iris portion, and determining the start and end coordinates of the iris region by the inner and outer boundaries for cutting;
[0056] S34: performing normalization processing on the extracted iris region, converting iris images of different sizes and shapes into uniform sizes and shapes, specifically using a polar coordinate conversion method, and the normalization formula is:
[0057] ,
[0058] in, is the normalized iris image; is the original iris image; The conversion formula from polar coordinates to rectangular coordinates is ; The conversion formula from polar coordinates to rectangular coordinates is ;
[0059] S35: Correcting the normalized iris image to eliminate distortion caused by shooting angle or user posture, and ensuring that all iris images have the same alignment and ratio.
[0060] Furthermore, the S4 specifically includes:
[0061] S41: dividing the normalized iris image into a plurality of small blocks, each small block containing a predetermined number of pixels for extracting local features;
[0062] S42: performing 3D local binary pattern coding processing on each block of the iris image, wherein the 3D-LBP generates a binary code for each pixel by analyzing the grayscale changes of the image block at different scales and directions;
[0063] S43: combining the 3D-LBP coding values of all pixels in each block into a coding map, where the coding map is used to represent local features of the iris image at different scales and directions;
[0064] S44: performing statistical analysis on the coding map, calculating the occurrence frequency of each coding pattern in the entire iris image, generating a feature histogram, and combining the data in the feature histogram into a high-dimensional feature vector, which is used to represent the local texture features and global feature distribution of the iris image;
[0065] S45: Optimizing the generated high-dimensional feature vector, reducing the dimension of the feature vector by using dimensionality reduction technology, retaining predetermined feature information, and removing redundant and noise data.
[0066] Furthermore, the step of 3D local binary pattern encoding processing in S42 includes:
[0067] S421: Select a pixel in each block as a central pixel;
[0068] S422: Collecting neighborhood pixels around the central pixel according to a preset radius and direction;
[0069] S423: Compare the grayscale value of the neighboring pixel with the grayscale value of the central pixel. If the grayscale value of the neighboring pixel is greater than or equal to the grayscale value of the central pixel, the position of the neighboring pixel is recorded as 1, otherwise it is recorded as 0;
[0070] S424: Combining the binary results of all neighboring pixels into a binary number as the 3D-LBP encoding value of the central pixel.
[0071] Furthermore, the S44 specifically includes:
[0072] S441: Perform statistical analysis on the coding map generated for each block, and calculate the frequency of occurrence of each 3D-LBP coding mode in the block. The specific method is to traverse each pixel in the coding map and record the number of times each coding mode appears. First, preset the first pixel in the coding map. The number of times the encoding mode appears in the block is ;
[0073] S442: The number of occurrences of each coding mode counted in S441 Divide by the total number of pixels in the tile , calculate the frequency of each coding mode , the calculation formula is as follows: ;
[0074] S443: The frequencies of all coding modes in each block are Combined into a feature histogram, the histogram represents the feature distribution of the iris image in different local areas. Suppose there are encoding mode, the feature histogram is a A vector of entries , the formula is: ;
[0075] S444: Merge the feature histograms of all the blocks to generate a global feature histogram of the entire iris image. Specifically, accumulate the frequencies of the same coding mode in different blocks to obtain the coding mode frequency distribution in the global range. The frequency of this coding pattern in the entire iris image is , the calculation formula is:
[0076] ,in, is the total number of blocks; For the The encoding mode is The frequency in each block;
[0077] S445: Arrange the data in the global feature histogram in order to form a high-dimensional feature vector. The frequency of each encoding mode As a dimension in the feature vector, the feature vector is used to represent the overall texture feature distribution of the iris image. It is expressed as: ,in, is a high-dimensional feature vector.
[0078] Furthermore, the S5 specifically includes:
[0079] S51: Extract the iris feature vector The feature vector stored in the pre-registered iris feature library in the smart lock Compare them one by one and calculate the similarity between feature vectors. Specifically, cosine similarity is used to calculate the similarity value. , the calculation formula is: ,in, is the newly extracted iris feature vector; The first eigenvectors; · is the vector dot product; is the magnitude of the vector;
[0080] S52: Calculate the similarity value Similarity threshold with the preset For comparison, if Greater than or equal to threshold , then the match is considered successful and the user's identity is identified; otherwise, the match fails;
[0081] S53: Regularly update the pre-registered iris feature database dynamically to maintain the accuracy and latest status of the feature database. The dynamic update mechanism includes the following steps:
[0082] S531: The iris feature vectors successfully matched within a predetermined time are Store in temporary storage area;
[0083] S532: Perform statistical analysis on the feature vectors in the temporary storage area, calculate the frequency and stability of the feature vectors of each user, and select feature vectors with high frequency and stability as update candidates;
[0084] S533: Compare the updated candidate feature vector with the corresponding feature vector in the existing feature library. If the new feature vector has higher stability and matching accuracy, replace the old feature vector with the new feature vector to update the feature library.
[0085] S54: According to actual application requirements, set periodic parameters for updating the feature library to ensure that the feature library can reflect changes in the user's iris features in a timely manner.
