Search

CN-121982805-A - Hotel check-in method, device, equipment and storage medium based on face recognition

CN121982805ACN 121982805 ACN121982805 ACN 121982805ACN-121982805-A

Abstract

The invention relates to the technical field of intelligent hotel entrance control and discloses a hotel check-in method, a device, equipment and a storage medium based on face recognition, which comprise the steps of collecting radio frequency coupling strength and communication response time delay data during interaction of a radio frequency card reading module and a first near field communication unit of a smart card, resolving an environment coupling fluctuation value and a protocol time sequence fluctuation value, and configuring dynamic isolation time length according to the environment coupling fluctuation value and the protocol time sequence fluctuation value so as to delay activation of a second near field communication unit; according to the identification interval duration and the fluctuation value, the interaction stability parameter is calculated, the identity of the first near field communication unit is mapped to a preset face orientation angle, the face deflection angle of the current user is collected, the angle matching score is calculated, the interaction stability parameter is utilized to dynamically adjust the angle allowable deviation range and the stability weight, and finally the access control synthesized value is calculated to drive the door lock to be opened.

Inventors

  • MA XIANJUN

Assignees

  • 安徽云视通讯有限公司

Dates

Publication Date
20260505
Application Date
20260212

Claims (10)

  1. 1. The hotel check-in method based on face recognition is applied to an access control system comprising a radio frequency card reading module and a camera acquisition module, and is characterized by comprising the following steps: During the process that the radio frequency card reading module establishes a communication link with a first near field communication unit in the intelligent card, a radio frequency coupling strength data set and a communication response time delay data set are collected, an environment coupling fluctuation value and a protocol time sequence fluctuation value are calculated, and dynamic isolation duration is configured according to the environment coupling fluctuation value and the protocol time sequence fluctuation value so as to delay the activation operation of the radio frequency card reading module on a second near field communication unit in the intelligent card; According to the identification interval duration of the radio frequency card reading module to the first near field communication unit and the second near field communication unit, the environmental coupling fluctuation value and the protocol time sequence fluctuation value, calculating an interaction stability parameter, converting the identity of the first near field communication unit into an activation sequence code, and mapping to generate a preset face orientation angle; Acquiring the face deflection angle of the current user through a camera acquisition module, calculating an angle matching score aiming at the preset face orientation angle, and dynamically adjusting an angle allowable deviation range and a stability weight by utilizing the interaction stability parameter; And calculating an access control synthesized value by integrating the stability weight, the angle matching score and the face similarity, and sending an opening instruction to a door lock driving unit when the access control synthesized value meets a system threshold condition.
  2. 2. The face recognition-based hotel check-in method of claim 1, wherein collecting the radio frequency coupling strength data set and the communication response delay data set during the time period when the radio frequency card reading module establishes the communication link with the first near field communication unit in the smart card, and resolving the environmental coupling fluctuation value and the protocol time sequence fluctuation value, comprises: Setting a fixed sampling time window and a sampling frequency, and determining the number of sampling samples; In the fixed sampling time window, acquiring a radio frequency coupling strength data set by reading a received signal strength register of a radio frequency card reading module and carrying out normalization processing; Recording a time difference from the instruction sending starting time to the response receiving interruption time by using a microsecond counter, and acquiring a communication response time delay data set; calculating the square root of the variance of the radio frequency coupling strength dataset as an environmental coupling fluctuation value; And dividing the communication response time delay data set by a preset reference time delay scale for normalization processing, and calculating the square root of the variance of the normalized communication response time delay data set as a protocol time sequence fluctuation value.
  3. 3. The face recognition-based hotel check-in method of claim 2, wherein configuring the dynamic isolation duration to delay the activation of the second near field communication unit in the smart card by the radio frequency card reading module according to the environmental coupling fluctuation value and the protocol timing fluctuation value comprises: Calculating the change rate of adjacent sampling time by using the radio frequency coupling strength data set, calculating the root mean square value of the change rate, and obtaining an action disturbance value after adjustment of a preset scaling constant; Summing the environment coupling fluctuation value, the protocol time sequence fluctuation value, the action disturbance value and a preset minimum positive number to obtain a total jitter scale; multiplying the total jitter scale by a preset beat delay proportional coefficient, and adding the total jitter scale with a preset basic beat time length to obtain the dynamic isolation time length; And sending a suspension instruction to a first near field communication unit in the smart card, and starting an activation operation of a second near field communication unit in the smart card after waiting for the end of the dynamic isolation duration.
  4. 4. The face recognition-based hotel check-in method as claimed in claim 3, wherein the calculating the interaction stability parameter according to the recognition interval duration of the radio frequency card reading module to the first near field communication unit and the second near field communication unit and the environmental coupling fluctuation value and the protocol time sequence fluctuation value comprises: during the activation and identification of a second near field communication unit in the smart card by the radio frequency card reading module, a second radio frequency coupling strength data set and a second communication response time delay data set are collected; Calculating the variance of the second radio frequency coupling strength data set, adding the variance with the square of the environment coupling fluctuation value, and then opening a root number to obtain a comprehensive environment coupling fluctuation value; Normalizing the second communication response time delay data set, calculating variance, adding square of the protocol time sequence fluctuation value, and then opening root number to obtain a comprehensive protocol time sequence fluctuation value; calculating a second action disturbance value according to the second radio frequency coupling strength data set, and calculating the square root of the sum of the square of the action disturbance value and the square of the second action disturbance value to obtain a comprehensive action disturbance value; Dividing the identification interval duration by a preset time normalization reference constant to obtain a normalization interval, and dividing the normalization interval by the sum of the comprehensive environment coupling fluctuation value, the comprehensive protocol time sequence fluctuation value, the comprehensive action disturbance value and a preset minimum positive number to obtain the interaction stability parameter.
