A hotel self-service check-in intelligent registration management system and method

Abstract

The present invention discloses a hotel self-service check-in intelligent registration management system and method, which relates to the field of registration management technology. The combination of a camera and a secondary camera captures information from the front and side, making the recognition of facial features more comprehensive, and effectively solving the problem of feature loss caused by a single angle in the traditional system. By fusing the front and side features, a more complete facial total feature vector V is formed. Z , improving the system's ability to recognize complex facial features. Combining convolutional neural networks for feature matching and taking advantage of deep learning, it significantly improves recognition accuracy and speed, and has strong adaptability. Based on the weighted accumulation similarity score calculation method, the weight factor can be flexibly adjusted to optimize the matching effect for different situations, ensuring the intelligence and efficiency of the system. The system's dynamic database traversal mechanism when it fails to recognize ensures the final confirmation of the customer's identity, improving user experience and security.

Description

Technical Field

[0001] The present invention relates to the technical field of registration management, and in particular to a hotel self-service check-in intelligent registration management system and method. Background Art

[0002] The hotel self-service check-in smart registration management system is a direct reflection of this trend. It belongs to the field of intelligent service systems and aims to improve customer check-in experience and hotel operation efficiency. In this system, the application of face recognition technology has greatly simplified the traditional check-in process, thus realizing convenient services without ID card or credit card verification.

[0003] Long queues and cumbersome procedures may lead to customer dissatisfaction, affecting the hotel's reputation and return rate. Moreover, most hotels currently still rely on traditional manual verification methods during the check-in process. Guests are required to show their ID or credit card, which is not only time-consuming and labor-intensive, but also increases security risks. Traditional registration methods often lead to long waiting times, affecting customers' check-in experience. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides a hotel self-service check-in intelligent registration management system and method, which solves the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: a hotel self-service check-in intelligent registration and management system, including a facial feature collection module, a facial feature matching module, an identity authentication module, a room allocation and payment confirmation module and an intelligent unlocking and check-in confirmation module;

[0006] The facial feature acquisition module uses a high-resolution camera to collect the guest's facial features and generate a total facial feature vector V Z , where the high-resolution cameras include a main camera and two sub-cameras, and the high-resolution cameras are arranged in a triangle;

[0007] The facial feature matching module uses a convolutional neural network to obtain the total facial feature vector V Z Compare with the stored guest data in the hotel database and return the similarity score;

[0008] The identity authentication module calls the identity information in the database based on the information obtained by the face feature matching module, and converts the collected facial total feature vector V Z Compare with the data on the Internet to further verify the guest's identity;

[0009] The room allocation and payment confirmation module automatically allocates rooms according to the identity information after successful recognition, and prompts the guest to pay the room fee through mobile payment;

[0010] The smart unlocking and check-in confirmation module sends the unlocking permission to the guest's mobile phone when the payment is successful, and the guest can unlock the door lock through the mobile phone App.

[0011] Preferably, the facial feature acquisition module includes a multi-angle camera acquisition unit and a feature vector fusion unit;

[0012] The multi-angle camera acquisition unit captures the guest's facial image from different angles through a main camera and two auxiliary cameras to generate a facial feature vector V i ; Among them, the main camera captures the front image features and generates the front facial feature vector V M , the two secondary cameras capture the side features from the left and right angles respectively, and generate the left facial feature vector V L and the right facial feature vector V R , combine these three images to obtain more complete facial information to reduce the loss of feature points caused by angle problems;

[0013] For the facial feature vector V i Expressed as: V i ;

[0014] In the formula, Represents the N identified in the image f feature points, N f Indicates the number of feature points;

[0015] The feature vector fusion unit extracts features from the images captured by the main camera and the auxiliary camera, and then uses the feature vector fusion algorithm to combine the frontal facial feature vector V M , left facial feature vector V L and the right facial feature vector V R Integrate into the total facial feature vector V Z ;

[0016] The total facial feature vector V Z Obtained by the following formula:

[0017] V Z ;

[0018] In the formula, Represents the weight value of the main camera, represents the weight values ​​of the two secondary cameras, and =1, .

[0019] Preferably, the facial feature matching module includes a facial feature matching unit and a similarity scoring unit;

[0020] The facial feature matching unit uses a convolutional neural network to collect the total facial feature vector V Z Compare it with the feature vector stored in the database to extract the total facial feature vector V Z The vector value F of the feature point Zi , extract the vector value F of the feature point in the feature vector stored in the database Si , calculate the difference between the two ;

[0021] The difference Obtained by the following formula:

[0022] ;

[0023] In the formula, F Zi Represents the total facial feature vector V Z The value of the i-th feature point in F Si Represents the value of the i-th feature point in the feature vector stored in the database;

[0024] By calculating the difference of each feature point, we get the similarity score S M , used to determine whether the collected facial features best match the faces stored in the database;

[0025] The similarity scoring unit calculates the difference between each feature point Perform weighted accumulation to calculate the final similarity score S M , to judge the guest’s identity. The lower the score, the more similar the collected facial features are to those in the database;

[0026] The similarity score S M Obtained by the following formula:

[0027] S M ;

[0028] Where n represents the number of facial feature points, It represents the weight factor, which reflects the importance of each feature point. The weights of different feature points may be different. For example, the weights of feature points such as eyes, nose, and mouth are larger, while the weights of features such as contours are smaller.

