Method and system for guiding old people to go home through door lock

Abstract

The invention discloses a method and system for guiding an old person to go home through a door lock, and the method comprises the steps: calculating a lost score value through a lost score model according to a condition that a target door lock A detects that the old person fails in continuous verification for multiple times and stays overtime in front of the door, and judging that the old person is in a lost state when the score value exceeds a threshold value, and obtaining a lost judgment result; according to the door lock A, encrypting and broadcasting the elderly biological characteristics to all the door locks, receiving confirmation information returned by the successfully matched door lock B, and obtaining a target room positioning result; according to the floor topological data of the door lock A and the door lock B, generating an optimal guide path, and obtaining path data containing a passing door lock sequence; according to the path data, controlling lamp strips along the way to be lightened according to a preset mode, and cooperating with a directional sound box to play voice guidance; and sending a reset instruction to the path door lock according to the fact that the door lock B detects that the old person successfully unlocks or arrives. According to the embodiment of the invention, convenient, efficient and intelligent door lock guiding can be carried out, and precise guiding and supporting for the elderly are realized.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent door locks, and particularly relates to a method and system for guiding the elderly to go home by a door lock. Background Art

[0002] With the intensification of population aging, the problem of the travel safety of the elderly has become increasingly prominent. Getting lost is particularly common in unfamiliar environments, bringing emotional anxiety and potential safety hazards to the lives of the elderly. Existing positioning and navigation systems are mostly designed for young people and lack special designs for the elderly. Moreover, traditional door lock technologies mainly focus on security and fail to effectively identify and address the problems encountered by the elderly. Summary of the Invention

[0003] The purpose of the present invention is to provide a method and system for guiding the elderly to go home by a door lock, so as to solve the deficiencies in the prior art, and be able to perform convenient, efficient and intelligent door lock guidance, and achieve precise guidance and support for the elderly.

[0004] An embodiment of the present application provides a method for guiding the elderly to go home by a door lock, and the method includes: Judgment of lost behavior: According to the target door lock A detecting that the elderly fails to verify continuously for multiple times and stays in front of the door for overtime, calculate the lost score value through a lost score model. When the score value exceeds the threshold, it is determined to be in a lost state; wherein, the lost score model fuses the number of verification failures, the stay duration, and the radar trajectory confidence to obtain a lost judgment result; Cross-room identity matching: Encrypt and broadcast the biometric characteristics of the elderly by the door lock A to all door locks, and receive the confirmation information returned by the successfully matched door lock B; wherein, the biometric characteristics include a face feature vector and a fingerprint hash value, to obtain the target room positioning result; Dynamic path planning: Generate an optimal guiding path according to the floor topology data of the door lock A and the door lock B; wherein, the path on the same floor is directly planned, and the path across floors is transferred through the connecting point door lock, to obtain path data including the sequence of passing door locks; Multi-modal guidance execution: According to the path data, control the strip lights along the way to be lit in a preset mode, and cooperate with the directional speaker to play voice guidance; wherein, the strip lights display direction arrows and the non-path segments remain extinguished, to obtain a real-time guidance effect; End point confirmation and reset: According to the door lock B detecting that the elderly successfully unlocks or arrives, send a reset instruction to the path door locks; wherein, the instruction triggers the strip lights to go out and the log to be uploaded, to complete the guidance closed-loop.

[0005] Optionally, the judgment of lost behavior includes: Multi-modal Verification Failure Detection: Based on Door Lock A, when it detects that the elderly person fails in face / fingerprint / password verification three times consecutively, record the failure count N; where the interval between each verification is less than 5 seconds to obtain a verification failure record; Behavioral Feature Assisted Judgment: Based on the millimeter-wave radar, when it monitors that the elderly person stays in front of the door for more than 30 seconds without leaving, record the stay duration T; where the radar synchronously generates a trajectory confidence level S to obtain behavioral monitoring data; Lost Score Calculation: According to the formula P = α⋅N + β⋅T + γ⋅S, calculate the lost score value; where α, β, and γ are weight coefficients, N is the verification failure count, T is the stay duration, and S is the radar trajectory confidence level. When P > 1, trigger a lost determination to obtain a lost status confirmation signal.

[0006] Optionally, the cross-room identity matching includes: Feature Broadcasting: Encrypt a broadcast packet containing a face feature vector and a fingerprint hash value according to Door Lock A and send it to all door locks via a local area network; where the broadcast packet is appended with the ID of Door Lock A to obtain a feature diffusion result; Feature Comparison: According to Door Lock B receiving the broadcast packet, compare its face feature with the local registered feature library; where the cosine similarity algorithm is used, and when the similarity exceeds 95%, it is determined to be a match to obtain a matching result; Target Confirmation: According to Door Lock B, unicast a confirmation message containing its own ID to Door Lock A; where the message is encrypted by AES to obtain a precise positioning of the target room.