  5. 5. The face recognition-based hotel check-in method of claim 4, wherein converting the identity of the first near field communication unit to an activation order code to map to generate a preset face orientation angle, comprises: the method comprises the steps of storing a first chip identifier and a second chip identifier in advance, and setting a first sequence target angle and a second sequence target angle; Comparing the identity of a first near field communication unit with the first chip identifier and the second chip identifier, and if the first chip identifier is matched, assigning the activation sequence code as a positive unit value; calculating a first linear interpolation weight and a second linear interpolation weight based on the activation order encoding, wherein the first linear interpolation weight is half of a sum of one and the activation order encoding, and the second linear interpolation weight is half of a difference between one and the activation order encoding; And calculating a first product of the first sequence target angle and the first linear interpolation weight and a second product of the second sequence target angle and the second linear interpolation weight, and adding the first product and the second product to obtain the preset face orientation angle.
  6. 6. The face recognition-based hotel check-in method of claim 5, wherein obtaining the face deflection angle of the current user by the camera acquisition module, calculating an angle matching score for the preset face orientation angle, comprises: acquiring an image containing the face of a user through a camera acquisition module, extracting two-dimensional key points, solving a rotation matrix relative to a coordinate system of the camera acquisition module by combining preset three-dimensional head model key points, and calculating a four-quadrant arctangent value by using elements of the rotation matrix to obtain the face deflection angle; Calculating the square of the difference between the face deflection angle and the preset face orientation angle as a deviation square term; calculating the product of the square of the angle allowable deviation range and a preset constant two as a normalization factor; calculating the inverse of the quotient of the square deviation term divided by the normalization factor as a gaussian index; And calculating a function value taking the natural constant as a base and taking the Gaussian index as an index to obtain the angle matching score.
  7. 7. The face recognition-based hotel check-in method of claim 6, wherein dynamically adjusting the angle tolerance range and the stability weight by using the interactive stability parameter, integrating the stability weight, the angle matching score and the face similarity, calculating an entrance guard control composite value, and when the entrance guard control composite value meets a system threshold condition, sending an opening instruction to a door lock driving unit, comprising: calculating a function value taking a natural constant as a base and taking the first index term as an index, multiplying the function value by a preset variable tolerance amplitude, and adding the function value to a preset minimum tolerance constant to obtain the angle allowable deviation range; calculating a function value taking a natural constant as a bottom and taking the second index term as an index, adding the function value and a constant one to obtain a middle sum value, and calculating the reciprocal of the middle sum value to obtain the stability weight; Carrying out validity check on the smart card, setting the validity verification value of the smart card to be one when the verification is passed, otherwise, setting the validity verification value to be zero; Extracting a face feature vector of a current user and a prestored resident registration template vector, and calculating cosine similarity as the face feature similarity; calculating the continuous product of the intelligent card validity verification value, the stability weight, the angle matching score and the normalized face similarity to obtain the access control synthesized value; And calculating a difference value between the door control synthesized value and a preset system opening threshold value, judging the difference value by using a step function, and generating an opening instruction and sending the opening instruction to a door lock driving unit when the difference value is greater than or equal to zero.
  8. 8. The hotel check-in device based on face recognition, which is applied to an access control system comprising a radio frequency card reading module and a camera acquisition module, comprises the steps of implementing the hotel check-in method based on face recognition as claimed in any one of claims 1 to 7, and is characterized by comprising the following steps: The dynamic isolation control module is used for collecting a radio frequency coupling strength data set and a communication response time delay data set during the period that the radio frequency card reading module establishes a communication link with a first near field communication unit in the intelligent card, resolving an environment coupling fluctuation value and a protocol time sequence fluctuation value, and configuring dynamic isolation time length according to the environment coupling fluctuation value and the protocol time sequence fluctuation value so as to delay the activation operation of the radio frequency card reading module on a second near field communication unit in the intelligent card; The parameter resolving and mapping module is used for resolving the interaction stability parameter according to the identification interval duration of the radio frequency card reading module to the first near field communication unit and the second near field communication unit, the environment coupling fluctuation value and the protocol time sequence fluctuation value, and converting the identity of the first near field communication unit into an activation sequence code so as to map and generate a preset face orientation angle; The visual analysis and adjustment module is used for acquiring the face deflection angle of the current user through the camera acquisition module, calculating an angle matching score aiming at the preset face orientation angle, and dynamically adjusting the angle allowable deviation range and the stability weight by utilizing the interaction stability parameter; And the entrance guard decision executing module is used for calculating an entrance guard control synthesized value by integrating the stability weight, the angle matching score and the face similarity, and sending an opening instruction to the door lock driving unit when the entrance guard control synthesized value meets the system threshold condition.
  9. 9. An electronic device, comprising: the device comprises a memory, a processor, a radio frequency card reading module and a camera acquisition module; The memory is used for storing a computer program; The processor is configured to execute the computer program to implement the steps of the face recognition based hotel check-in method of any one of claims 1 to 7.
  10. 10. A computer readable storage medium, on which a computer program is stored, characterized in that the computer program, when being executed by a processor, implements the steps of the face recognition based hotel check-in method as claimed in any one of claims 1 to 7.

