Elevator intelligent early warning method and system for special personnel

Abstract

The invention discloses an intelligent elevator early warning method and system for special personnel, and relates to the technical field of intelligent monitoring and safety protection. Related check-in personnel storage characteristic data information is obtained according to personnel data and information provided when a resident handles a check-in procedure; related passenger recognition feature data information is obtained, a cosine similarity algorithm is combined, the face similarity coefficient Xxs of a plurality of elevator passengers is calculated, after similarity comparison, recognized resident passenger crowds are constructed, and a child independent driving analysis instruction is sent out; and then, the personnel information of the identified resident passenger crowd is retrieved and matched, a set matching algorithm is utilized to obtain a single-walking child proportion coefficient Zcp, after comparison, whether children taking the elevator alone exist in the identified resident passenger crowd in the current elevator or not is judged so as to generate corresponding-level danger early warning, and corresponding early warning reminding is executed.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent monitoring and safety protection, and in particular to an elevator intelligent early warning method and system for special personnel. Background Art

[0002] With the rapid development of information technology and intelligent technology, intelligent monitoring and security protection have gradually become important means to ensure public safety. With the advancement of technology, the application of artificial intelligence, big data and Internet of Things technologies has made security monitoring gradually shift to intelligence, real-time and efficiency. The application of these technologies not only improves monitoring efficiency, but also can more accurately identify potential safety risks. Under this trend, elevators, as an indispensable part of urban construction, have also gradually introduced intelligent management systems, especially monitoring of passenger safety in elevators, which has improved the safety and reliability of elevators during use.

[0003] Although the current elevator management system has achieved basic passenger identification and management, there are still some difficult-to-solve safety hazards when dealing with special groups taking the elevator. First, most of the existing elevator management systems rely on traditional card swiping, password and manual input methods to verify the identity of passengers, lacking in-depth analysis and dynamic identification of passenger identities. In addition, the existing system does not fully consider the risks of special groups in the use of elevators, especially for children under 6 years old taking the elevator. It fails to effectively identify whether the children are accompanied by family members, and it is difficult to issue effective warnings in time for children taking the elevator alone, which can easily lead to potential elevator safety accidents. Therefore, there is an urgent need for an intelligent, real-time monitoring elevator intelligent warning system that can identify and detect children taking the elevator alone, immediately issue a warning and take corresponding emergency measures to ensure the safety of passengers. Summary of the invention

[0004] In view of the deficiencies in the prior art, the present invention provides an elevator intelligent early warning method and system for special personnel, which solves the problems in the above-mentioned background technology.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an elevator intelligent early warning method for special personnel, comprising the following steps:

[0006] S1. Based on the ID information and facial information provided by the occupants when checking in, the OCR text recognition technology and FaceNet algorithm are used to construct the storage feature data information of the relevant occupants, and multiple sets of cameras are arranged inside the elevator, combined with wireless network communication technology, to obtain the identification feature data information of the relevant passengers;

[0007] S2. Extract the stored characteristic data information of the relevant occupants and the identification characteristic data information of the relevant passengers, and combine the cosine similarity algorithm to obtain the face similarity coefficient Xxs of several elevator passengers. After the similarity comparison, construct the identified resident passenger group, and issue a child lone travel analysis instruction;

[0008] S3. After receiving the instruction for analyzing the children traveling alone, the personnel information of the identified resident passenger group is searched and matched to obtain the data information of the identified resident passenger, and the coefficient of the proportion of children traveling alone Zcp is obtained by using the set matching algorithm;

[0009] S4. According to the value of the coefficient of proportion of children traveling alone Zcp, it is determined whether there are children riding the elevator alone among the identified resident passengers in the current elevator, so as to generate a corresponding level of danger warning and execute a corresponding warning reminder.

[0010] Preferably, the specific steps of S1 include:

[0011] S11. First, when the resident goes through the check-in procedure at the property center, the identity information of the resident is obtained by using the OCR text recognition technology based on the ID card, household registration booklet and property certificate provided by the resident. The facial information of the resident is collected through the face recognition camera and the 3D structured light camera, and the facial feature information of the resident is obtained by combining the FaceNet algorithm. The identity information of the resident includes the name, age Gnr and mobile phone number of each resident; the facial feature information of the resident includes the target facial feature vector Fmb of each resident.

[0012] S12. Perform feature recognition on the facial feature information and identity information of the occupants, and combine with AES-256 encryption technology to construct relevant occupant storage feature data information, and store the encrypted relevant occupant storage feature data information in the cloud database of the elevator management system. When a person takes the elevator, the relevant data stored in the cloud database is called for identity authentication; wherein, the relevant occupant storage feature data information includes the age Gnr of each occupant, the target face feature vector Fmb and the family ID number Bid.

[0013] Preferably, the specific step S1 also includes:

[0014] S13. When people take the elevator, based on the need for identity recognition of the elevator passengers, multiple groups of cameras are deployed inside the elevator, and the multiple groups of cameras are wirelessly connected to the elevator management system in combination with wireless network communication technology; and the identity characteristics of the elevator passengers are recognized based on the multiple groups of cameras, and relevant passenger recognition feature data information is obtained; wherein the multiple groups of cameras include RGB cameras, infrared cameras and depth cameras, and the relevant passenger recognition feature data information includes the passenger face feature vector Fcz of each passenger.

[0015] Preferably, the specific steps of S2 include:

[0016] S21, extracting features from the storage feature data information of the relevant occupants and the identification feature data information of the relevant passengers stored in the cloud database, and analyzing the similarity between the facial features of the passengers and the occupants in combination with the cosine similarity algorithm to construct a facial similarity coefficient Xxs, which is specifically obtained in the following manner:

[0017] ;

[0018] Where Fcz represents the passenger face feature vector of the corresponding passenger, and Fmb represents the target face feature vector of the occupants in the cloud database. It is represented by the length of the facial feature vector of the corresponding passenger. Represented as the length of the target facial feature vector of the occupants in the cloud database.