[0007] Optionally, the dynamic path planning includes: Floor Topology Analysis: According to Door Lock A, retrieve the locally stored floor topology map to determine whether Door Lock B is on the same floor; where the topology map contains door lock positions and connection point information to obtain a floor relationship determination; Path Generation: According to the same-floor determination result, if on the same floor, plan a straight path, and if across floors, add stair / elevator connection point door locks; where when the floor interval is 1 floor, the path priority of the stairs > the path priority of the elevator, and when the floor interval is greater than or equal to 2 floors, the path priority of the elevator > the path priority of the stairs to obtain an optimized path sequence; Path Binding: According to the generated path, bind the elderly person's face feature to the ID of the passing door locks; where the bound data is saved to the cache of each door lock to obtain a path persistence result.

[0008] Optionally, the multi-modal guidance execution includes: Send a lighting instruction to the corresponding light strip according to the path door lock; where the instruction contains color and arrow direction, and the response delay of the light strip < 200ms to obtain a visual guidance effect; Voice collaboration: According to the trigger of the door lock at the connection point, the directional speaker plays guiding voice; among them, the voice content includes the name of the elderly and the turning instruction, and the broadcast interval is 10 seconds to obtain the auditory guiding effect; Abnormal handling: According to the camera detecting that the elderly person goes in the wrong direction, control the light strip in the opposite direction to turn red and play corrective voice; among them, the inner tube of the light strip on the correct path is synchronously activated to obtain the deviation correction guiding effect.

[0009] Optionally, the end point confirmation and reset include: Successful arrival determination: According to the door lock B detecting that the elderly person passes the verification or the camera confirming that they enter the room, generate an unlocking success event; among them, the event includes the facial features of the elderly person to obtain the end point confirmation signal; System reset: According to the door lock B broadcasting a reset instruction to the path door locks; among them, the instruction triggers the light strip to go out, clears the cached feature data and uploads the log to obtain the system reset result; Emergency handling: According to the event of not arriving within the time limit, the door lock B activates the red flashing light strip and notifies the caregiver; among them, continue to guide until manual intervention to obtain the emergency response result.

[0010] Optionally, the method further includes: According to the door lock A detecting that the elderly person X gets lost and generating the first path plan, broadcast the path data including the first light strip identifier to the passing door locks to obtain the first activated path; The door lock B detects that the elderly person Y gets lost and broadcasts the path planning information to the path door locks. The door lock L is on the first activated path and receives the path planning from the door lock B, and feeds back a path conflict notice to the door lock B to obtain the conflict detection result; According to the conflict detection result, the door lock B selects the second light strip identifier to re-broadcast the path data, and the passing door locks synchronously update the light strip display mode to obtain a differentiated path; When the door lock C detects that the elderly person Z gets lost and receives the conflict notice from the door lock L, after planning the third light strip path, before the door lock J is about to plan the path, it receives a conflict notice including three light strip identifiers, and waits for any path to end and then activates the path with the third light strip identifier to obtain the time-division multiplexing guiding effect.

[0011] Optionally, the method further includes: According to the non-path door lock E detecting that the elderly person X walks in the reverse direction through the camera, broadcast a path acquisition request including their facial features to obtain the reverse walking event; According to the door lock A that initiates the navigation to respond to the request and return the original path data, the door lock E combines the floor topology to generate a supplementary path including the error section to obtain enhanced path information; According to the door lock E synchronously detecting that the elderly person Y walks in the wrong direction in the same direction, activate the correct direction indication corresponding to the second light strip identifier in the supplementary path to obtain a dual-path deviation correction interface; Based on the light strip system that simultaneously displays red error warnings and two-color correct path identifications, combined with voice differentiation prompts, a multi-person collaborative deviation correction effect is obtained.

[0012] Another embodiment of the present application provides a system for guiding the elderly home by a door lock. The system includes: A determination module for determining lost behavior: According to the target door lock A, it detects that the elderly person fails to verify continuously for multiple times and stays in front of the door for an excessive time. The lost score value is calculated through a lost score model. When the score value exceeds the threshold, it is determined as a lost state; wherein, the lost score model fuses the number of verification failures, the stay duration, and the radar trajectory confidence to obtain a lost determination result; A matching module for cross-room identity matching: According to the door lock A, the biometric characteristics of the elderly person are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received; wherein, the biometric characteristics include a face feature vector and a fingerprint hash value to obtain a target room positioning result; A planning module for dynamic path planning: According to the floor topology data of the door lock A and the door lock B, an optimal guiding path is generated; wherein, the path on the same floor is directly planned, and the path across floors is transferred through the connecting point door lock to obtain path data including the sequence of passing door locks; An execution module for multi-modal guiding execution: According to the path data, control the light strips along the way to be lit in a preset mode, and cooperate with the directional speaker to play voice guidance; wherein, the light strip displays a direction arrow and the non-path segments remain extinguished to obtain a real-time guiding effect; A reset module for end confirmation and reset: According to the door lock B detecting that the elderly person has successfully unlocked or arrived, send a reset instruction to the path door locks; wherein, the instruction triggers the light strip to go out and the log to be uploaded to complete the guiding closed-loop.