[0029] When the similarity score S M <Threshold T M , it means the system has successfully identified the guest’s identity;

[0030] When the similarity score S M >Threshold T M, it means that the system has not recognized the guest's identity. The system will traverse all the stored feature vectors in the database and recalculate the similarity score S M , until the similarity score S M <Threshold T M Until then, the match is successful and the guest's identity information is returned.

[0031] Preferably, the identity authentication module includes an identity information matching unit and an identity authentication probability calculation unit;

[0032] The identity information matching unit calls the identity information in the database based on the face feature matching module and compares it with the collected total facial feature vector V Z Further comparison is performed, and the system cross-verifies the collected facial features with the identity data in the external network database to ensure that the guest's identity information is up-to-date and valid;

[0033] When the similarity score S returned by the facial feature matching unit M <Threshold T M =0.85, the system calls the internal database and cross-checks the guest identity information in the external network database to ensure that the guest identity is up-to-date and correct;

[0034] The identity authentication probability calculation unit is based on the similarity score S M , calculate the probability of successful authentication P through the probability model S ,The identity authentication probability calculation unit uses an exponential decay function to calculate the probability of successful ,authentication and determine the guest’s identity;

[0035] According to the similarity score S M , calculate the probability of successful verification P S ;

[0036] The probability P S Obtained by the following formula:

[0037] P S ;

[0038] In the formula, k represents the adjustment coefficient;

[0039] When the similarity score S M The lower the probability of successful authentication, P S The closer to 1;

[0040] When the similarity score S M The higher the probability of successful authentication, P S The closer to 0.

[0041] Preferably, the room allocation and payment confirmation module includes a room allocation unit and a payment confirmation unit;

[0042] The room allocation unit allocates the room according to the number of remaining rooms N on that day. a and the total number of hotel rooms N n , calculate the room availability A R , based on room availability A R And preset room allocation rules to select the most suitable room for guests;

[0043] The room availability rate A R Obtained by the following formula:

[0044] A R ;

[0045] Room availability A R It indicates the proportion of available rooms in the current hotel. The higher the value, the more rooms can be allocated to guests. R When the room availability rate is low, the system may prioritize allocating smaller and more economical rooms to improve room utilization. R , dynamically adjust the room allocation strategy to ensure that hotel room resources are maximized and the occupancy rate is improved.

[0046] Preferably, the payment confirmation unit prompts the guest to pay the room fee through the mobile payment platform according to the room assigned by the room assignment unit; the payment success rate P P Depending on the payment platform and the current network environment, the system will dynamically detect whether the payment is successful;

[0047] Payment success rate P P Determined by the current payment environment, mainly depending on the availability of the payment platform and the network environment:

[0048] P P ;

[0049] In the formula, represents the stability coefficient of the payment platform, t represents the payment time, represents the indicator function;

[0050] Payment success rate P P Affected by network conditions and the stability of the payment platform, the longer the payment time t is, the higher the payment success rate P P Will increase over time.

[0051] Preferably, the smart unlocking and check-in confirmation module includes a permission generation unit, a mobile app unlocking unit and a real-time feedback and support unit;

[0052] After the payment confirmation unit successfully pays, the permission generation unit generates a unique unlocking permission code Uk, and sends the unlocking permission code Uk and related information including the room number and the permission validity period T U , sent to the guest’s mobile app;

[0053] The judgment formula of the authority validity of the authority generation unit is:

[0054] Effectiveness ;

[0055] Where, T t Indicates the difference between the current time and the permission generation time;

[0056] When payment is successful, the management system of the smart unlocking and check-in confirmation module sends the unlocking permission to the guest's mobile phone, and the guest can unlock the door lock through the mobile phone App.

[0057] Preferably, the mobile app unlocking unit requires the guest to input and automatically send the unlocking authority code Uk through the mobile app for verification with the room door lock system;

[0058] When the unlock code entered by the guest is the same as the unlock authority code Uk, the door will be opened;

[0059] When the unlock code entered by the guest is different from the unlock authority code Uk, the door will not be opened. If the unlock code entered three times does not correspond to the unlock authority code Uk, the door will be locked and a unique unlock authority code Uk needs to be regenerated.