Description

Hotel check-in method, device, equipment and storage medium based on face recognition Technical Field The invention relates to the technical field of intelligent hotel entrance guard control, in particular to a hotel check-in method, a hotel check-in device, hotel check-in equipment and a hotel check-in storage medium based on face recognition. Background With the development of smart hotels, a non-contact radio frequency identification (RFID/NFC) technology and a face recognition technology are widely applied to a guest room door lock system. Existing high security level check-in procedures typically require the resident to hold a house card and coordinate with face verification. In order to improve anti-counterfeiting capability and interaction experience, a smart card with embedded dual communication units (dual chips) is partially adopted, and an attempt is made to trigger different verification logics by detecting the activation sequence of two communication units (for example, identifying unit A and then identifying unit B or vice versa), for example, requiring a user to cooperate to complete living body detection of a left face or a right face. However, in a practical hotel engineering scenario, the above solution faces challenges. Firstly, the panel structure of the hotel door lock is made of metal materials such as zinc alloy, and the environment of the material can cause serious resonance elimination effect of near field communication in the 13.56MHz frequency band, so that the antenna coupling efficiency is reduced and is not uniform. Secondly, when the house card of the double communication unit enters the sensing area of the card reader, an anti-collision and card selection process of the ISO/IEC-14443 standard is triggered. Under the interference of metal environment, the time of selection of the two communication units by the card reader is not simply dependent on the card swinging direction of the user, but is influenced by the double effects of environment coupling fluctuation and the randomness of the communication protocol time slot. In addition, when a resident handles to check in or open a door, short residence actions such as quick card swinging and hand shielding are often accompanied. During such fast and unstable operation, the signal strength and response time delay collected by the reader can produce severe jitter. If the user intention is judged and the corresponding face orientation verification is triggered only by relying on a simple sequence of identification, the misjudgment of the system is easily caused by signal drift (for example, the user wants to trigger the left face verification, the system firstly reads an incorrect chip to request the right face due to signal fluctuation), so that verification failure or door opening can not be caused, and the living experience and passing efficiency of residents are seriously affected. Disclosure of Invention The invention provides a hotel check-in method, a hotel check-in device, a hotel check-in equipment and a hotel check-in storage medium based on face recognition, which solve the technical problems in the background art. In a first aspect, a hotel check-in method based on face recognition is applied to an access control system including a radio frequency card reading module and a camera acquisition module, and includes: During the process that the radio frequency card reading module establishes a communication link with a first near field communication unit in the intelligent card, a radio frequency coupling strength data set and a communication response time delay data set are collected, an environment coupling fluctuation value and a protocol time sequence fluctuation value are calculated, and dynamic isolation duration is configured according to the environment coupling fluctuation value and the protocol time sequence fluctuation value so as to delay the activation operation of the radio frequency card reading module on a second near field communication unit in the intelligent card; According to the identification interval duration of the radio frequency card reading module to the first near field communication unit and the second near field communication unit, the environmental coupling fluctuation value and the protocol time sequence fluctuation value, calculating an interaction stability parameter, converting the identity of the first near field communication unit into an activation sequence code, and mapping to generate a preset face orientation angle; Acquiring the face deflection angle of the current user through a camera acquisition module, calculating an angle matching score aiming at the preset face orientation angle, and dynamically adjusting an angle allowable deviation range and a stability weight by utilizing the interaction stability parameter; And calculating an access control synthesized value by integrating the stability weight, the angle matching score and the face similarity, and sending an opening instr