[0013] Another embodiment of the present application provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the method described in any one of the above.

[0014] Compared with the prior art, a method for guiding the elderly to go home by a door lock provided by the present invention detects that the elderly have failed to verify continuously for multiple times and the stay time in front of the door exceeds the time limit through the target door lock A, calculates the lost score value through the lost score model, determines the lost state when the score value exceeds the threshold, and obtains the lost determination result; encrypts and broadcasts the biometric characteristics of the elderly to all door locks according to the door lock A, receives the confirmation information returned by the successfully matched door lock B, and obtains the target room positioning result; generates the optimal guiding path according to the floor topology data of the door lock A and the door lock B, and obtains the path data including the sequence of passing door locks; controls the strip lights along the way to be lit in a preset mode according to the path data, and collaborates with the directional speaker to play voice guidance; sends a reset instruction to the path door locks according to the detection that the elderly have successfully unlocked or arrived by the door lock B, so as to be able to perform convenient, efficient and intelligent door lock guidance, and realize precise guidance and support for the elderly. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a hardware structure block diagram of a computer terminal for a method for guiding the elderly to go home by a door lock provided by an embodiment of the present invention; Figure 2 It is a flowchart of a method for guiding the elderly to go home by a door lock provided by an embodiment of the present invention; Figure 3 It is a structural diagram of a system for guiding the elderly to go home by a door lock provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0016] The embodiments described below by referring to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as a limitation to the present invention.

[0017] An embodiment of the present invention first provides a method for guiding the elderly to go home by a door lock. This method can be applied to electronic devices, such as computer terminals, specifically ordinary computers, etc.

[0018] The following takes running on a computer terminal as an example to explain it in detail. Figure 1 It is a hardware structure block diagram of a computer terminal for a method for guiding the elderly to go home by a door lock provided by an embodiment of the present invention. As Figure 1 shown, the computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory may include a non-volatile storage medium and an internal memory.

[0019] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any method for guiding the elderly to go home by a door lock.

[0020] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0021] The internal memory provides an environment for the operation of a computer program in a non-volatile storage medium. When the computer program is executed by a processor, the processor can be made to execute any method for guiding an elderly person home by a door lock.

[0022] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 1 the structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the computer device to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have a different component layout.

[0023] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0024] See Figure 2 , embodiments of the present invention provide a method for guiding an elderly person home by a door lock, which may include the following steps: S201, lost behavior determination: According to the target door lock A, it is detected that the elderly person fails to verify continuously for multiple times and stays in front of the door for an overtime. The lost score value is calculated through a lost score model. When the score value exceeds the threshold, it is determined to be in a lost state; wherein, the lost score model fuses the number of verification failures, the stay duration, and the radar trajectory confidence to obtain a lost determination result; specifically, the lost behavior determination includes: Multi-modal verification failure detection: According to the door lock A, it is detected that the elderly person fails in face / fingerprint / password verification continuously 3 times, and the number of failures N is recorded; wherein, the interval between each verification is less than 5 seconds to obtain a verification failure record; Behavior feature assisted determination: According to the millimeter-wave radar, it is monitored that the elderly person stays in front of the door for more than 30 seconds without leaving action, and the stay duration T is recorded; wherein, the radar synchronously generates the trajectory confidence S to obtain behavior monitoring data; Lost path score calculation: Calculate the lost path score value according to the formula P = α⋅N + β⋅T + γ⋅S; where α, β, and γ are weight coefficients, N is the number of verification failures, T is the stay duration, S is the radar trajectory confidence. When P > 1, trigger the lost path determination to obtain the lost path status confirmation signal.

[0025] S202, Cross-room identity matching: Encrypt and broadcast the biometric characteristics of the elderly to all door locks according to door lock A, and receive the confirmation information returned by the successfully matched door lock B; where the biometric characteristics include the face feature vector and the fingerprint hash value to obtain the target room positioning result; specifically, the cross-room identity matching includes: Feature broadcast: Encrypt a broadcast packet containing the face feature vector and the fingerprint hash value according to door lock A, and send it to all door locks through the local area network; where the broadcast packet is attached with the ID of door lock A to obtain the feature diffusion result; Feature comparison: According to door lock B receiving the broadcast packet, compare its face feature with the local registered feature library; where the cosine similarity algorithm is used, and when the similarity exceeds 95%, it is determined as a match to obtain the matching result; Target confirmation: According to door lock B, unicast a confirmation message containing its own ID to door lock A; where the message is encrypted by AES to obtain the precise positioning of the target room.