[0019] Preferably, the specific step S2 also includes:

[0020] S22. According to the calculation method of step S21, the facial similarity coefficients Xxs of several elevator passengers are obtained respectively, and a similarity threshold is pre-set. By comparing the facial similarity coefficients Xxs of several elevator passengers with the similarity threshold, it is identified whether the elevator passenger is a resident. If the facial similarity coefficient Xxs does not exceed the similarity threshold, it means that the current elevator passenger is not a resident. At this time, the corresponding elevator passenger is included in the temporary passenger group after identification. If the facial similarity coefficient Xxs exceeds the similarity threshold, it means that the current elevator passenger is a resident and the person is marked. At this time, the marked elevator passenger is included in the resident passenger group after identification, and a child walking alone analysis instruction is issued.

[0021] Preferably, the specific steps of S3 include:

[0022] S31. After receiving the child travel alone analysis instruction, use the cloud database indexing technology to search and match the personnel information of the identified resident passenger group to obtain the identified resident passenger data information, which includes the age Gnr and family ID number Bid of each identified resident passenger, wherein the persons of the same family have the same family ID number Bid;

[0023] S32, extracting features from the identified resident passenger data information obtained in step S31, and obtaining the number of children Nr and the number of adults Nc according to the age Gnr value of each identified resident passenger;

[0024] ;

[0025] In the formula, It is represented by the age of the i-th identified resident passenger, i=1, 2, 3, ..., N, N is the total number of identified resident passengers, where, represents the indicator function, when the age of the resident passenger after the i-th identification If it is less than 6, it is a child and returns 1, otherwise it returns 0; represents the indicator function, when the age of the resident passenger after the i-th identification If the person is over 18, the result is 1, otherwise it returns 0.

[0026] Preferably, the specific step S3 also includes:

[0027] S33, according to the number of children Nr and the number of adults Nc obtained in step S32, and in combination with the identified household passenger data information, a set matching algorithm is used to obtain the single child ratio coefficient Zcp, which is specifically obtained by the following formula:

[0028] ;

[0029] In the formula, It is the family ID number of the r-th child, r=1, 2, 3, ..., Nr, Nr is the number of children, It is represented by the family ID number of the jth adult, j=1, 2, 3, ..., Nc, Nc is the number of adults, where represents an indicator function. When the family ID number of the rth child is the same as the family ID number of the jth adult, it means that the two belong to the same family and returns 1, otherwise it returns 0; Indicates the proportion of all children who ride the elevator alone.

[0030] Preferably, the specific steps of S4 include:

[0031] S41. According to the value of the coefficient of proportion of children traveling alone Zcp, it is determined whether there are children traveling alone in the current elevator among the identified resident passengers, so as to generate a corresponding level of danger warning. The specific contents are as follows:

[0032] If the coefficient of the proportion of children traveling alone Zcp>0, it means that there are children traveling alone in the elevator among the identified residents and passengers in the elevator, indicating that there is a safety risk that the children in the elevator are not accompanied by adults. At this time, a level 1 danger warning is issued.

[0033] If the coefficient of the proportion of children traveling alone Zcp=0, it means that there are no children riding alone in the elevator among the residents and passengers identified in the current elevator, which means that there is no safety risk for children in the current elevator without adult accompaniment. At this time, a second-level danger warning is issued.

[0034] Preferably, the specific step S4 also includes:

[0035] S42: According to the first-level danger warning and the second-level danger warning issued in step S41, corresponding warning reminders are executed. The specific execution contents are as follows:

[0036] When receiving a level 1 danger warning, the warning reminder will be: an emergency voice prompt will be played in the elevator, "Warning! There is a child alone in the elevator. Please arrange for an adult to accompany the child as soon as possible. If there is an emergency, please contact the property management center and call the emergency contact number immediately." The elevator management system will send the warning information to the parents and the property staff of the building where the elevator is located via SMS and phone. If the child fails to be accompanied by an adult immediately after the warning is issued, the elevator management system will initiate the emergency response process and further track the child through the surveillance camera inside the elevator;

[0037] When receiving a Level 2 danger warning, the following warning reminder will be executed: a voice prompt will be played in the elevator: "All children in the elevator are currently accompanied by adults. If you need anything, please contact the property center." At the same time, a reminder message will be sent to the property staff to inform them that there are no children riding alone in the elevator, and the surveillance cameras inside the elevator will continue to monitor whether there are children riding alone among the new entrants in the elevator.

[0038] An intelligent elevator warning system for special personnel, including a personnel collection module, a similarity analysis module, a child walking alone analysis module and a level warning module;

[0039] The personnel collection module is used to construct the storage feature data information of the relevant occupants based on the certificate information and facial information provided by the occupants when checking in, using OCR text recognition technology and FaceNet algorithm, and to obtain the identification feature data information of the relevant passengers by deploying multiple groups of cameras inside the elevator and combining wireless network communication technology;

[0040] The similarity analysis module is used to extract the feature data information stored by the relevant occupants and the feature data information of the relevant passengers, and combine the cosine similarity algorithm to obtain the face similarity coefficient Xxs of several elevator passengers. After the similarity comparison, the identified occupant passenger group is constructed, and a child lone travel analysis instruction is issued to the outside;

[0041] The single-child analysis module is used to retrieve and match the personnel information of the identified resident passenger group after receiving the single-child analysis instruction, obtain the identified resident passenger data information, and use the set matching algorithm to obtain the single-child ratio coefficient Zcp;

[0042] The level warning module is used to determine whether there are children riding alone in the elevator among the identified residents and passengers in the current elevator according to the numerical value of the lone children proportion coefficient Zcp, so as to generate the corresponding level of danger warning and execute the corresponding warning reminder.