[0086] Furthermore, the S6 specifically includes:
[0087] S61: receiving the matching result of the similarity judgment in S5, if the matching is successful, proceeding to the next step of user identity verification; if the matching fails, refusing to unlock and issuing a warning signal;
[0088] S62: The control module searches for corresponding user authority information from a pre-stored user authority database based on the successfully matched user feature vector, and verifies whether the user has unlocking authority, including user identity verification and authority level verification. Only users with unlocking authority can proceed to the next step;
[0089] S63: If the user passes the authority verification, the control module will generate an unlocking instruction and send the instruction to the driving circuit of the actuator. The instruction includes specific operation steps of starting the motor and opening the lock cylinder;
[0090] S64: After receiving the unlocking command sent by the control module, the actuator starts the motor to drive the lock cylinder to rotate, completing the unlocking operation;
[0091] S65: After the motor completes the unlocking operation, the sensor in the actuator detects the state of the lock cylinder to confirm whether the unlocking is successful. If the unlocking is successful, the sensor feeds back a success signal to the control module;
[0092] S66: After receiving the unlocking success signal, the control module generates an unlocking record, including information on user identity, unlocking time, and unlocking location;
[0093] S67: Finally, the generated unlocking record and user information are uploaded to the cloud server through wireless communication technology.
[0094] The beneficial effects of the present invention are:
[0095] 1) The present invention solves the stability problem of iris image acquisition under different lighting conditions by introducing multimodal fusion processing and dynamic iris positioning algorithm. Through the coordinated work of high-resolution camera and infrared light source, combined with multi-frame image synthesis and image quality detection technology, it ensures that clear iris images can be collected in complex lighting environments such as low light and backlight, thereby improving image quality and providing a reliable foundation for the subsequent recognition process;
[0096] 2) The present invention uses the 3D-LBP algorithm to extract features from iris images and generates high-dimensional iris feature vectors. By performing detailed statistical analysis on the coding map and optimizing the feature vectors, the accuracy and robustness of iris feature extraction are improved. At the same time, the application of dynamic matching and encrypted transmission technology ensures the security and accuracy of the iris feature vectors during transmission and matching, effectively preventing data leakage and matching errors.
[0097] 3) The present invention regularly updates the iris feature library through a dynamic update mechanism to adapt to changes in user iris features and ensure the long-term stability and accuracy of the recognition system. The control module performs user identity authentication based on the matching results, and uploads the unlocking records and user information to the cloud server through the wireless communication module, realizing remote data management and auditing, significantly improving the security, accuracy and user experience of smart locks, and having high practical value and promotion prospects. BRIEF DESCRIPTION OF THE DRAWINGS
[0098] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only for the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0099] Figure 1 This is a flow chart of the iris recognition control on / off method according to an embodiment of the present invention;
[0100] Figure 2 3D-LBP encoding process flow diagram of an embodiment of the present invention;
[0101] Figure 3 This is a comparison chart of the performance indicators of different feature extraction algorithms under the conditions of 100 users and 3000 samples;
[0102] Figure 4 This is a comparison chart of the improvement in accurate matching rate before and after the update, with 20 users and 6000 samples.
[0103] Figure 5 This is a comparison chart of the overall performance of the method of the present invention and the control group method under the conditions of 50 users and 1500 operations. DETAILED DESCRIPTION
[0104] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.
[0105] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. The words "include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects.
[0106] like Figure 1-Figure 2 As shown, the iris recognition control opening and closing method based on the smart lock includes the following steps:
[0107] S1: The iris acquisition module on the smart lock collects the user's iris image. The iris acquisition module includes a high-resolution camera and an infrared light source to ensure clear images under various lighting conditions.
[0108] S2: Perform multimodal fusion processing on the iris image collected by S1, and fuse the infrared spectrum image and the visible spectrum image to enhance the clarity of the iris features;
[0109] S3: Use a preset positioning algorithm to locate the iris image, extract the iris area, and perform real-time normalization processing to adapt to the changes in iris size and shape of different users and ensure the consistency of feature extraction;
[0110] S4: 3D-LBP (three-dimensional local binary pattern) is used to extract features from the normalized iris image to generate a high-dimensional iris feature vector. 3D-LBP captures the texture features of the iris image at different scales to enhance the recognition accuracy and anti-interference ability.
[0111] S5: Dynamically match the extracted iris feature vector with the pre-registered iris feature library stored in the smart lock. The preset encryption transmission and processing technology is used in the process to ensure the security of data during transmission and matching, and the iris feature library is regularly updated through a dynamic update mechanism;
[0112] S6: Perform user identity authentication based on the matching result of S5, and determine whether the user has the unlocking authority through the control module of the smart lock. If the verification is successful, the control module sends an unlocking command to the actuator to drive the motor to perform the unlocking operation. After the unlocking is successful, the unlocking record and user information are uploaded to the cloud server through the wireless communication module.