[0026] S203, Dynamic path planning: Generate the optimal guiding path according to the floor topology data of door lock A and door lock B; where the paths on the same floor are directly planned, and the paths across floors are transferred through the connecting point door locks to obtain the path data including the sequence of passing door locks; specifically, the dynamic path planning includes: Floor topology analysis: Retrieve the local stored floor topology map according to door lock A to determine whether door lock B is on the same floor; where the topology map includes door lock positions and connection point information to obtain the floor relationship determination; Path generation: According to the same-floor determination result, if on the same floor, plan a straight path, and if across floors, add the stair / elevator connecting point door locks; where when the floor interval is 1 floor, the path priority of the stairs > the path priority of the elevator, and when the floor interval is greater than or equal to 2 floors, the path priority of the elevator > the path priority of the stairs to obtain the optimized path sequence; Path binding: According to the generated path, bind the face feature of the elderly with the ID of the passing door locks; where the binding data is saved to the cache of each door lock to obtain the path persistence result.

[0027] S204, Multimodal guidance execution: According to the path data, control the strip lights along the way to light up in a preset mode, and cooperate with the directional speaker to play voice guidance; where the strip lights display direction arrows and the non-path segments remain off to obtain the real-time guidance effect; specifically, the multimodal guidance execution includes: Send a lighting instruction to the affiliated light strip according to the path door lock; among them, the instruction includes color and arrow direction, and the response delay of the light strip < 200ms to obtain a visual guidance effect; Voice collaboration: Trigger the directional speaker to play guiding voice according to the connection point door lock; among them, the voice content includes the name of the elderly and the turning instruction, and the broadcast interval is 10 seconds to obtain an auditory guidance effect; Abnormal handling: Detect that the elderly go in the wrong direction according to the camera, control the light strip in the reverse direction to turn red and play error correction voice; among them, synchronously activate the inner tube of the correct path light strip to obtain a deviation correction guidance effect.

[0028] S205, End point confirmation and reset: Send a reset instruction to the path door lock according to the door lock B detecting that the elderly have successfully unlocked or arrived; among them, the instruction triggers the light strip to go out and the log to be uploaded to complete the guidance closed loop. Specifically, the end point confirmation and reset includes: Successful arrival determination: Generate an unlocking success event according to the door lock B detecting that the elderly have passed the verification or the camera confirming that they have entered the room; among them, the event includes the face features of the elderly to obtain an end point confirmation signal; System reset: Broadcast a reset instruction to the path door lock according to the door lock B; among them, the instruction triggers the light strip to go out, clears the cached feature data and uploads the log to obtain a system reset result; Emergency handling: According to the event of not arriving within the time limit, the door lock B activates the red flashing light strip and notifies the caregiver; among them, continue to guide until manual intervention to obtain an emergency response result.

[0029] Further, the method further includes: Detect that the elderly X is lost according to the door lock A and generate a first path plan, and broadcast path data including the first light strip identifier (thick green line) to the passing door locks to obtain a first activated path; The door lock B detects that the elderly Y is lost and broadcasts path planning information to the path door lock. The door lock L is on the first activated path and receives the path planning of the door lock B. If it judges that the paths overlap, it will feedback a path conflict notification to the door lock B to obtain a conflict detection result; According to the conflict detection result, the door lock B selects the second light strip identifier (blue dotted line) to re-broadcast the path data, and the passing door locks synchronously update the light strip display mode to obtain a differentiated path; After the door lock C detects that the elderly Z is lost and receives the conflict notification from the door lock L, and plans the third light strip path, before the door lock J prepares to plan the path, it receives a conflict notification including three light strip identifiers, and waits for any path to end and then activates the path with the third light strip identifier (thin yellow line) to obtain a time-division multiplexing guidance effect.

[0030] Further, the method further includes: Detect that the elderly X is walking in the reverse direction according to the non-path door lock E through the camera, and broadcast a path acquisition request including his face features to obtain a reverse walking event; According to the door lock A that initiates navigation to respond to the request and return the original path data, the door lock E combines the floor topology to generate a supplementary path including the error section, obtaining enhanced path information; According to the door lock E synchronously detecting the wrong walking of the elderly Y in the same direction, activating the correct direction indication corresponding to the second light strip identifier in the supplementary path, obtaining a dual-path deviation correction interface; According to the light strip system simultaneously displaying a red error warning and a two-color correct path identifier, combined with voice differentiation prompts ("X, please return to the green path, Y, please return to the blue path"), obtaining a multi-person collaborative deviation correction effect.