[0043] The present invention provides an elevator intelligent early warning method and system for special personnel, which has the following beneficial effects:

[0044] (1) By innovatively combining OCR text recognition technology, FaceNet algorithm, face recognition technology, cosine similarity algorithm and cloud database technology, the safety management level of special personnel in elevators is improved, especially in identifying and preventing the risks of children riding alone in elevators. Compared with traditional elevator management systems, it can identify the identities of passengers in elevators in real time and accurately determine whether there are safety hazards of children riding alone in elevators. By efficiently extracting facial feature information of occupants, combining wireless network and multiple groups of cameras inside the elevator to authenticate passengers in real time, the in-depth analysis and dynamic identification of passengers in the elevator are ensured. By calculating and analyzing the coefficient of the proportion of single children Zcp, it is determined whether there are children riding alone in the current resident passenger crowd in the elevator after identification, and the corresponding level of danger warning is automatically generated, and voice prompts, SMS notifications and emergency response instructions are issued according to the risk level, thereby effectively avoiding the loopholes in traditional elevator management systems that are difficult to timely identify the problem of children riding alone in elevators. This fully automated intelligent elevator management system improves the safety of elevators and the protection capabilities of special personnel, and can provide more accurate and reliable elevator safety guarantees for property management and residents.

[0045] (2) Accurate identity recognition and feature extraction are effectively achieved by using a variety of technical means. Through the automatic recognition of the ID documents of the occupants by OCR text recognition technology and the collection and comparison of facial features by the FaceNet algorithm, the identity information and facial data of each occupant can be efficiently stored and managed. At the same time, the multi-dimensional information collection method combined with the RGB camera, infrared camera and depth camera can achieve more accurate identity recognition and real-time monitoring, especially when the flow of people in the elevator is relatively complex. This feature effectively improves the recognition accuracy of the elevator management system and divides the group of temporary passengers after recognition and the group of resident passengers after recognition, so as to issue a child walking alone analysis instruction, providing the prerequisite and data support for the subsequent child walking alone analysis. At the same time, this data collection method effectively makes up for the fact that most existing elevator management systems rely on traditional card swiping, passwords and manual input methods to verify the identity of passengers, lacking in-depth analysis and dynamic recognition of passenger identities, and effectively avoiding safety hazards caused by human errors or time delays.

[0046] (3) The innovative design of the lone child analysis and danger warning module is a highlight. By setting up the lone child ratio coefficient Zcp, it is possible to dynamically analyze and accurately calculate whether there are children riding alone in the elevator based on the personnel information of the identified resident passenger group. When the lone child ratio coefficient Zcp>0, it can quickly trigger a level one danger warning, and promptly notify relevant personnel (such as parents, property personnel) through voice prompts and SMS notifications, thereby providing the first-time protection for the child's safety. Compared with the traditional elevator management system, which is difficult to automatically detect whether a child is accompanied by an adult, this method can effectively warn at an early stage and avoid potential dangers. In addition, different levels of danger warnings are set according to the lone child ratio coefficient Zcp, and the response level can be flexibly adjusted according to actual conditions. For example, when the lone child ratio coefficient Zcp=0, only a level two danger warning is issued, and the surveillance camera inside the elevator continues to monitor whether there are children riding alone among the new entrants in the elevator, and further observes the dynamics of the people in the elevator. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 This is a flow chart of an intelligent early warning method for elevators for special personnel according to the present invention;

[0048] Figure 2 This is a block diagram of an elevator intelligent early warning system for special personnel according to the present invention;

[0049] Figure 3 This is a logical thinking diagram of an elevator intelligent early warning method for special personnel according to the present invention;

[0050] Figure 4The present invention is a personnel relationship Venn diagram of an elevator intelligent early warning method for special personnel. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] Example 1, please refer to Figure 1 , Figure 3 and Figure 4 The present invention provides an elevator intelligent early warning method for special personnel, comprising the following steps:

[0053] S1. Based on the ID information and facial information provided by the occupants when checking in, the OCR text recognition technology and FaceNet algorithm are used to construct the storage feature data information of the relevant occupants, and multiple sets of cameras are arranged inside the elevator, combined with wireless network communication technology, to obtain the identification feature data information of the relevant passengers;

[0054] S2. Extract the stored characteristic data information of the relevant occupants and the identification characteristic data information of the relevant passengers, and combine the cosine similarity algorithm to obtain the face similarity coefficient Xxs of several elevator passengers. After the similarity comparison, construct the identified resident passenger group, and issue a child lone travel analysis instruction;

[0055] S3. After receiving the instruction for analyzing the children traveling alone, the personnel information of the identified resident passenger group is searched and matched to obtain the data information of the identified resident passenger, and the coefficient of the proportion of children traveling alone Zcp is obtained by using the set matching algorithm;

[0056] S4. According to the value of the coefficient of proportion of children traveling alone Zcp, it is determined whether there are children riding the elevator alone among the identified resident passengers in the current elevator, so as to generate a corresponding level of danger warning and execute a corresponding warning reminder.

[0057] In this embodiment, through the combination of OCR text recognition technology and FaceNet algorithm with wireless network communication technology, the identity information of passengers in the elevator can be accurately identified, especially the timely warning of children walking alone; this method effectively solves the deficiency of the existing elevator management system that fails to timely identify and warn children walking alone when people take the elevator, and improves the safety of elevator use; through feature extraction and similarity comparison, it can effectively distinguish the identities of different passengers and accurately identify children, and then calculate the coefficient Zcp of the proportion of children walking alone, so as to generate different levels of danger warnings; this technical means not only improves the intelligence level of the elevator, but also can trigger emergency response measures such as voice prompts and SMS notifications at critical moments, ensure the safety of children taking the elevator, and reduce safety hazards; the implementation of this method enhances the real-time monitoring capability of the elevator management system for special groups taking the elevator, meets the increasing demand for elevator safety in modern cities, and effectively solves the deficiencies of traditional elevator management systems in dynamic recognition and immediate response.