[0060] Preferably, the real-time feedback and support unit provides a real-time feedback channel, allowing guests to contact the hotel customer service immediately when they encounter problems during the self-service check-in process to ensure that the problem is solved in a timely manner;

[0061] The guest’s feedback will be recorded for subsequent service adjustments;

[0062] ;

[0063] In the formula, Indicates the guest's feedback record. Represents the jth piece of feedback information.

[0064] A hotel self-service check-in intelligent registration management method comprises the following steps:

[0065] Step 1: The facial feature acquisition module uses a high-resolution camera to collect the guest's facial features and generate a total facial feature vector V Z , where the high-resolution cameras include a main camera and two sub-cameras, and the high-resolution cameras are arranged in a triangle;

[0066] Step 2: The facial feature matching module uses a convolutional neural network to obtain the total facial feature vector V Z Compare with the stored guest data in the hotel database and return the similarity score;

[0067] Step 3: The identity authentication module calls the identity information in the database based on the information obtained by the face feature matching module and converts the collected facial feature vector V Z Compare with the data on the Internet to further verify the guest's identity;

[0068] Step 4: The room allocation and payment confirmation module automatically allocates rooms based on the identity information after successful recognition, and prompts the guest to pay the room fee through mobile payment;

[0069] Step 5: Smart unlocking and check-in confirmation module When the payment is successful, the unlocking permission will be sent to the guest’s mobile phone, and the guest can unlock the door lock through the mobile app.

[0070] The present invention provides a hotel self-service check-in intelligent registration management system and method, which has the following beneficial effects:

[0071] (1) The use of high-resolution cameras arranged in a triangle improves the accuracy of facial feature acquisition and ensures that more detailed information is captured. Feature matching through convolutional neural networks greatly improves the speed and accuracy of facial recognition and reduces the possibility of misidentification. Combining network data for identity confirmation increases the level of verification and improves the reliability of identity confirmation. The system allocates rooms based on real-time data, improves operational efficiency and reduces human errors. Unlocking the door lock through the mobile phone app provides great convenience for customers and further improves the check-in experience.

[0072] (2) The combination of the front camera and the secondary camera captures information from the front and side, making the recognition of facial features more comprehensive and effectively solving the problem of feature loss caused by a single angle in traditional systems. By fusing the front and side features, a more complete facial feature vector V is formed. Z , improving the system's ability to recognize complex facial features. Combining convolutional neural networks for feature matching and taking advantage of deep learning, it significantly improves recognition accuracy and speed, and has strong adaptability.

[0073] Based on the weighted cumulative similarity score calculation method, the weight factor can be flexibly adjusted to optimize the matching effect for different situations, ensuring the intelligence and efficiency of the system. The dynamic database traversal mechanism when the system fails to identify ensures the final confirmation of the customer's identity and improves user experience and security.

[0074] (3) By cross-verifying identity information with external network databases, the system ensures that the data used is up-to-date and valid, thereby improving the security and accuracy of identification. The identity authentication probability calculation unit calculates the identity information based on the similarity score S M , through the operation of exponential decay function, the authentication result is more scientific and reliable. The room availability rate A is calculated by the ratio of the real-time room number to the total room number R , making room allocation more flexible and efficient, and improving hotel resource utilization. The design of the payment confirmation unit allows the system to monitor the payment success rate P in real time P , ensuring a smooth payment experience in various network environments. R Dynamically adjusting room allocation can meet the needs of different guests and improve customer satisfaction and hotel operation efficiency.

[0075] (4) Generate a unique unlocking authorization code Uk after payment confirmation, ensuring the security of each guest's identity and room information and avoiding potential security loopholes. Customers unlock door locks through mobile phone apps, reducing the use of traditional keys and improving the convenience and modernity of check-in. The real-time feedback and support unit allows customers to contact the hotel in time when they encounter problems during self-service check-in, reducing customer anxiety and enhancing customer satisfaction. Limit the number of unlocking code inputs to prevent criminals from attempting to unlock the room, ensuring customer safety and privacy. By recording and analyzing customer feedback, hotels can continuously optimize service processes and improve the overall customer experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] Figure 1 This is a flowchart diagram of a hotel self-service check-in intelligent registration management system of the present invention;

[0077] Figure 2 The present invention is a schematic diagram of the steps of a hotel self-service check-in intelligent registration management method. DETAILED DESCRIPTION

[0078] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0079] Example 1

[0080] The present invention provides a hotel self-service check-in intelligent registration management system, please refer to Figure 1 , including facial feature acquisition module, facial feature matching module, identity authentication module, room allocation and payment confirmation module and smart unlocking and check-in confirmation module;

[0081] The facial feature acquisition module uses a high-resolution camera to collect the guest's facial features and generate a total facial feature vector V Z , where the high-resolution cameras include a main camera and two sub-cameras, and the high-resolution cameras are arranged in a triangle;