[0113] S1 specifically includes:
[0114] S11: The iris acquisition module on the smart lock includes an ambient light sensor, which is used to detect the light intensity of the surrounding environment in real time and determine whether the current lighting conditions are suitable for iris acquisition;
[0115] S12: According to the ambient light intensity detected by S11, the control unit of the iris acquisition module automatically adjusts the brightness and angle of the infrared light source to provide optimal lighting conditions, ensuring that a clear iris image can be obtained even in low-light or backlit environments;
[0116] S13: a high-resolution camera in the iris acquisition module, used to capture an iris image of the user, where the high-resolution camera has a resolution of at least 1080P to ensure high detail and clarity of the image;
[0117] S14: During the acquisition process, a high-resolution camera is used to capture infrared spectrum images and visible spectrum images at the same time. By synchronously acquiring images of the two spectrums, complete iris image information is ensured under different lighting conditions;
[0118] S15: The iris image obtained in S14 is detected by an image quality detection algorithm to ensure the clarity, contrast and noise-freeness of the image. When the image quality does not meet the preset standards, return to S12 for re-capture until an iris image that meets the quality requirements is obtained. Through the above steps, the iris acquisition module can stably and clearly capture the user's iris image under various lighting conditions, thereby improving the reliability and adaptability of the iris recognition system and ensuring the accuracy and security of subsequent processing and recognition.
[0119] S2 specifically includes:
[0120] S21: preprocessing the infrared spectrum image and visible spectrum image obtained from S1, including grayscale transformation, noise filtering and image enhancement, so as to improve the quality of the image and ensure the accuracy in the fusion process;
[0121] S22: extracting features from the preprocessed infrared spectrum image and visible spectrum image respectively, using a preset image feature extraction algorithm to obtain feature information of each image, ensuring that effective information of each spectrum image can be retained during fusion;
[0122] S23: using a preset feature matching algorithm, matching the features of the infrared spectrum image and the visible spectrum image, aligning corresponding points in the two images, and ensuring that the two images are consistent in spatial position;
[0123] S24: Determine a fusion strategy based on the feature matching result, specifically use a preset weighted average method to assign weights to the infrared spectrum image and the visible spectrum image, and the weight assignment is based on parameters including image contrast, brightness, and feature richness; then fuse the infrared spectrum image and the visible spectrum image, and synthesize a fused image through the weighted average method to ensure that the fused image retains the features of the two spectrum images and enhances the clarity and stability of the iris features;
[0124] S25: Optimizing the fused image, including contrast enhancement and detail enhancement, to ensure that the final fused image has high-quality visual effects and provides high-quality image data for subsequent iris positioning and feature extraction; through the above steps, the infrared spectrum image and the visible spectrum image are effectively fused, the feature clarity and stability of the iris image are improved, the accuracy and robustness of the iris recognition system are enhanced, and high-quality iris images can be obtained under various lighting conditions.
[0125] S24 specifically includes:
[0126] S241: Calculate contrast, brightness, and feature richness of the preprocessed infrared spectrum image and visible spectrum image;
[0127] The contrast ratio is calculated by calculating the standard deviation of the image gray value. The calculation formula is:
[0128] ,
[0129] ;
[0130] in, is the contrast of infrared spectrum image; Represents the infrared spectrum image The gray value of pixels; is the average gray value of the infrared spectrum image; is the total number of pixels of the infrared spectrum image; is the contrast of the visible spectrum image; Represents the visible spectrum image The gray value of pixels; is the average gray value of the visible spectrum image; is the total number of pixels of the visible spectrum image;
[0131] The brightness is calculated by the method of calculating the average gray value of the image. The formula is:
[0132] ,
[0133] ;
[0134] in, is the brightness of the infrared spectrum image; The infrared spectrum image The gray value of pixels; is the brightness of the visible spectrum image; The visible spectrum image The gray value of pixels;
[0135] Feature richness uses edge detection algorithms (such as Canny edge detection) to calculate the number of edges in the image. The formula is:
[0136] ,
[0137] ,
[0138] in, The richness of the external spectral image; Represents the infrared spectrum image A binary flag indicating whether a pixel is an edge pixel; is the richness of the visible spectrum image; Represents the visible spectrum image A binary flag indicating whether a pixel is an edge pixel;
[0139] S242: According to the parameters calculated in S241, weights are assigned to the infrared spectrum image and the visible spectrum image. The weight of the infrared spectrum image is , the weight of the visible spectrum image is , then the weight calculation formulas are:
[0140] Contrast Weight: , ;
[0141] Luminance Weight: , ;
[0142] Feature richness weight: , ;
[0143] S243: Comprehensively combine various weight parameters to obtain the final fusion weight, and set the comprehensive weight to and , then the calculation formula of the comprehensive weight is:
[0144] ,
[0145] ;
[0146] in, , , are the weight coefficients of contrast, brightness and feature richness, which need to be satisfied ;
[0147] S244: According to the comprehensive weight calculated in S243, the infrared spectrum image and the visible spectrum image are weighted average fused, and the fused image The calculation formula is:
[0148] ,
[0149] in, is the gray value matrix of the infrared spectrum image; is the gray value matrix of the visible spectrum image.
[0150] Through the above steps, the infrared spectrum image and the visible spectrum image are effectively weighted and fused under different lighting conditions, which improves the feature clarity and stability of the fused image, thereby enhancing the accuracy and reliability of the iris recognition system.