[0031] In practical applications, the elderly in nursing homes often go to the wrong room due to cognitive decline. Traditional door locks only provide identity verification functions and cannot actively identify and correct wrong behaviors. Single biometric recognition (such as face recognition) has a high misjudgment rate when the face of the elderly changes (such as wearing a mask or expression), and cannot distinguish between the intentions of "going to the wrong room" and "visiting". Existing guiding systems rely on fixed signs or manual assistance and lack the ability of real-time dynamic guidance, especially causing anxiety among the elderly at night.

[0032] Scheme summary: The present invention identifies lost behaviors through the number of failures of multi-modal identity verification (face, password, fingerprint) and behavioral characteristics (such as wandering duration), dynamically generates the optimal path from room to room through low-power light strips, and collaborates with voice prompts to solve the problem of active guidance for the elderly with cognitive impairments after going to the wrong room.

[0033] Solution: I. Prerequisites Intelligent door locks are installed for each household on each floor of the nursing home. The door locks are equipped with basic modules such as cameras and millimeter-wave radars, and the door locks can communicate with each other. The floor topology of this floor (door lock position data, whether it is close to the cross-floor connection point) is pre-stored in the door locks on each floor. The guiding light strips are interconnected on the ground between adjacent door locks. Combining the internal structure of the building, the guiding light strips on the same floor lead to the cross-floor connection point (staircase or elevator).

[0034] Each door lock stores the face feature vectors of its own family members.

[0035] II. Scheme steps 1: Determination of lost behavior a. Triggered by multi-modal verification failure: The elderly stands in front of room A (taking room number 303 as an example, door lock number 3-03), tries face recognition (completed within 1 second), fingerprint verification mode or password login mode. If the verification fails continuously 3 times (including face + fingerprint / password), the indicator light on the top of the door lock turns orange and flashes slowly. During this period, the millimeter-wave radar is activated to judge the behavior of the elderly.

[0036] b. Behavioral assistance determination: The millimeter-wave radar detects that the elderly person stays and wanders in front of the door for more than 30 seconds without leaving, and records the staying and wandering time.

[0037] C. Define the lost path scoring formula: P_lost = α ⋅ N_verification_failure + β ⋅ T_stay + γ ⋅ S_radar_trajectory_confidence. Where α, β, and γ are weight coefficients, N is the number of verification failures, T is the staying duration, and S is the radar trajectory confidence. For example, α = 0.04, β = 0.03, γ = 0.1. When P_lost > 1, the behavior detection and judgment module is triggered, and the elderly person is likely to be lost and needs behavior guidance.

[0038] 2: Cross-room identity matching a. Door lock A broadcasts the encrypted face feature vector to all smart locks through the protocol, including: the face feature vector of the current elderly person, the fingerprint feature hash value, and the door lock ID of door lock A (door lock ID - A).

[0039] b. After receiving it, other smart locks compare the face feature vector through the identity matching engine. Door lock B finds that the elderly person is a resident of its own room and unicasts a confirmation message to door lock A, including the ID of door lock B and the above face feature vector.

[0040] Implementation example: Elderly person -> Door lock A: Try face recognition (the 1st failure); Door lock A -> Door lock A: Record the number of failures N = 1; Elderly person -> Door lock A: Try fingerprint verification (the 2nd failure); Door lock A -> Door lock A: Update N = 2 and start wandering monitoring; Door lock A -> Door lock A: Monitor continuous residence T = 25 seconds; Door lock A -> Door lock A: Calculate the lost path score P = 0.93 (not reaching the threshold, taking the radar trajectory confidence as 100% as an example); Elderly person -> Door lock A: Enter the password (the 3rd failure); Door lock A -> Door lock A: N = 3, T = 32 seconds, P = 1.18; Door lock A -> All door locks: Send an encrypted feature packet (including face feature vector + fingerprint hash); Door lock B -> Door lock A: Feature matching is successful, locate the target room B.

[0041] 3: Path planning If the door lock B successfully matches the face feature vector of the elderly person and the door lock A receives the confirmation message from the door lock B, the door lock A starts the path planning to the door lock B. Based on the floor topologies pre-stored in each door lock, the door lock A and the door lock B exchange their floor topologies with each other and compare them with the local topology. If the door lock A finds that the door lock B (such as room number 308, door lock number 3-08) is on the same floor as itself, it is judged to be on the same floor, and the door lock A selects the direct path to the door lock B as the shortest path; if the door lock B (such as room number 203, door lock number 2-03) and the door lock A are not on the same floor, it is judged to be on different floors, and the door lock A selects to pass through the door lock N at the adjacent connection point on this floor to reach the door lock M at the adjacent connection point on the floor of the door lock B, and finally reaches the door lock B. When the floor interval is 1 floor, the door lock A preferentially selects the door lock passing through the stairwell as the floor connection point to reach the door lock B; when the floor interval is greater than or equal to 2 floors, the door lock A preferentially selects the door lock passing through the elevator as the floor connection point to reach the door lock B.