[0082] The facial feature matching module uses a convolutional neural network to obtain the total facial feature vector V Z Compare with the stored guest data in the hotel database and return the similarity score;

[0083] The identity authentication module calls the identity information in the database based on the information obtained by the face feature matching module, and converts the collected facial total feature vector V Z Compare with the data on the Internet to further verify the guest's identity;

[0084] The room allocation and payment confirmation module automatically allocates rooms according to the identity information after successful recognition, and prompts the guest to pay the room fee through mobile payment;

[0085] The smart unlocking and check-in confirmation module sends the unlocking permission to the guest's mobile phone when the payment is successful, and the guest can unlock the door lock through the mobile phone App.

[0086] In this embodiment, a high-resolution camera arranged in a triangle is used to improve the accuracy of facial feature acquisition and ensure that more detailed information is captured. Feature matching through convolutional neural networks greatly improves the speed and accuracy of facial recognition and reduces the possibility of misidentification. Combining network data for identity confirmation increases the level of verification and improves the reliability of identity confirmation. The system allocates rooms based on real-time data, improves operational efficiency and reduces human errors. Unlocking the door lock through the mobile phone app provides great convenience for customers and further enhances the check-in experience.

[0087] The self-service check-in system reduces the time customers spend waiting at the front desk, making the check-in process faster and smoother. High-precision facial recognition and identity verification reduce the risk of impersonation and ensure customer safety. Convenient mobile payment and room unlocking functions are provided to improve customer satisfaction and convenience. The automated check-in management process reduces dependence on front desk staff and reduces human resource expenditure. The integrated database system updates guest information in real time, improving the accuracy and efficiency of information management. Introducing advanced technical means to enhance the hotel's modern image and attract more young customers.

[0088] Example 2

[0089] This embodiment is explained in Example 1, please refer to Figure 1 ,Specifically: the facial feature acquisition module includes a multi-angle camera acquisition unit and a feature vector fusion unit;

[0090] The multi-angle camera acquisition unit captures the guest's facial image from different angles through a main camera and two auxiliary cameras to generate a facial feature vector V i ; Among them, the main camera captures the front image features and generates the front facial feature vector V M , the two secondary cameras capture the side features from the left and right angles respectively, and generate the left facial feature vector V L and the right facial feature vector V R , combine these three images to obtain more complete facial information to reduce the loss of feature points caused by angle problems;

[0091] For the facial feature vector V i Expressed as: V i ;

[0092] In the formula, Represents the N identified in the image f feature points, N f Indicates the number of feature points;

[0093] The feature vector fusion unit extracts features from the images captured by the main camera and the auxiliary camera, and then uses the feature vector fusion algorithm to combine the frontal facial feature vector V M , left facial feature vector V L and the right facial feature vector V R Integrate into the total facial feature vector V Z ;

[0094] The total facial feature vector V Z Obtained by the following formula:

[0095] V Z ;

[0096] In the formula, Represents the weight value of the main camera, represents the weight values ​​of the two secondary cameras, and =1, .

[0097] The facial feature matching module includes a facial feature matching unit and a similarity scoring unit;

[0098] The facial feature matching unit uses a convolutional neural network to collect the total facial feature vector V Z Compare it with the feature vector stored in the database to extract the total facial feature vector VZ The vector value F of the feature point Zi , extract the vector value F of the feature point in the feature vector stored in the database Si , calculate the difference between the two ;

[0099] The difference Obtained by the following formula:

[0100] ;

[0101] In the formula, F Zi Represents the total facial feature vector V Z The value of the i-th feature point in F Si Represents the value of the i-th feature point in the feature vector stored in the database;

[0102] By calculating the difference of each feature point, we get the similarity score S M , used to determine whether the collected facial features best match the faces stored in the database;

[0103] The similarity scoring unit calculates the difference between each feature point Perform weighted accumulation to calculate the final similarity score S M , to judge the guest’s identity. The lower the score, the more similar the collected facial features are to those in the database;

[0104] The similarity score S M Obtained by the following formula:

[0105] S M ;

[0106] Where n represents the number of facial feature points, represents the weight factor;

[0107] When the similarity score S M <Threshold T M , it means the system has successfully identified the guest’s identity;

[0108] When the similarity score S M >Threshold T M , it means that the system has not recognized the guest's identity. The system will traverse all the stored feature vectors in the database and recalculate the similarity score S M , until the similarity score S M <Threshold T M Until then, the match is successful and the guest's identity information is returned.