[0151] S3 specifically includes:
[0152] S31: Use the improved Canny edge detection algorithm to perform edge detection on the iris image and extract the edges of the iris and pupil. The specific formula is:
[0153] ,
[0154] in, is the gradient amplitude; For images in Directional gradient; For images in Directional gradient;
[0155] S32: Perform circular Hough transform on the edge image extracted in S31 to identify the inner and outer boundaries of the iris. The formula of circular Hough transform is: ,in, is the coordinate of the center of the circle; is the radius; is the coordinate of the edge point;
[0156] S33: extracting the iris region according to the inner and outer boundaries of the iris identified in S32, generating an image containing only the iris portion, and determining the start and end coordinates of the iris region by the inner and outer boundaries for cutting;
[0157] S34: performing normalization processing on the extracted iris region, converting iris images of different sizes and shapes into uniform sizes and shapes, specifically using a polar coordinate conversion method, and the normalization formula is:
[0158] ,
[0159] in, is the normalized iris image; is the original iris image; The conversion formula from polar coordinates to rectangular coordinates is ; The conversion formula from polar coordinates to rectangular coordinates is ;
[0160] S35: Correcting the normalized iris image to eliminate distortion caused by shooting angle or user posture, and ensuring that all iris images have the same alignment and proportion.
[0161] Through the above steps, using the preset dynamic iris positioning algorithm and real-time normalization processing method, the iris area can be accurately located and extracted, and the iris images of different users can be standardized, providing consistent and high-quality data input for subsequent feature extraction and recognition, thereby enhancing the accuracy and robustness of the iris recognition system.
[0162] S4 specifically includes:
[0163] S41: dividing the normalized iris image into a plurality of small blocks, each of which contains a predetermined number of pixels, for extracting local features, and through the block processing, more accurately capturing the local texture features in the iris image;
[0164] S42: performing 3D local binary pattern (3D-LBP) encoding processing on each block of the iris image, where 3D-LBP generates a binary code for each pixel by analyzing grayscale changes of the image block at different scales and directions;
[0165] S43: combining the 3D-LBP coding values of all pixels in each block into a coding map, where the coding map is used to represent local features of the iris image at different scales and directions;
[0166] S44: performing statistical analysis on the coding map, calculating the occurrence frequency of each coding pattern in the entire iris image, generating a feature histogram, and combining the data in the feature histogram into a high-dimensional feature vector, which is used to represent the local texture features and global feature distribution of the iris image;
[0167] S45: Optimizing the generated high-dimensional feature vector, using dimensionality reduction technology to reduce the dimension of the feature vector, retaining predetermined feature information, removing redundant and noise data, so as to enhance the discriminative power and stability of the feature vector; through the above steps, using 3D-LBP to extract features from the normalized iris image, it is possible to effectively capture the local texture features and global feature distribution of the iris image, generate a high-dimensional iris feature vector, provide high-precision and high-stability feature data for subsequent dynamic feature matching and user identity authentication, and improve the accuracy and reliability of the iris recognition system.
[0168] The steps of 3D local binary pattern encoding processing in S42 include:
[0169] S421: Select a pixel in each block as a central pixel;
[0170] S422: Collecting neighborhood pixels around the central pixel according to a preset radius and direction;
[0171] S423: Compare the grayscale value of the neighboring pixel with the grayscale value of the central pixel. If the grayscale value of the neighboring pixel is greater than or equal to the grayscale value of the central pixel, the position of the neighboring pixel is recorded as 1, otherwise it is recorded as 0;
[0172] S424: Combining the binary results of all neighboring pixels into a binary number as the 3D-LBP encoding value of the central pixel.
[0173] S44 specifically includes:
[0174] S441: Perform statistical analysis on the coding map generated for each block, and calculate the frequency of occurrence of each 3D-LBP coding mode in the block. The specific method is to traverse each pixel in the coding map and record the number of times each coding mode appears. First, preset the first pixel in the coding map. The number of times the encoding mode appears in the block is ;
[0175] S442: The number of occurrences of each coding mode counted in S441 Divide by the total number of pixels in the tile , calculate the frequency of each coding mode , the calculation formula is as follows: ;
[0176] S443: The frequencies of all coding modes in each block are Combined into a feature histogram, the histogram represents the feature distribution of the iris image in different local areas. Suppose there are encoding mode, the feature histogram is a A vector of entries , the formula is: ;
[0177] S444: Merge the feature histograms of all the blocks to generate a global feature histogram of the entire iris image. Specifically, accumulate the frequencies of the same coding mode in different blocks to obtain the coding mode frequency distribution in the global range. The frequency of this coding pattern in the entire iris image is , the calculation formula is:
[0178] ,in, is the total number of blocks; For the The encoding mode is The frequency in each block;
[0179] S445: Arrange the data in the global feature histogram in order to form a high-dimensional feature vector. The frequency of each encoding mode As a dimension in the feature vector, the feature vector is used to represent the overall texture feature distribution of the iris image. It is expressed as: ,in, It is a high-dimensional feature vector. Through the above steps, the coding map is statistically analyzed and a feature histogram is generated, which can accurately calculate the frequency of occurrence of each coding pattern in the iris image, and combine these frequency data into a high-dimensional feature vector, providing a comprehensive and detailed description of the iris image features, providing high-precision and high-reliability feature data for subsequent dynamic feature matching and user identity authentication, and improving the accuracy and robustness of the iris recognition system.