[0058] Example 2, please refer to Figure 1 , specifically: S1 specific steps include:

[0059] S11. First, when the resident goes through the check-in procedure at the property center, the identity information of the resident is obtained by using the OCR text recognition technology based on the ID card, household registration booklet and property certificate provided by the resident. The facial information of the resident is collected through the face recognition camera and the 3D structured light camera, and the facial feature information of the resident is obtained by combining the FaceNet algorithm. The identity information of the resident includes the name, age Gnr and mobile phone number of each resident; the facial feature information of the resident includes the target facial feature vector Fmb of each resident.

[0060] It should be noted that OCR text recognition technology is a technology used to convert text information in an image into editable and searchable text data; by analyzing the shape, structure and pattern of the text in the image, the text information of the scanned or photographed documents, identity cards, household registration books, and real estate certificates is extracted and converted into a format that can be understood by the computer; in the present invention, OCR text recognition technology is used to extract name, age GNR and mobile phone number information from the identity card, household registration book and real estate certificate documents provided by the occupants as basic data for identity verification of the occupants; when the occupants check in, the occupants' faces are scanned and photographed by a 3D structured light camera and a face recognition camera to obtain a high-definition facial image; then, using the FaceNet deep learning algorithm specifically for facial recognition and verification, a convolutional neural network CNN is used to extract facial features, and by learning a large-scale facial data set, the facial image of each occupant can be converted into a 128-dimensional feature vector, thereby obtaining a 128-dimensional target face feature vector Fmb; this target face feature vector Fmb is a unique facial identity identifier for each occupant, which can accurately reflect the facial features of the person, and has uniqueness and high discrimination.

[0061] S12. Perform feature recognition on the facial feature information and identity information of the occupants, and combine with AES-256 encryption technology to construct relevant occupant storage feature data information, and store the encrypted relevant occupant storage feature data information in the cloud database of the elevator management system. When a person takes the elevator, the relevant data stored in the cloud database is called for identity authentication; wherein, the relevant occupant storage feature data information includes the age Gnr of each occupant, the target face feature vector Fmb and the family ID number Bid.

[0062] Specifically, the specific steps of S1 also include:

[0063] S13. When people take the elevator, based on the need for identity recognition of the elevator passengers, multiple groups of cameras are deployed inside the elevator, and the multiple groups of cameras are wirelessly connected to the elevator management system in combination with wireless network communication technology; and the identity characteristics of the elevator passengers are recognized based on the multiple groups of cameras, and relevant passenger recognition feature data information is obtained; wherein the multiple groups of cameras include RGB cameras, infrared cameras and depth cameras, and the relevant passenger recognition feature data information includes the passenger face feature vector Fcz of each passenger.

[0064] It should be noted that the passenger facial feature vector Fcz is obtained by collecting facial image data of passengers in the elevator through multiple groups of cameras (including RGB cameras, infrared cameras and depth cameras), and processing the images using the FaceNet algorithm; these cameras provide facial image information from multiple angles and under different lighting conditions. By converting these images into high-dimensional facial feature vectors Fcz, the unique facial features of the passengers can be accurately captured and represented; the role of the passenger facial feature vector Fcz is to provide each passenger with a unique digital identity, which can be compared with the resident facial feature vectors stored in the cloud database to help identify the passenger in real time.

[0065] In this embodiment, the present invention realizes highly intelligent identification of elevator passengers by collecting and processing multi-dimensional identity information of occupants when checking in at the property center. First, OCR text recognition technology is used to accurately obtain the identity card, household registration booklet and real estate certificate information of the occupants, and then the face recognition camera and 3D structured light camera are used to collect the faces of the occupants, and the facial feature information of each resident is obtained in combination with the FaceNet algorithm. This method not only ensures the accuracy of the identity information of each resident during collection, but also establishes an independent, encrypted and traceable occupant data model by combining facial feature vectors and identity information. The data in this process is securely stored using AES-256 encryption technology, which improves the security and privacy protection of the data. The encrypted data stored in the cloud database will provide continuous identity authentication support for the elevator system. Multiple groups of cameras are deployed inside the elevator, including RGB cameras, infrared cameras and depth cameras. The camera is connected to the elevator management system through a wireless network to achieve real-time and dynamic identification of personal identity features. This combination of multiple cameras not only improves the adaptability of the elevator environment to changes, but also ensures that the facial feature information of passengers can be efficiently captured under different lighting, angles and environmental conditions. The identification feature data information of elevator passengers is extracted in a timely manner and compared with the occupant data stored in the cloud database to ensure that the identity of each passenger is accurately verified. This method combines a variety of sensor technologies, wireless communication and encryption storage technologies to ensure that the identification of personnel during the elevator ride is safer, smarter and real-time, significantly reducing potential safety risks, especially when dealing with special personnel riding the elevator, it can quickly identify and respond, and improve the safety protection capabilities of the elevator. This method not only improves the intelligence level of the elevator system, but also provides passengers with a safer riding environment, and effectively addresses the shortcomings of traditional elevator management systems in identity recognition accuracy, real-time and safety.