[0109] In this embodiment, the use of multi-angle cameras enables more comprehensive capture of facial features, significantly improves recognition accuracy, and reduces feature point loss due to angle problems. Through the feature vector fusion algorithm, facial features from different cameras are effectively integrated to generate a more representative total facial feature vector V Z , which improves the recognition ability of the system. The use of convolutional neural networks for feature matching accelerates the comparison process with the feature vectors stored in the database and shortens the waiting time for customers to check in. The similarity score calculation method makes identity judgment more accurate by weighted accumulation of the difference values ​​of each feature point, which improves the intelligence level of the system.

[0110] Set the similarity threshold to ensure that the system can accurately identify the guest's identity, while preventing misidentification and ensuring the safety of the customer. When the recognition fails, the system will automatically traverse the database and recalculate the similarity score to ensure the maximum recognition of the customer's identity and improve the system's fault tolerance.

[0111] The combination of the main camera and the secondary camera captures information from the front and side, making the recognition of facial features more comprehensive, effectively solving the problem of feature loss caused by a single angle in traditional systems. By fusing the front and side features, a more complete facial feature vector V is formed. Z , improving the system's ability to recognize complex facial features. Combining convolutional neural networks for feature matching and taking advantage of deep learning, it significantly improves recognition accuracy and speed, and has strong adaptability.

[0112] Based on the weighted cumulative similarity score calculation method, the weight factor can be flexibly adjusted to optimize the matching effect for different situations, ensuring the intelligence and efficiency of the system. The dynamic database traversal mechanism when the system fails to identify ensures the final confirmation of the customer's identity and improves user experience and security.

[0113] Example 3

[0114] This embodiment is explained in Example 2. Please refer to Figure 1 ,Specifically: the identity authentication module includes an identity information matching unit and an identity authentication probability calculation unit;

[0115] The identity information matching unit calls the identity information in the database based on the face feature matching module and compares it with the collected total facial feature vector V Z Further comparison is performed, and the system cross-verifies the collected facial features with the identity data in the external network database to ensure that the guest's identity information is up-to-date and valid;

[0116] When the similarity score S returned by the facial feature matching unit M <Threshold TM =0.85, the system calls the internal database and cross-checks the guest identity information in the external network database to ensure that the guest identity is up-to-date and correct;

[0117] The identity authentication probability calculation unit is based on the similarity score S M , calculate the probability of successful authentication P through the probability model S ,The identity authentication probability calculation unit uses an exponential decay function to calculate the probability of successful ,authentication and determine the guest’s identity;

[0118] According to the similarity score S M , calculate the probability of successful verification P S ;

[0119] The probability P S Obtained by the following formula:

[0120] P S ;

[0121] In the formula, k represents the adjustment coefficient;

[0122] When the similarity score S M The lower the probability of successful authentication, P S The closer to 1;

[0123] When the similarity score S M The higher the probability of successful authentication, P S The closer to 0.

[0124] The room allocation and payment confirmation module includes a room allocation unit and a payment confirmation unit;

[0125] The room allocation unit allocates the room according to the number of remaining rooms N on that day. a and the total number of hotel rooms N n , calculate the room availability A R , based on room availability A R And preset room allocation rules to select the most suitable room for guests;

[0126] The room availability rate A R Obtained by the following formula:

[0127] A R ;

[0128] By room availabilityA R, dynamically adjust the room allocation strategy to ensure that the hotel room resources are maximized and the occupancy rate is improved. The payment confirmation unit prompts the guest to pay the room fee through the mobile payment platform according to the room allocated by the room allocation unit; the payment success rate P P Depending on the payment platform and the current network environment, the system will dynamically detect whether the payment is successful;

[0129] Payment success rate P P Determined by the current payment environment, mainly depending on the availability of the payment platform and the network environment:

[0130] P P ;

[0131] In the formula, represents the stability coefficient of the payment platform, t represents the payment time, represents the indicator function;

[0132] Payment success rate P P Affected by network conditions and the stability of the payment platform, the longer the payment time t is, the higher the payment success rate P P Will increase over time.

[0133] In this embodiment, by cross-verifying the identity information in the internal and external databases, the customer's identity information is ensured to be up-to-date and accurate, and the risk of misidentification is reduced. The combination of the identity information matching unit and the probability calculation unit improves the intelligent level of identity authentication and makes the identification process more efficient. Based on the real-time room availability calculation, the system can flexibly adjust the room allocation strategy to maximize the use of hotel room resources and improve the occupancy rate. Dynamic detection of payment success rate P P The mechanism can provide timely feedback on payment status, reduce customer waiting time, and improve user experience. By using an exponential decay function to calculate the probability of successful identity authentication, a more scientific basis for judgment is provided, which is helpful for risk management. The dynamic adjustment capability for different network environments and payment platforms enables the system to maintain efficient operation under various conditions, improving the stability and reliability of the system.