[0180] S5 specifically includes:
[0181] S51: Extract the iris feature vector The feature vector stored in the pre-registered iris feature library in the smart lock Compare them one by one and calculate the similarity between feature vectors. Specifically, cosine similarity is used to calculate the similarity value. , the calculation formula is: ,in, is the newly extracted iris feature vector; The first eigenvectors; · is the vector dot product; is the magnitude of the vector;
[0182] S52: Calculate the similarity value Similarity threshold with the preset For comparison, if Greater than or equal to threshold , then the match is considered successful and the user's identity is identified; otherwise, the match fails;
[0183] S53: Regularly update the pre-registered iris feature database dynamically to maintain the accuracy and latest status of the feature database. The dynamic update mechanism includes the following steps:
[0184] S531: The iris feature vectors successfully matched within a predetermined time are Store in temporary storage area;
[0185] S532: Perform statistical analysis on the feature vectors in the temporary storage area, calculate the frequency and stability of the feature vectors of each user, and select feature vectors with high frequency and stability as update candidates;
[0186] S533: Compare the updated candidate feature vector with the corresponding feature vector in the existing feature library. If the new feature vector has higher stability and matching accuracy, replace the old feature vector with the new feature vector to update the feature library.
[0187] S54: According to actual application requirements, set the periodic parameters of the feature library update to ensure that the feature library can timely reflect the changes in the user's iris features, while avoiding the waste of system resources caused by frequent updates. The update cycle can be dynamically adjusted according to the user's usage frequency and matching accuracy. Through the above steps, the extracted iris feature vector is dynamically matched with the pre-registered iris feature library stored in the smart lock, which can improve the accuracy and security of recognition. Through the dynamic update mechanism, the iris feature library is regularly updated to ensure the accuracy and currency of the feature library, effectively respond to changes in user iris features, and improve the long-term stability and reliability of the iris recognition system.
[0188] S6 specifically includes:
[0189] S61: receiving the matching result of the similarity judgment in S5, if the matching is successful, proceeding to the next step of user identity verification; if the matching fails, refusing to unlock and issuing a warning signal;
[0190] S62: The control module searches for corresponding user authority information from a pre-stored user authority database based on the successfully matched user feature vector, and verifies whether the user has unlocking authority, including user identity verification and authority level verification. Only users with unlocking authority can proceed to the next step;
[0191] S63: If the user passes the authority verification, the control module will generate an unlocking instruction and send the instruction to the driving circuit of the actuator. The instruction includes specific operation steps of starting the motor and opening the lock cylinder;
[0192] S64: After receiving the unlocking command sent by the control module, the actuator starts the motor to drive the lock core to rotate, completing the unlocking operation. The motor drive circuit in the actuator controls the rotation direction and rotation angle of the motor to ensure that the lock core can be accurately opened;
[0193] S65: After the motor completes the unlocking operation, the sensor in the actuator detects the state of the lock cylinder to confirm whether the unlocking is successful. If the unlocking is successful, the sensor feeds back a success signal to the control module;
[0194] S66: After receiving the unlocking success signal, the control module generates an unlocking record, including information on user identity, unlocking time, and unlocking location. These records are used for subsequent auditing and management;
[0195] S67: Finally, the generated unlocking record and user information are uploaded to the cloud server through wireless communication technology. Wireless communication can adopt Wi-Fi, Bluetooth or cellular network communication methods to ensure timely and secure data transmission; through the above steps, user identity authentication is performed according to the matching result of S5 to ensure that only authorized users can unlock the lock. The unlocking command is sent through the control module and the motor is driven to perform the unlocking operation, which improves the safety and reliability of the smart lock. The unlocking record and user information are uploaded to the cloud server through the wireless communication module, which realizes remote data management and auditing, and improves the management efficiency and security of the smart lock system.
[0196] In order to verify the effectiveness of the iris recognition control opening and closing method based on the smart lock of the present invention, a series of experimental tests were designed to evaluate its safety, accuracy and practicality. The experiments included iris image acquisition stability test, feature extraction and matching accuracy test, feature library dynamic update mechanism effect test and system overall performance test.
[0197] Experiment 1: Iris image acquisition stability test
[0198] Purpose: To verify whether the iris acquisition module can stably and clearly acquire the user's iris image under different lighting conditions.
[0199] The specific method is as follows:
[0200] 1. Set different lighting conditions in the laboratory (low light: 10 lux, normal light: 500 lux, high light: 2000 lux, backlight: 500 lux backlight);
[0201] 2. Select 100 volunteers and use the smart lock of the present invention to collect iris images of each volunteer under the above lighting conditions;
[0202] 3. Analyze the collected iris images using image quality assessment indicators, including clarity, contrast, and noise-free rate.
[0203] Table 1 Results of Experiment 1
[0204]
[0205] It can be seen from Table 1 that the iris acquisition module of the present invention can stably and clearly acquire iris images under different lighting conditions, and the image quality meets the expected standard, and is suitable for a variety of lighting environments in practical applications.