[0042] The door lock A sends the path data to the door lock B to all the door locks on the path. (Taking the floor interval of 1 as an example: "Passing through door locks 3-03, 3-04, 3-05... on the third floor, the adjacent connection point door lock 3-09 goes downstairs to the stairwell, and the exit of the second-floor stairwell passes through the adjacent connection point door locks 209, 208, 207... 204 to reach the target door lock 203"), and the above message also includes the face feature vector of the above-mentioned elderly person.

[0043] All the door locks on the path from the door lock A to the door lock B record the face feature vector and path information of the above-mentioned elderly person and bind them.

[0044] 4: Voice prompt The door lock A plays the voice: "Dear X (the name of the elderly person who went to the wrong room), the current location is room 305 on the third floor, and your room is room 203 on the second floor". Please follow the guidance of the green light path to reach your room.

[0045] 5: Dynamic path generation and guidance After the path planning is completed, all the door locks on the path simultaneously send a lighting instruction to the light strip system. The light strip system and the door locks are networked through a low-power wireless protocol (such as ZigBee Mesh) to ensure that the instruction transmission delay < 200ms.

[0046] a. Light strip activation: Each door lock on the path from the door lock A to the door lock B sends a lighting signal to the light strip from this door lock to the next door lock. The light strips along the path are all lit (green), and each light strip simultaneously displays an arrow as the forward direction, and the non-path light strips remain in the sleep state.

[0047] b. Lighting guidance: After the elderly person walks along the green light strip (correct direction), the light strip in the forward direction continues to be highly illuminated. When passing the door lock of the next room along the way (door lock C), this door lock continues to judge the forward direction of the elderly person. If the direction is correct, the light strip between door lock C and door lock A is turned off.

[0048] c. Wrong-way guidance: If the camera of any door lock on the path (such as door lock E) detects that the elderly person is walking in the reverse direction (and it does not save the face feature vector of this elderly person), then door lock E lights up the light strip it belongs to in red and plays a voice prompt: "Dear X (name of the elderly person who has taken the wrong room), your forward direction is incorrect. Please return to the green light path."

[0049] 6: Voice collaborative prompt Deploy directional speakers at the cross-floor connection points (such as stairwells). The speakers and the nearby door locks are networked through a low-power wireless protocol (such as ZigBee) to ensure that the instruction transmission delay < 200 ms. The speaker continuously plays the guiding voice (such as "Dear X, please turn right into the stairwell and go downstairs to the second floor, then turn left into the second-floor corridor"). After arriving at the second floor, the speakers at the connection points on this floor also play a voice broadcast for the path guidance on this floor, such as "Dear XX, please turn left when you exit the stairwell and continue to follow the guidance of the green light path to reach your room").

[0050] 7: Endpoint confirmation and system reset a. After the elderly person arrives at door lock B and unlocks it through any verification method, door lock B sends a "successful unlock" message containing the face feature vector of this elderly person to all the door locks on the corresponding path of this elderly person. After each door lock on the path receives the "successful unlock" message, it turns off the light strip it is responsible for on the path corresponding to the above face feature vector, exits the path guidance mode, and uploads the log to the intelligent lock management platform.

[0051] b. If the door is opened in advance by other relatives or friends, when the camera of door lock B detects that the elderly person has arrived, it automatically plays a voice: "Dear X, you have arrived at your room safely. The guiding light strip will be turned off for you soon. Wish you good health and a happy life"; door lock B also sends a "successful unlock" message to all the door locks on the corresponding path of this elderly person. All the door locks on the corresponding path of this elderly person send a "successful unlock" message, exit the path guidance mode, and upload the log to the intelligent lock management platform.

[0052] 8: Emergency response a. When the elderly do not reach the target room within more than 5 minutes, the target door lock (lock B) notifies the caregiver through an alarm sound and activates the guiding light strip to an emergency state (red flashing) until the elderly arrive at the target room accompanied by the caregiver, and after step 8, key confirmation and system reset are performed. If the elderly have a sudden emergency and need to seek medical treatment or go home and cannot reach the target room, after the caregiver confirms by face recognition, each door lock exits the path guidance mode and uploads the log to the intelligent lock management platform.