[0134] By cross-verifying identity information with external network databases, the system ensures that the data used is up-to-date and valid, thereby improving the security and accuracy of identification. M , through the operation of exponential decay function, the authentication result is more scientific and reliable. The room availability rate A is calculated by the ratio of the real-time room number to the total room number R , making room allocation more flexible and efficient, and improving hotel resource utilization. The design of the payment confirmation unit allows the system to monitor the payment success rate P in real time P, ensuring a smooth payment experience in various network environments. R Dynamically adjusting room allocation can meet the needs of different guests and improve customer satisfaction and hotel operation efficiency.

[0135] Example 4

[0136] This embodiment is explained in Example 3, please refer to Figure 1 ,Specifically: the smart unlocking and check-in confirmation module includes a permission generating unit, a mobile App unlocking unit and a real-time feedback and support unit;

[0137] After the payment confirmation unit successfully pays, the permission generation unit generates a unique unlocking permission code Uk, and sends the unlocking permission code Uk and related information including the room number and the permission validity period T U , sent to the guest’s mobile app;

[0138] The judgment formula of the authority validity of the authority generation unit is:

[0139] Effectiveness ;

[0140] Where, T t Indicates the difference between the current time and the permission generation time;

[0141] When payment is successful, the management system of the smart unlocking and check-in confirmation module sends the unlocking permission to the guest's mobile phone, and the guest can unlock the door lock through the mobile phone App.

[0142] The mobile app unlocking unit requires the guest to input and automatically send the unlocking authority code Uk through the mobile app for verification with the room door lock system;

[0143] When the unlock code entered by the guest is the same as the unlock authority code Uk, the door will be opened;

[0144] When the unlock code entered by the guest is different from the unlock authority code Uk, the door will not be opened. If the unlock code entered three times does not correspond to the unlock authority code Uk, the door will be locked and a unique unlock authority code Uk needs to be regenerated.

[0145] The real-time feedback and support unit provides a real-time feedback channel, allowing guests to contact the hotel customer service immediately when they encounter problems during the self-service check-in process to ensure that the problem is solved in a timely manner;

[0146] The guest’s feedback will be recorded for subsequent service adjustments;

[0147] ;

[0148] In the formula, Indicates the guest's feedback record. Indicates the jth feedback information

[0149] In this embodiment, by generating a unique unlocking permission code Uk, it is ensured that only guests who have successfully paid can obtain access rights, thereby improving security. Guests can check in without a key through the mobile phone App, which simplifies the check-in process and improves the user experience. The real-time feedback and support unit provides an instant communication channel, so that guests can quickly contact the hotel customer service when they encounter problems to ensure that the problems are handled in a timely manner. The judgment formula for the validity of the permission can monitor the validity of the unlocking permission in real time, enhancing the intelligent management capabilities of the system. By limiting the number of unlocking code inputs, safety risks caused by misoperation are prevented, ensuring that the door cannot be opened without authorization. Recording guest feedback information provides a basis for subsequent service adjustments, which helps to improve the service quality of the hotel.

[0150] After payment is confirmed, a unique unlocking authorization code Uk is generated to ensure the security of each guest's identity and room information and avoid potential security loopholes. Customers unlock door locks through mobile phone apps, reducing the use of traditional keys and improving the convenience and modernity of check-in. The real-time feedback and support unit allows customers to contact the hotel in time when they encounter problems during self-service check-in, reducing their sense of uneasiness and enhancing their satisfaction. Limiting the number of unlocking code inputs prevents criminals from trying to unlock the room and ensures customer safety and privacy. By recording and analyzing customer feedback, hotels can continuously optimize service processes and improve the overall customer experience.

[0151] Example 5

[0152] A hotel self-service check-in intelligent registration management method, please refer to Figure 2 , specifically: including the following steps:

[0153] Step 1: The facial feature acquisition module uses a high-resolution camera to collect the guest's facial features and generate a total facial feature vector V Z , where the high-resolution cameras include a main camera and two sub-cameras, and the high-resolution cameras are arranged in a triangle;

[0154] Step 2: The facial feature matching module uses a convolutional neural network to obtain the total facial feature vector V Z Compare with the stored guest data in the hotel database and return the similarity score;

[0155] Step 3: The identity authentication module calls the identity information in the database based on the information obtained by the face feature matching module and converts the collected facial feature vector V Z Compare with the data on the Internet to further verify the guest's identity;

[0156] Step 4: The room allocation and payment confirmation module automatically allocates rooms based on the identity information after successful recognition, and prompts the guest to pay the room fee through mobile payment;

[0157] Step 5: Smart unlocking and check-in confirmation module When payment is successful, the unlocking permission will be sent to the guest's mobile phone, and the guest can unlock the door lock through the mobile app

[0158] In this embodiment, the facial features of the guest are collected by a high-resolution camera to generate a total facial feature vector V Z A main camera and two sub-cameras are used to capture the guest's facial images simultaneously in a triangle arrangement. The main camera captures the front image, while the two sub-cameras capture images from the left and right sides, ensuring that the acquired facial information is more comprehensive and reducing the loss of feature points caused by angles.