[0066] Example 3, please refer to Figure 1 , specifically: S2 specific steps include:

[0067] S21, extracting features from the storage feature data information of the relevant occupants and the identification feature data information of the relevant passengers stored in the cloud database, and analyzing the similarity between the facial features of the passengers and the occupants in combination with the cosine similarity algorithm to construct a facial similarity coefficient Xxs, which is specifically obtained in the following manner:

[0068] ;

[0069] Where Fcz represents the passenger face feature vector of the corresponding passenger, and Fmb represents the target face feature vector of the occupants in the cloud database. It is represented by the length of the facial feature vector of the corresponding passenger. Represented as the length of the target facial feature vector of the occupants in the cloud database.

[0070] It should be noted that the cosine similarity algorithm is a mathematical method for calculating the similarity of two vectors, and the similarity is judged by measuring the cosine value of the angle between the two vectors; the value range of the cosine similarity is between -1 and 1, where 1 indicates that the two vectors are exactly the same, 0 indicates that the two vectors are orthogonal, and -1 indicates that they are completely opposite; in the present invention, the cosine similarity algorithm is used to calculate the similarity between the passenger face feature vector Fcz of the corresponding passenger and the target face feature vector Fmb of the occupant in the cloud database; by calculating the cosine similarity, a face similarity coefficient Xxs is generated, which reflects the similarity of the facial features of the two; the function is to determine whether the passenger in the elevator matches the facial features of a certain occupant through the face similarity coefficient Xxs, so as to perform identity authentication; if the similarity exceeds the set threshold, it can be determined that the passenger is a legal resident, and further analysis can be performed, such as determining whether a child is riding the elevator alone; this method effectively solves the shortcomings of identity authentication in traditional elevator management systems, provides a more accurate and real-time passenger identification function, and ensures the safety of passengers.

[0071] Specifically, the specific steps of S2 also include:

[0072] S22. According to the calculation method of step S21, the facial similarity coefficients Xxs of several elevator passengers are obtained respectively, and a similarity threshold is pre-set. By comparing the facial similarity coefficients Xxs of several elevator passengers with the similarity threshold, it is identified whether the elevator passenger is a resident. If the facial similarity coefficient Xxs does not exceed the similarity threshold, it means that the current elevator passenger is not a resident. At this time, the corresponding elevator passenger is included in the temporary passenger group after identification. If the facial similarity coefficient Xxs exceeds the similarity threshold, it means that the current elevator passenger is a resident and the person is marked. At this time, the marked elevator passenger is included in the resident passenger group after identification, and a child walking alone analysis instruction is issued.

[0073] In this embodiment, the present invention introduces accurate face similarity analysis technology in the elevator passenger identification. By combining the storage feature data information of relevant occupants and the identification feature data information of relevant passengers stored in the cloud database, an efficient and accurate passenger identity authentication system is constructed; the identity features of passengers in the elevator are compared with those of residents through the cosine similarity algorithm, and the face similarity coefficient Xxs is calculated; this process can accurately evaluate the similarity by comparing the passenger face feature vector Fcz of the corresponding passenger with the target face feature vector Fmb of the occupants in the cloud database; according to the calculation results, it can be judged whether the current elevator passenger is a resident, and the passenger identity can be compared according to the set similarity threshold. The face recognition algorithm is used to classify the identities of passengers in the elevator; when the face similarity coefficient Xxs does not exceed the threshold, the passenger is identified as a temporary passenger and corresponding processing is performed; if the face similarity coefficient Xxs exceeds the threshold, the passenger is confirmed as a resident, further marked, and the child lone travel analysis instruction is triggered; the combination of face recognition and similarity algorithm not only effectively improves the accuracy of passenger identity recognition in the elevator, but also can realize fast and real-time identity authentication, effectively avoiding the uncertainty and delay problems of traditional manual judgment; through step S2, the present invention improves the intelligence and safety of the elevator management system, improves the user experience, especially when the flow of people in the elevator is relatively complex, it can provide more efficient and accurate identity recognition.

[0074] Example 4, please refer to Figure 1 , specifically: S3 specific steps include:

[0075] S31. After receiving the child travel alone analysis instruction, use the cloud database indexing technology to search and match the personnel information of the identified resident passenger group to obtain the identified resident passenger data information, which includes the age Gnr and family ID number Bid of each identified resident passenger, wherein the persons of the same family have the same family ID number Bid;

[0076] S32, extracting features from the identified resident passenger data information obtained in step S31, and obtaining the number of children Nr and the number of adults Nc according to the age Gnr value of each identified resident passenger;

[0077] ;

[0078] In the formula, It is represented by the age of the i-th identified resident passenger, i=1, 2, 3, ..., N, N is the total number of identified resident passengers, where, represents the indicator function, when the age of the resident passenger after the i-th identification If it is less than 6, it is a child and returns 1, otherwise it returns 0; represents the indicator function, when the age of the resident passenger after the i-th identification If the person is over 18, the result is 1, otherwise it returns 0.

[0079] Specifically, the S3 specific steps also include:

[0080] S33, according to the number of children Nr and the number of adults Nc obtained in step S32, and in combination with the identified household passenger data information, a set matching algorithm is used to obtain the single child ratio coefficient Zcp, which is specifically obtained by the following formula:

[0081] ;

[0082] In the formula, It is the family ID number of the r-th child, r=1, 2, 3, ..., Nr, Nr is the number of children, It is represented by the family ID number of the jth adult, j=1, 2, 3, ..., Nc, Nc is the number of adults, where represents an indicator function. When the family ID number of the rth child is the same as the family ID number of the jth adult, it means that the two belong to the same family and returns 1, otherwise it returns 0; Indicates the percentage of children who ride elevators alone among all children. for , indicating that the current child is not accompanied by a family member (that is, the current child’s family ID is different from the family ID of any adult in the elevator), returns 1, otherwise it returns 0.