[0053] 9: Further improvement (1): How to distinguish when multiple people guide simultaneously This solution supports up to 3 elderly people for simultaneous lost path guidance. Each door lock's corresponding light strip contains 4 inner tubes that can display different colors and states (such as a thick green line, a dotted blue line, a thin yellow line, and a red wrong path warning line). When 3 elderly people need guidance on an intersecting path, the door locks adopt a peak-shifting guidance strategy: For example, elderly person X gets lost at door lock A and the actual destination is door lock E; elderly person Y gets lost at door lock B and the actual destination is door lock F; elderly person Z gets lost at door lock C and the actual destination is door lock G. However, the paths from door lock A to door lock E for X, from door lock B to door lock F for Y, and from door lock C to door lock G for Z all overlap. When door locks A, B, and C perform path planning (broadcasting path information to all door locks on the path), each door lock on the path considers itself still in the navigation state until it receives the "successful unlocking" message from the destination door lock. If a door lock L is on the path of an elderly person and receives another path planning, door lock L unicasts a "path conflict" message to the door lock responsible for the planning. For example, if the first door lock for planning is door lock A and no other door lock reports a "path conflict" at this time, door lock A defaults to choosing the thick green line and fills in the "thick green line" information when sending the path planning to all door locks on the path. After all door locks on the path receive it, they simultaneously display the "thick green line" on the light strip. The second door lock for planning is door lock B, which broadcasts the path information to all door locks on the path. At this time, door lock L is on the path planned by door lock A. After receiving the path planning sent by B, it unicasts a "path conflict" message to door lock B, including the current light strip color as "thick green line". After receiving the "path conflict" message, door lock B chooses to use the dotted blue line and resends the path planning to all door locks on the path, filling in the "dotted blue line" information. After all door locks on the path receive it, they simultaneously display the "dotted blue line" on the light strip. Door lock L also saves the path "thick green line" of door lock A and the path "dotted blue line" of door lock B. If there are other door locks besides door lock L that also detect a path conflict, they all send a "path conflict" message to door lock B. For the same information, door lock B only responds once (resends the path planning, filling in the "dotted blue line" information). When door lock B itself is on the path planned by door lock A, door lock B already has the path and "thick green line" information of door lock A, so it directly uses the "dotted blue line" information and broadcasts the planned path to all door locks on the path. The third door lock for planning is door lock C. Among them, door lock L reports a conflict message to door lock C, including the current light strip color states as "thick green line" and "dotted blue line", then door lock C chooses yellow. If multiple door locks report conflict messages, such as door lock L reporting including "thick green line" and "dotted blue line", and door lock J reporting a conflict message including "thick green line" and "thin yellow line", it means that the lost path guidance paths of 3 elderly people have overlapped simultaneously.Then the door lock C temporarily stops planning and tells the door lock that reported the conflict message, "If you receive the message indicating the end of planning, report it to me." After the door lock that reported the conflict message of "green and yellow" replies that the path planning is completed, the door lock C selects "yellow" for planning.

[0054] The door locks A, B, and C give voice prompts according to the names of different elderly people (e.g., for the elderly person X: "Dear X, your current location is Room 305 on the 3rd floor, and your room is Room 203 on the 2nd floor. Please follow the guidance of the green thick-line lighting path to reach your room." For the elderly person Y: "Dear Y, your current location is Room 307 on the 3rd floor, and your room is Room 402 on the 4th floor. Please follow the guidance of the blue dotted-line lighting path to reach your room."). The directional speakers at the cross-floor connection points broadcast the voice guidance information of one of the elderly people every 10 seconds.