[0206] Experiment 2: Feature extraction and matching accuracy test
[0207] Objective: To verify the feature extraction effect of 3D-LBP algorithm and the accuracy of feature vector matching.
[0208] The specific method is as follows:
[0209] 1. Select 100 volunteers, each of whom collects iris images at different times (morning, noon, evening) and different angles (frontal, left side, right side), and collect iris images 30 times for each person, for a total of 3,000 samples;
[0210] 2. Use 3D-LBP and 2D-LBP algorithms to extract the feature vector of each sample;
[0211] 3. Use cosine similarity to calculate the similarity between feature vectors and evaluate the matching accuracy;
[0212] Table 2 Results of Experiment 2
[0213]
[0214] Through Table 2 and Figure 3 It can be seen that compared with the 2D-LBP algorithm, the 3D-LBP algorithm can extract iris features more accurately, and the accuracy of feature vector matching is high, which ensures the reliability of iris recognition, and the false acceptance rate and false rejection rate are both within an acceptable range.
[0215] Experiment 3: Test the effect of dynamic update mechanism of feature library
[0216] Purpose: To verify the effectiveness of the dynamic update mechanism of the signature database and ensure the accuracy and currency of the signature database.
[0217] Here’s how:
[0218] 1. Select 20 volunteers, each of whom collects iris images 10 times a day for 30 days, for a total of 6,000 samples;
[0219] 2. Store the collected iris feature vectors in a temporary storage area and update the feature library regularly;
[0220] 3. Compare the matching accuracy of the feature library before and after the dynamic update mechanism is introduced;
[0221] Table 3 Results of Experiment 3
[0222]
[0223] Through Table 3 and Figure 4 It can be seen that through the dynamic update mechanism, the feature library has been effectively updated, the matching accuracy has been significantly improved, and the long-term stability and accuracy of the system have been ensured.
[0224] Experiment 4: System overall performance test
[0225] Purpose: To verify the overall performance of the system, including recognition speed, unlocking success rate and data transmission stability.
[0226] Here’s how:
[0227] 1. Select 50 volunteers, each of whom performs 30 iris recognition and unlocking operations using the method described in the present invention and the method described in the document "Wang Yimou. Design and Implementation of Iris Recognition System Device End [D]. Beijing University of Posts and Telecommunications, 2017.", for a total of 1500 times;
[0228] 2. Measure and compare the recognition time, unlocking success rate and data transmission time;
[0229] 3. Upload unlocking records and user information to the cloud server through the wireless communication module to evaluate the stability and security of data transmission.
[0230] Table 4 Results of Experiment 4
[0231]
[0232] Through Table 4 and Figure 5 It can be seen that the overall performance of the system is excellent. On the basis of almost no sacrifice in recognition speed, the unlocking success rate is higher and the data transmission is more stable and secure, which fully verifies the practicability and reliability of the present invention.
[0233] Through the above-mentioned experimental tests, the present invention can stably collect high-quality iris images under different lighting conditions, the 3D-LBP algorithm can efficiently extract iris features, the dynamic update mechanism effectively maintains the latest and accuracy of the feature library, and the overall performance is excellent. These results verify the practicality of the present invention.
[0234] The present invention is intended to cover all such substitutions, modifications and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.
Claims
1. An iris recognition control opening and closing method based on a smart lock, characterized in that: The following steps are involved: S1: Collecting the user's iris image through the iris acquisition module on the smart lock, where the iris acquisition module includes a high-resolution camera and an infrared light source to ensure that clear images are collected under various lighting conditions; S2: Perform multimodal fusion processing on the iris image collected by S1, and fuse the infrared spectrum image and the visible spectrum image to enhance the clarity of the iris features; Specifically include: S21: preprocessing the infrared spectrum image and the visible spectrum image obtained from S1, specifically including grayscale transformation, noise filtering and image enhancement to improve the image quality; S22: extracting features from the preprocessed infrared spectrum image and visible spectrum image respectively, and using a preset image feature extraction algorithm to obtain feature information of each image; S23: using a preset feature matching algorithm, matching the features of the infrared spectrum image and the visible spectrum image, aligning corresponding points in the two images, and ensuring that the two images are consistent in spatial position; S24: Determine a fusion strategy based on the feature matching result, specifically use a preset weighted average method to assign weights to the infrared spectrum image and the visible spectrum image, and the weight assignment is based on parameters including image contrast, brightness, and feature richness; then fuse the infrared spectrum image and the visible spectrum image, and synthesize a fused image by using the weighted average method to ensure that the fused image retains the features of the two spectrum images; S241: Calculate contrast, brightness, and feature richness of the preprocessed infrared spectrum image and visible spectrum image; The contrast ratio is calculated by calculating the standard deviation of the image gray value, and the calculation formula is: , ; in, is the contrast of infrared spectrum image; Represents the infrared spectrum image The gray value of pixels; is the average gray value of the infrared spectrum image; is the total number of pixels of the infrared spectrum image; is the contrast of the visible spectrum image; Represents the visible spectrum image The gray value of pixels; is the average gray value of the visible spectrum image; is the total number of pixels of the visible spectrum image; The brightness is calculated by the method of calculating the average gray value of the image, and the formula is: , ; in, is the brightness of the infrared spectrum image; The infrared spectrum image The gray value of pixels; is the brightness of the visible spectrum image; The visible spectrum image The gray value of pixels; The feature richness uses an edge detection algorithm to calculate the number of edges in the image, and the formula is: , , in, is the richness of infrared spectral image; Represents the