[0083] It should be noted that the coefficient of the proportion of children traveling alone Zcp is an indicator used to measure the proportion of children riding the elevator alone; the number of children Nr and the number of adults Nc of all passengers in the elevator are analyzed in combination with the set matching algorithm to calculate how many children are not accompanied by adults among all children; specifically, by comparing the family ID number of each child with the family ID number of the adults in the elevator, it is determined whether each child rides the elevator with adults from the same family; if the family ID numbers are the same, it means that the child is accompanied by an adult, otherwise it means that the child is riding the elevator alone. Finally, by calculating the proportion of children who are not accompanied by adults among all children, the coefficient of the proportion of children traveling alone Zcp is obtained. The value of the coefficient of the proportion of children traveling alone Zcp is between 0 and 1.

[0084] Specific example: Assume that there are 5 residents and passengers in the elevator after identification;

[0085] Step 1: Retrieve the resident passenger information from the database through cloud database indexing technology and obtain the following data:

[0086] Passenger 1 (adult, age 30, family ID number Bid=A);

[0087] Passenger 2 (adult, age 35, family ID number Bid=A);

[0088] Passenger 3 (child, age 5, family ID number Bid=B);

[0089] Passenger 4 (child, age 4, family ID number Bid=C);

[0090] Passenger 5 (adult, age 40, family ID number Bid=C);

[0091] Step 2: Extract features from passenger information and classify passengers according to their age Gnr:

[0092] Children: passengers under the age of 6;

[0093] Adult: Passengers aged 18 or above;

[0094] Results extracted based on age:

[0095] Number of children Nr = 2 (passenger 3 and passenger 4);

[0096] Number of adults Nc = 3 (passenger 1, passenger 2, passenger 5);

[0097] Step 3: Use the set matching algorithm to calculate the single child ratio coefficient Zcp. The family IDs of child passengers 3 and 4 are B and C respectively, and there are no adult passengers who match these two family IDs:

[0098] According to the calculation formula, the coefficient of the proportion of children traveling alone, Zcp, is 1;

[0099] In this embodiment, the present invention further optimizes the detection and warning mechanism of children walking alone in the intelligent early warning system of the elevator, especially when processing the information of residents and passengers, an efficient cloud database index and set matching algorithm is adopted, so that the judgment of the situation of children walking alone in the elevator is more accurate and reliable; first, the personnel information of the identified resident passenger group is retrieved and matched by using the cloud database indexing technology to obtain the data information of the identified resident passengers, and the number of children Nr and the number of adults Nc of the identified resident passengers are obtained through age screening to distinguish children from adults; in this way, the recognition accuracy of different age groups is improved; then, through independent family ID matching and calculation of the proportion of children, it is further determined whether there are children riding the elevator alone, and converted into the coefficient of proportion of children walking alone Zcp; according to the coefficient of proportion of children walking alone Zcp; this method not only improves the accuracy of monitoring children walking alone, but also enhances the adaptability to the dynamics of family passengers, effectively solving the problem of difficulty in timely identification of children walking alone in the elevator in traditional elevator monitoring, thereby providing more personalized and refined safety protection means to ensure the safety of children riding the elevator.

[0100] Example 5, please refer to Figure 1 , specifically: S4 specific steps include:

[0101] S41. According to the value of the coefficient of proportion of children traveling alone Zcp, it is determined whether there are children traveling alone in the current elevator among the identified resident passengers, so as to generate a corresponding level of danger warning. The specific contents are as follows:

[0102] If the coefficient of the proportion of children traveling alone Zcp>0, it means that there are children traveling alone in the elevator among the identified residents and passengers in the elevator, indicating that there is a safety risk that the children in the elevator are not accompanied by adults. At this time, a level 1 danger warning is issued.

[0103] If the coefficient of the proportion of children traveling alone Zcp=0, it means that there are no children riding alone in the elevator among the residents and passengers identified in the current elevator, which means that there is no safety risk for children in the current elevator without adult accompaniment. At this time, a second-level danger warning is issued.

[0104] Specifically, the specific steps of S4 also include:

[0105] S42: According to the first-level danger warning and the second-level danger warning issued in step S41, corresponding warning reminders are executed. The specific execution contents are as follows:

[0106] When receiving a level 1 danger warning, the warning reminder will be: an emergency voice prompt will be played in the elevator, "Warning! There is a child alone in the elevator. Please arrange for an adult to accompany the child as soon as possible. If there is an emergency, please contact the property management center and call the emergency contact number immediately." The elevator management system will send the warning information to the parents and the property staff of the building where the elevator is located via SMS and phone. If the child fails to be accompanied by an adult immediately after the warning is issued, the elevator management system will initiate the emergency response process and further track the child through the surveillance camera inside the elevator;

[0107] When receiving a Level 2 danger warning, the following warning reminder will be executed: a voice prompt will be played in the elevator: "All children in the elevator are currently accompanied by adults. If you need anything, please contact the property center." At the same time, a reminder message will be sent to the property staff to inform them that there are no children riding alone in the elevator, and the surveillance cameras inside the elevator will continue to monitor whether there are children riding alone among the new entrants in the elevator.