[0159] The total facial feature vector V Z Compare it with the guest data in the hotel database. Use convolutional neural network to Z Processing is performed to calculate the similarity score with each guest feature vector in the database, and the score is returned to evaluate the accuracy of the recognition result.

[0160] If the similarity score meets the set threshold, the system calls the internal database and the external network database for cross-comparison to confirm whether the guest's identity information is up to date and valid. Automatically allocate rooms and prompt payment. Based on the successfully identified identity information, automatically select the appropriate room and notify the guest to complete the room fee payment through the mobile payment platform. The system dynamically calculates the room availability to optimize the room allocation strategy. After the payment is successful, the system generates a unique unlocking authorization code and sends it to the guest's mobile app. The guest can enter the unlocking code or directly use the app to unlock the door lock and complete the check-in.

[0161] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A hotel self-service check-in intelligent registration management system, characterized by: It includes facial feature acquisition module, facial feature matching module, identity authentication module, room allocation and payment confirmation module, and smart unlocking and check-in confirmation module; The facial feature acquisition module uses a high-resolution camera to collect the guest's facial features and generate a total facial feature vector V Z , where the high-resolution cameras include a main camera and two sub-cameras, and the high-resolution cameras are arranged in a triangle; The facial feature matching module uses a convolutional neural network to obtain the total facial feature vector V Z Compare with the stored guest data in the hotel database and return the similarity score; The facial feature matching module includes a facial feature matching unit and a similarity scoring unit; The facial feature matching unit uses a convolutional neural network to collect the total facial feature vector V Z Compare it with the feature vector stored in the database to extract the total facial feature vector V Z The vector value F of the feature point Zi , extract the vector value F of the feature point in the feature vector stored in the database Si , calculate the difference ΔF between the two i ; The difference ΔF i Obtained by the following formula: ΔF i =|F Zi -F Si |; In the formula, F Zi Represents the total facial feature vector V Z The value of the i-th feature point in F Si Represents the value of the i-th feature point in the feature vector stored in the database; By calculating the difference of each feature point, we get the similarity score S M , used to determine whether the collected facial features best match the faces stored in the database; The similarity scoring unit calculates the difference ΔF of each feature point. i Perform weighted accumulation to calculate the final similarity score S M , to judge the guest’s identity. The lower the score, the more similar the collected facial features are to those in the database; The similarity score S M Obtained by the following formula: Where n represents the number of facial feature points, ω i represents the weight factor; When the similarity score S M <Threshold T M , it means the system has successfully identified the guest’s identity; When the similarity score S M >Threshold T M , it means that the system has not recognized the guest's identity. The system will traverse all the stored feature vectors in the database and recalculate the similarity score S M , until the similarity score S M <Threshold T M Until finally the match is successful and the guest's identity information is returned; The identity authentication module calls the identity information in the database based on the information obtained by the face feature matching module, and converts the collected facial total feature vector V Z Compare with the data on the Internet to further verify the guest's identity; The identity authentication module includes an identity information matching unit and an identity authentication probability calculation unit; The identity information matching unit calls the identity information in the database based on the face feature matching module and compares it with the collected total facial feature vector V Z Further comparison is performed, and the system cross-verifies the collected facial features with the identity data in the external network database to ensure that the guest's identity information is up-to-date and valid; When the similarity score S returned by the facial feature matching unit M <Threshold T M =0.85, the system calls the internal database and cross-checks the guest identity information in the external network database to ensure that the guest identity is up-to-date and correct; The identity verification probability calculation unit is based on the similarity score S M , calculate the probability of successful authentication P through the probability model S ,The identity authentication probability calculation unit uses an exponential decay function to calculate the probability of successful ,authentication and determine the guest’s identity; According to the similarity score S M , calculate the probability of successful verification P S ; The probability P S Obtained by the following formula: P S =1-e x (-k*S M ); In the formula, k represents the adjustment coefficient, e x represents the exponential function; When the similarity score S M The lower the probability of successful authentication, P S The closer to 1; When the similarity score S M The higher the probability of successful authentication, P S The closer to 0; The room allocation and payment confirmation module automatically allocates rooms according to the identity information after successful recognition, and prompts the guest to pay the room fee through mobile payment; The room allocation and payment confirmation module includes a room allocation unit and a payment confirmation unit; The room allocation unit allocates the room according to the number of remaining rooms N on that day. a and the total number of hotel rooms N n , calculate the room availability A R , based on room availability A R And preset room allocation rules to select the most suitable room for guests; The room availability rate A R Obtained by the following formula: By room availabilityA R , dynamically adjust room allocation strategies to ensure that hotel room resources are maximized and occupancy rates are increased; The payment confirmation unit prompts the guest to pay the room fee through the mobile payment platform according to the room assigned by the room assignment unit; the payment success rate P P Depending on the payment platform and the current network environment, the system will dynamically detect whether the payment is successful; Payment success rate P P Determined by the current payment environment, mainly depending on the availability of the payment platform and the network environment: P P =1-e x (-λ*t); In the formula, λ represents the stability coefficient of the payment platform, t represents the payment time, and e x represents the exponential function; Payment success rate P P Affected by network conditions and the stability of the payment platform, the longer the payment time t is, the higher the payment success rate P P will increase over time; The smart unlocking and check-in confirmation module sends the unlocking permission to the guest's mobile phone when the payment is successful, and the guest can unlock the door lock through the mobile phone App.