infrared spectrum image A binary flag indicating whether a pixel is an edge pixel; is the richness of the visible spectrum image; Represents the visible spectrum image A binary flag indicating whether a pixel is an edge pixel; S242: According to the parameters calculated in S241, weights are assigned to the infrared spectrum image and the visible spectrum image. The weight of the infrared spectrum image is , the weight of the visible spectrum image is , then the weight calculation formulas are: Contrast Weight: , ; Luminance Weight: , ; Feature richness weight: , ; S243: Comprehensively combine various weight parameters to obtain the final fusion weight, and set the comprehensive weight to and , then the calculation formula of the comprehensive weight is: , ; in, , , are the weight coefficients of contrast, brightness and feature richness, which need to be satisfied ; S244: According to the comprehensive weight calculated in S243, the infrared spectrum image and the visible spectrum image are weighted average fused, and the fused image The calculation formula is: , in, is the gray value matrix of the infrared spectrum image; is the gray value matrix of the visible spectrum image; S25: Optimizing the fused image, including contrast enhancement and detail enhancement, to provide high-quality image data for subsequent iris positioning and feature extraction; S3: Using a preset positioning algorithm to locate the iris image, extracting the iris area, and performing real-time normalization processing; S4: Using 3D-LBP algorithm to extract features from the normalized iris image and generate a high-dimensional iris feature vector; S5: dynamically matching the extracted iris feature vector with the pre-registered iris feature library stored in the smart lock, using a preset encryption transmission and processing technology, and regularly updating the iris feature library through a dynamic update mechanism; S6: Perform user identity authentication based on the matching result of S5, and determine whether the user has the unlocking authority through the control module of the smart lock. If the verification is successful, the control module sends an unlocking command to the actuator to drive the motor to perform the unlocking operation. After the unlocking is successful, the unlocking record and user information are uploaded to the cloud server through the wireless communication module.
2. The iris recognition control opening and closing method based on the smart lock according to claim 1 is characterized in that: The S1 specifically includes: S11: The iris acquisition module on the smart lock includes an ambient light sensor, which is used to detect the light intensity of the surrounding environment in real time and determine whether the current lighting conditions are suitable for iris acquisition; S12: According to the ambient light intensity detected by S11, the control unit of the iris acquisition module automatically adjusts the brightness and angle of the infrared light source to provide optimal lighting conditions, ensuring that a clear iris image can be obtained even in low-light or backlit environments; S13: a high-resolution camera in the iris acquisition module, used to capture an iris image of the user, the high-resolution camera having a resolution of at least 1080P; S14: During the acquisition process, a high-resolution camera is used to capture infrared spectrum images and visible spectrum images at the same time. By synchronously acquiring images of the two spectrums, complete iris image information is ensured under different lighting conditions; S15: The iris image obtained in S14 is detected by an image quality detection algorithm to ensure the clarity, contrast and noise-freeness of the image. When the image quality does not meet the preset standard, return to S12 for re-collection until an iris image that meets the quality requirements is obtained.
3. The method for controlling opening and closing of an intelligent lock by iris recognition according to claim 1 is characterized in that: The S3 specifically includes: S31: Use the improved Canny edge detection algorithm to perform edge detection on the iris image and extract the edges of the iris and pupil. The specific formula is: , in, is the gradient amplitude; For images in Directional gradient; For images in Directional gradient; S32: Perform circular Hough transform on the edge image extracted in S31 to identify the inner and outer boundaries of the iris. The formula of circular Hough transform is: ,in, is the coordinate of the center of the circle; is the radius; is the coordinate of the edge point; S33: extracting the iris region according to the inner and outer boundaries of the iris identified in S32, generating an image containing only the iris portion, and determining the start and end coordinates of the iris region by the inner and outer boundaries for cutting; S34: performing normalization processing on the extracted iris region, converting iris images of different sizes and shapes into uniform sizes and shapes, specifically using a polar coordinate conversion method, and the normalization formula is: , in, is the normalized iris image; is the original iris image; The conversion formula from polar coordinates to rectangular coordinates is ; The conversion formula from polar coordinates to rectangular coordinates is ; S35: Correcting the normalized iris image to eliminate distortion caused by shooting angle or user posture, and ensuring that all iris images have the same alignment and ratio.
4. The iris recognition control opening and closing method based on the smart lock according to claim 1 is characterized in that: The S4 specifically includes: S41: dividing the normalized iris image into a plurality of small blocks, each small block containing a predetermined number of pixels for extracting local features; S42: performing 3D local binary pattern coding processing on each block of the iris image, wherein the 3D-LBP generates a binary code for each pixel by analyzing the grayscale changes of the image block at different scales and directions; S43: combining the 3D-LBP coding values of all pixels in each block into a coding map, where the coding map is used to represent local features of the iris image at different scales and directions; S44: performing statistical analysis on the coding map, calculating the occurrence frequency of each coding pattern in the entire iris image, generating a feature histogram, and combining the data in the feature histogram into a high-dimensional feature vector, which is used to represent the local texture features and global feature distribution of the iris image; S45: Optimizing the generated high-dimensional feature vector, reducing the dimension of the feature vector by using dimensionality reduction technology, retaining predetermined feature information, and removing redundant and noise data.