[0108] In this embodiment, through real-time monitoring and analysis of the coefficient of proportion of children walking alone Zcp in step S4, the protection measures for the safety of children in the elevator are significantly improved, especially with the cooperation of multiple danger warning mechanisms, which ensures the rapid response and handling of children riding alone in the elevator; through the judgment of the coefficient of proportion of children walking alone Zcp, it is possible to accurately identify whether there are children without adult accompaniment in the elevator, and issue corresponding danger warnings according to the degree of risk; if the coefficient of proportion of children walking alone Zcp is greater than 0, a first-level danger warning will be issued immediately, indicating that there are children walking alone in the current elevator, and an emergency voice reminder will be activated immediately, and parents and property personnel will be notified through text messages, phone calls and other methods to ensure that the relevant responsible parties can intervene in time and take necessary measures; at the same time, an emergency response process is also configured to further ensure that when the danger is not dealt with in time When the alarm is released, the elevator management system can continue to track through surveillance cameras to ensure the real-time safety of children; for the case where the coefficient of the proportion of children traveling alone Zcp is equal to 0, a second-level danger warning is used to confirm that all children have been accompanied by adults, and the changes in personnel in the elevator continue to be monitored, and whether there are children riding the elevator alone among the new entrants in the elevator; this dual warning mechanism combines voice prompts, information sending and monitoring tracking, and has efficient and comprehensive safety protection capabilities. It not only improves the accuracy of intelligent elevator management, but also can respond to emergencies in real time, reducing the safety hazards caused by children riding the elevator alone; through this comprehensive protection measure, a more intelligent and safe management solution is provided for the use of elevators in residential areas, ensuring the safety of children riding the elevator alone, and effectively avoiding the traditional elevator management system's neglect of children riding the elevator alone.

[0109] Example 6, please refer to Figure 1 and Figure 2, an elevator intelligent early warning system for special personnel, including a personnel collection module, a similarity analysis module, a child walking alone analysis module and a level early warning module;

[0110] The personnel collection module is used to construct the storage feature data information of the relevant occupants based on the certificate information and facial information provided by the occupants when checking in, using OCR text recognition technology and FaceNet algorithm, and to obtain the identification feature data information of the relevant passengers by deploying multiple groups of cameras inside the elevator and combining wireless network communication technology;

[0111] The similarity analysis module is used to extract the feature data information stored by the relevant occupants and the feature data information of the relevant passengers, and combine the cosine similarity algorithm to obtain the face similarity coefficient Xxs of several elevator passengers. After the similarity comparison, the identified occupant passenger group is constructed, and a child lone travel analysis instruction is issued to the outside;

[0112] The single-child analysis module is used to retrieve and match the personnel information of the identified resident passenger group after receiving the single-child analysis instruction, obtain the identified resident passenger data information, and use the set matching algorithm to obtain the single-child ratio coefficient Zcp;

[0113] The level warning module is used to determine whether there are children riding alone in the elevator among the identified residents and passengers in the current elevator according to the numerical value of the lone children proportion coefficient Zcp, so as to generate the corresponding level of danger warning and execute the corresponding warning reminder.

[0114] 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. An intelligent elevator early warning method for special personnel, characterized in that: The following steps are involved: S1. Based on the ID information and facial information provided by the occupants when checking in, the OCR text recognition technology and FaceNet algorithm are used to construct the storage feature data information of the relevant occupants, and multiple sets of cameras are arranged inside the elevator, combined with wireless network communication technology, to obtain the identification feature data information of the relevant passengers; S2. Extract the stored characteristic data information of the relevant occupants and the identification characteristic data information of the relevant passengers, and combine the cosine similarity algorithm to obtain the face similarity coefficient Xxs of several elevator passengers. After the similarity comparison, construct the identified resident passenger group, and issue a child lone travel analysis instruction; S3. After receiving the instruction for analyzing the children traveling alone, the personnel information of the identified resident passenger group is searched and matched to obtain the data information of the identified resident passenger, and the coefficient of the proportion of children traveling alone Zcp is obtained by using the set matching algorithm; S4. According to the value of the coefficient of proportion of children traveling alone Zcp, it is determined whether there are children riding the elevator alone among the identified resident passengers in the current elevator, so as to generate a corresponding level of danger warning and execute a corresponding warning reminder.

2. The intelligent early warning method for elevators for special personnel according to claim 1 is characterized in that: The specific steps of S1 include: S11. First, when the resident goes through the check-in procedure at the property center, the identity information of the resident is obtained by using the OCR text recognition technology based on the ID card, household registration booklet and property certificate provided by the resident, and the facial information of the resident is collected by the face recognition camera and the 3D structured light camera, and the facial feature information of the resident is obtained by combining the FaceNet algorithm, wherein the identity information of the resident includes the name, age Gnr and mobile phone number of each resident; the facial feature information of the resident includes the target facial feature vector Fmb of each resident; S12. Perform feature recognition on the facial feature information and identity information of the occupants, and combine with AES-256 encryption technology to construct relevant occupant storage feature data information, and store the encrypted relevant occupant storage feature data information in the cloud database of the elevator management system. When a person takes the elevator, the relevant data stored in the cloud database is called for identity authentication; wherein, the relevant occupant storage feature data information includes the age Gnr of each occupant, the target face feature vector Fmb and the family ID number Bid.

3. The intelligent early warning method for elevators for special personnel according to claim 2 is characterized in that: The specific steps of S1 also include: S13. When a person takes the elevator, multiple sets of cameras are arranged inside the elevator according to the need for identity recognition of the elevator passengers. The multiple sets of cameras are wirelessly connected to the elevator management system in combination with wireless network communication technology. The identity characteristics of elevator passengers are identified based on the multiple sets of cameras deployed to obtain relevant passenger identification feature data information; wherein the multiple sets of cameras include RGB cameras, infrared cameras and depth cameras, and the relevant passenger identification feature data information includes the passenger face feature vector Fcz of each passenger.

4. The intelligent early warning method for elevators for special personnel according to claim 3 is characterized in that: The specific steps of S2 include: S21, extracting features from the storage feature data information of the relevant occupants and the identification feature data information of the relevant passengers stored in the cloud database, and analyzing the similarity between the facial features of the passengers and the occupants in combination with the cosine similarity algorithm to construct a facial similarity coefficient Xxs, which is specifically obtained in the following manner: ; Where Fcz represents the passenger face feature vector of the corresponding passenger, and Fmb represents the target face feature vector of the occupants in the cloud database. It is represented by the length of the facial feature vector of the corresponding passenger. Represented as the length of the target facial feature vector of the occupants in the cloud database.