2. A hotel self-service check-in intelligent registration management system according to claim 1, characterized in that: The facial feature acquisition module includes a multi-angle camera acquisition unit and a feature vector fusion unit; The multi-angle camera acquisition unit captures the guest's facial image from different angles through a main camera and two auxiliary cameras to generate a facial feature vector V i ; Among them, the main camera captures the front image features and generates the front facial feature vector V M , the two secondary cameras capture the side features from the left and right angles respectively, and generate the left facial feature vector V L and the right facial feature vector V R ; For the facial feature vector V i It is expressed as: In the formula, Represents the N identified in the image f feature points, N f Indicates the number of feature points; The feature vector fusion unit extracts features from the images captured by the main camera and the auxiliary camera, and then uses the feature vector fusion algorithm to combine the frontal facial feature vector V M , left facial feature vector V L and the right facial feature vector V R Integrate into the total facial feature vector V Z ; The total facial feature vector V Z Obtained by the following formula: V Z =ω1*V M +ω2*V L +ω3*V R ; Wherein, ω1 represents the weight value of the main camera, ω2 and ω3 represent the weight values ​​of the two secondary cameras, and ω1+ω2+ω3=1, ω1>ω2=ω3.

3. A hotel self-service check-in intelligent registration management system according to claim 2, characterized in that: The smart unlocking and check-in confirmation module includes a permission generation unit, a mobile app unlocking unit and a real-time feedback and support unit; After the payment confirmation unit successfully pays, the permission generation unit generates a unique unlocking permission code Uk, and sends the unlocking permission code Uk and related information including the room number and the permission validity period T U , sent to the guest’s mobile app; The judgment formula of the authority validity of the authority generation unit is: Where, T t Indicates the difference between the current time and the permission generation time; When payment is successful, the management system of the smart unlocking and check-in confirmation module sends the unlocking permission to the guest's mobile phone, and the guest can unlock the door lock through the mobile phone App.

4. A hotel self-service check-in intelligent registration management system according to claim 3, characterized in that: The mobile app unlocking unit requires the guest to input and automatically send the unlocking authority code Uk through the mobile app for verification with the room door lock system; When the unlock code entered by the guest is the same as the unlock authority code Uk, the door will be opened; When the unlock code entered by the guest is different from the unlock authority code Uk, the door will not be opened. If the unlock code entered three times does not correspond to the unlock authority code Uk, the door will be locked and a unique unlock authority code Uk needs to be regenerated.

5. A hotel self-service check-in intelligent registration management system according to claim 4, characterized in that: The real-time feedback and support unit provides a real-time feedback channel, allowing guests to contact the hotel customer service immediately when they encounter problems during the self-service check-in process to ensure that the problem is solved in a timely manner; The guest’s feedback will be recorded for subsequent service adjustments; In the formula, F r Indicates the guest's feedback record, f j Represents the jth piece of feedback information.

6. A hotel self-service check-in intelligent registration management method, applied to a hotel self-service check-in intelligent registration management system according to any one of claims 1 to 5, characterized in that: The following steps are involved: Step 1: The facial feature acquisition module uses a high-resolution camera to collect the guest's facial features and generate a total facial feature vector V Z , where the high-resolution cameras include a main camera and two sub-cameras, and the high-resolution cameras are arranged in a triangle; Step 2: The facial feature matching module uses a convolutional neural network to obtain the total facial feature vector V Z Compare with the stored guest data in the hotel database and return the similarity score; Step 3: The identity authentication module calls the identity information in the database based on the information obtained by the face feature matching module and converts the collected facial feature vector V Z Compare with the data on the Internet to further verify the guest's identity; Step 4: The room allocation and payment confirmation module automatically allocates rooms based on the identity information after successful recognition, and prompts the guest to pay the room fee through mobile payment; Step 5: Smart unlocking and check-in confirmation module When the payment is successful, the unlocking permission will be sent to the guest’s mobile phone, and the guest can unlock the door lock through the mobile app.

Images