5. The iris recognition control opening and closing method based on the smart lock according to claim 4 is characterized in that: The step of 3D local binary pattern encoding processing in S42 includes: S421: Select a pixel in each block as a central pixel; S422: Collecting neighborhood pixels around the central pixel according to a preset radius and direction; S423: Compare the grayscale value of the neighboring pixel with the grayscale value of the central pixel. If the grayscale value of the neighboring pixel is greater than or equal to the grayscale value of the central pixel, the position of the neighboring pixel is recorded as 1, otherwise it is recorded as 0; S424: Combining the binary results of all neighboring pixels into a binary number as the 3D-LBP encoding value of the central pixel.
6. The iris recognition control opening and closing method based on the smart lock according to claim 5 is characterized in that: The S44 specifically includes: S441: Perform statistical analysis on the coding map generated for each block, and calculate the frequency of occurrence of each 3D-LBP coding mode in the block. The specific method is to traverse each pixel in the coding map and record the number of times each coding mode appears. First, preset the first pixel in the coding map. The number of times the encoding mode appears in the block is ; S442: The number of occurrences of each coding mode counted in S441 Divide by the total number of pixels in the tile , calculate the frequency of each coding mode , the calculation formula is as follows: ; S443: The frequencies of all coding modes in each block are Combined into a feature histogram, the histogram represents the feature distribution of the iris image in different local areas. Suppose there are encoding mode, the feature histogram is a A vector of entries , the formula is: ; S444: Merge the feature histograms of all the blocks to generate a global feature histogram of the entire iris image. Specifically, accumulate the frequencies of the same coding mode in different blocks to obtain the coding mode frequency distribution in the global range. The frequency of this coding pattern in the entire iris image is , the calculation formula is: ,in, is the total number of blocks; For the The encoding mode is The frequency in each block; S445: Arrange the data in the global feature histogram in order to form a high-dimensional feature vector. The frequency of each encoding mode As a dimension in the feature vector, the feature vector is used to represent the overall texture feature distribution of the iris image. It is expressed as: ,in, is a high-dimensional feature vector.
7. The iris recognition control opening and closing method based on the smart lock according to claim 1 is characterized in that: The S5 specifically includes: S51: Extract the iris feature vector The feature vector stored in the pre-registered iris feature library in the smart lock Compare them one by one and calculate the similarity between feature vectors. Specifically, cosine similarity is used to calculate the similarity value. , the calculation formula is: ,in, is the newly extracted iris feature vector; The first eigenvectors; · is the vector dot product; is the magnitude of the vector; S52: Calculate the similarity value Similarity threshold with the preset For comparison, if Greater than or equal to threshold , then the match is considered successful and the user's identity is identified; otherwise, the match fails; S53: Regularly update the pre-registered iris feature database dynamically to maintain the accuracy and latest status of the feature database. The dynamic update mechanism includes the following steps: S531: The iris feature vectors successfully matched within a predetermined time are Store in temporary storage area; S532: Perform statistical analysis on the feature vectors in the temporary storage area, calculate the frequency and stability of the feature vectors of each user, and select feature vectors with high frequency and stability as update candidates; S533: Compare the updated candidate feature vector with the corresponding feature vector in the existing feature library. If the new feature vector has higher stability and matching accuracy, replace the old feature vector with the new feature vector to update the feature library. S54: According to actual application requirements, set periodic parameters for updating the feature library to ensure that the feature library can reflect changes in the user's iris features in a timely manner.
8. The method for controlling opening and closing of an intelligent lock by iris recognition according to claim 1, characterized in that: The S6 specifically includes: S61: receiving the matching result of the similarity judgment in S5, if the matching is successful, proceeding to the next step of user identity verification; if the matching fails, refusing to unlock and issuing a warning signal; S62: The control module searches for corresponding user authority information from a pre-stored user authority database based on the successfully matched user feature vector, and verifies whether the user has unlocking authority, including user identity verification and authority level verification. Only users with unlocking authority can proceed to the next step; S63: If the user passes the authority verification, the control module will generate an unlocking instruction and send the instruction to the driving circuit of the actuator. The instruction includes specific operation steps of starting the motor and opening the lock cylinder; S64: After receiving the unlocking command sent by the control module, the actuator starts the motor to drive the lock cylinder to rotate, completing the unlocking operation; S65: After the motor completes the unlocking operation, the sensor in the actuator detects the state of the lock cylinder to confirm whether the unlocking is successful. If the unlocking is successful, the sensor feeds back a success signal to the control module; S66: After receiving the unlocking success signal, the control module generates an unlocking record, including information on user identity, unlocking time, and unlocking location; S67: Finally, the generated unlocking record and user information are uploaded to the cloud server through wireless communication technology.