5. The intelligent early warning method for elevators for special personnel according to claim 4 is characterized in that: The specific steps of S2 also include: S22. According to the calculation method of step S21, the facial similarity coefficients Xxs of several elevator passengers are obtained respectively, and a similarity threshold is pre-set. By comparing the facial similarity coefficients Xxs of several elevator passengers with the similarity threshold, it is identified whether the elevator passenger is a resident. If the facial similarity coefficient Xxs does not exceed the similarity threshold, it means that the current elevator passenger is not a resident. At this time, the corresponding elevator passenger is included in the temporary passenger group after identification. If the facial similarity coefficient Xxs exceeds the similarity threshold, it means that the current elevator passenger is a resident and the person is marked. At this time, the marked elevator passenger is included in the resident passenger group after identification, and a child walking alone analysis instruction is issued.

6. The intelligent early warning method for elevators for special personnel according to claim 5, characterized in that: The specific steps of S3 include: S31. After receiving the instruction for analyzing the child traveling alone, the cloud database indexing technology is used to search and match the personnel information of the identified resident passenger group to obtain the identified resident passenger data information, wherein the identified resident passenger data information includes the age Gnr and family ID number Bid of each identified resident passenger, wherein the persons of the same family have the same family ID number Bid; S32, extracting features from the identified resident passenger data information obtained in step S31, and obtaining the number of children Nr and the number of adults Nc according to the age Gnr value of each identified resident passenger; ; In the formula, It is represented by the age of the i-th identified resident passenger, i=1, 2, 3, ..., N, N is the total number of identified resident passengers, where, and are all expressed as indicator functions.

7. The intelligent early warning method for elevators for special personnel according to claim 6, characterized in that: The specific steps of S3 also include: S33, according to the number of children Nr and the number of adults Nc obtained in step S32, and in combination with the identified household passenger data information, a single child ratio coefficient Zcp is obtained through a set matching algorithm, which is specifically obtained through the following formula: ; In the formula, It is the family ID number of the r-th child, r=1, 2, 3, ..., Nr, Nr is the number of children, It is represented by the family ID number of the jth adult, j=1, 2, 3, ..., Nc, Nc is the number of adults, where represents the indicator function, Indicates the proportion of all children who ride the elevator alone.

8. The intelligent early warning method for elevators for special personnel according to claim 7, characterized in that: The specific steps of S4 include: S41. According to the value of the lone child proportion coefficient Zcp, it is determined whether there are children riding alone in the current elevator after identification of the resident passengers, so as to generate a corresponding level of danger warning, the specific contents of which are as follows: If the lone child ratio coefficient Zcp>0, it means that there are children riding alone in the elevator among the identified resident passengers, indicating that there is a safety risk of children in the elevator without adult accompaniment. At this time, a first-level danger warning is issued. If the lone children ratio coefficient Zcp=0, it means that there are no children riding alone in the elevator among the residents and passengers identified in the current elevator, which means that there is no safety risk for children in the current elevator without adult accompaniment. At this time, a second-level danger warning is issued.

9. The intelligent early warning method for elevators for special personnel according to claim 8, characterized in that: The specific steps of S4 also include: S42: According to the first-level danger warning and the second-level danger warning issued in step S41, corresponding warning reminders are executed, and the specific execution contents are as follows: When receiving a level 1 danger warning, the content of the warning reminder is: an emergency voice prompt will be played in the elevator, "Warning! There is a child alone in the elevator. Please arrange for an adult to accompany the child as soon as possible. If there is an emergency, please contact the property center and call the emergency contact number immediately." The elevator management system will send the warning information to the parents and the property staff of the building where the elevator is located via SMS and phone. If the child fails to be accompanied by an adult immediately after the warning is issued, the elevator management system will initiate the emergency response process and further track the child through the surveillance camera inside the elevator; When receiving a Level 2 danger warning, the following warning reminder will be executed: a voice prompt will be played in the elevator: "All children in the elevator are currently accompanied by adults. If you have any needs, please contact the property center." At the same time, a reminder message will be sent to the property staff to inform them that there are no children riding alone in the elevator, and the surveillance cameras inside the elevator will continue to monitor whether there are children riding alone among the new entrants in the elevator.

10. An elevator intelligent early warning system for special personnel, used to implement an elevator intelligent early warning method for special personnel as described in any one of claims 1 to 9, characterized in that: It includes personnel collection module, similarity analysis module, children walking alone analysis module and level warning module; The personnel collection module is used to construct storage feature data information of relevant occupants based on the certificate information and facial information provided by the occupants when checking in, using OCR text recognition technology and FaceNet algorithm, and to obtain identification feature data information of relevant passengers by arranging multiple groups of cameras inside the elevator in combination with wireless network communication technology; The similarity analysis module is used to extract features from the stored feature data information of relevant occupants and the identification feature data information of relevant passengers, and obtain the face similarity coefficient Xxs of several elevator passengers in combination with the cosine similarity algorithm. After the similarity comparison, the identified occupant passenger group is constructed, and a child lone travel analysis instruction is issued externally; The single-child analysis module is used to retrieve and match the personnel information of the identified resident passenger group after receiving the single-child analysis instruction, obtain the identified resident passenger data information, and use the set matching algorithm to obtain the single-child ratio coefficient Zcp; The level warning module is used to determine whether there are children riding alone in the elevator among the identified resident passengers in the current elevator according to the numerical value of the lone children proportion coefficient Zcp, so as to generate a corresponding level of danger warning and execute a corresponding warning reminder.

Images