[0055] 10: Further improvement (2): How to distinguish when multiple people go in the wrong direction When two or more elderly people go in the wrong direction simultaneously during the guiding process on the same floor, the red warning light strips may overlap, making it impossible to determine whether it is one of the elderly people who has gone in the wrong direction or multiple people who have gone in the wrong direction simultaneously, which may cause confusion and anxiety among the elderly. For example, when the elderly person X goes in the opposite direction, the door lock E discovers this situation through the camera. Since the door lock E is not on the planned path, at this time, the door lock E actively broadcasts a request for obtaining the lost navigation path to all the door locks on the same floor. The request contains the face feature vector of the elderly person X and the door lock ID of the door lock E. After receiving the request, the door lock A that initiates the navigation unicasts the navigation path to the door lock E. The door lock E combines the pre-stored floor topology data and supplements the wrong path to the navigation path, and at the same time broadcasts information to all the door locks on the path of the door lock A. At this time, the elderly person Y is also undergoing lost guiding and has also gone in the wrong direction. Due to the close distance, the red warning light strips of the two people actually overlap. Both X and Y think that the other person has gone in the wrong direction and continue to move forward in their wrong directions. At this time, the door lock E discovers through the camera that the elderly person Y has also gone in the wrong direction. The door lock E actively broadcasts a request for obtaining the lost navigation path to all the door locks on the same floor. The request contains the face feature vector of the elderly person Y and the door lock ID of the door lock E. After receiving the request, the door lock B that initiates the navigation unicasts the navigation path to the door lock E. At this time, E discovers that it is on the wrong path of the door lock A and then receives the navigation path of the door lock B, so it judges that the wrong paths have overlapped. At the same time, the door lock E combines the pre-stored floor topology data, supplements the wrong path to the navigation path, and broadcasts information to all the door locks on the path of the door lock B. On the basis of further improvement (1), the inner tubes of the guiding light strips of the elderly person X and the elderly person Y display different colors and states. After the door lock E and all the door locks on the navigation path supplement the wrong path information, the door lock E notifies the affiliated light strip. While lighting the inner tube of the red light strip, it also lights the inner tubes of the light strips of the correct paths of the elderly person X and the elderly person Y. At this time, combined with the voice prompt of the door lock E (such as for the elderly person X: "Dear X, your forward direction is incorrect. Please return to the path with the green thick-line light." For the elderly person Y: "Dear Y, your forward direction is incorrect. Please return to the path with the blue dotted-line light."), it is very easy for the elderly person X and the elderly person Y to find that the direction of their navigation path points backward in front of them, so as to quickly return to the correct forward direction.

[0056] Beneficial effects: The combination of multi-modal verification failure and behavior analysis makes the accuracy rate of lost recognition reach more than 90%. The dual auditory and visual guarantees of voice + guiding light strips strengthen the guiding effect, ensuring that elderly people with functional impairments can be accurately guided to their rooms after going into the wrong room.

[0057] It can be seen that according to the target door lock A detecting that the elderly person fails to verify continuously for multiple times and stays in front of the door for overtime, the lost score value is calculated through the lost score model. When the score value exceeds the threshold, it is determined to be in a lost state, and the lost determination result is obtained; according to the door lock A, the biometric characteristics of the elderly person are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received to obtain the target room positioning result; according to the floor topology data of the door lock A and the door lock B, the optimal guiding path is generated to obtain the path data including the sequence of passing door locks; according to the path data, the strip lights along the way are controlled to be lit in a preset mode, and the directional speaker is coordinated to play voice guidance; according to the door lock B detecting that the elderly person successfully unlocks or arrives, a reset instruction is sent to the path door locks, so that convenient, efficient and intelligent door lock guidance can be carried out, and accurate guidance and support for the elderly can be realized.

[0058] Another embodiment of the present invention provides a system for guiding an elderly person home by a door lock. Refer to Figure 3 , the system may include: A determination module 301 for determining lost behavior: According to the target door lock A detecting that the elderly person fails to verify continuously for multiple times and stays in front of the door for overtime, the lost score value is calculated through the lost score model. When the score value exceeds the threshold, it is determined to be in a lost state; wherein, the lost score model fuses the number of verification failures, the stay duration and the radar trajectory confidence to obtain the lost determination result; A matching module 302 for cross-room identity matching: According to the door lock A, the biometric characteristics of the elderly person are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received; wherein, the biometric characteristics include a face feature vector and a fingerprint hash value, to obtain the target room positioning result; A planning module 303 for dynamic path planning: According to the floor topology data of the door lock A and the door lock B, the optimal guiding path is generated; wherein, the same-floor path is directly planned, and the cross-floor path is transferred through the connecting point door lock, to obtain the path data including the sequence of passing door locks; An execution module 304 for multi-modal guidance execution: According to the path data, the strip lights along the way are controlled to be lit in a preset mode, and the directional speaker is coordinated to play voice guidance; wherein, the strip lights display direction arrows and the non-path segments remain extinguished to obtain the real-time guidance effect; A reset module 305 for end point confirmation and reset: According to the door lock B detecting that the elderly person successfully unlocks or arrives, a reset instruction is sent to the path door locks; wherein, the instruction triggers the strip lights to go out and the log to be uploaded to complete the guidance closed loop.

[0059] It can be seen that, based on the target door lock A detecting that the elderly person fails to verify continuously for multiple times and stays in front of the door for an excessive time, the lost score value is calculated through the lost score model. When the score value exceeds the threshold, it is determined as the lost state, and the lost determination result is obtained; according to the door lock A, the biometric features of the elderly person are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received to obtain the target room positioning result; according to the floor topology data of the door lock A and the door lock B, the optimal guiding path is generated to obtain the path data including the sequence of passing door locks; according to the path data, the strip lights along the way are controlled to be lit in a preset mode, and the directional speaker is coordinated to play voice guidance; according to the door lock B detecting that the elderly person successfully unlocks or arrives, a reset instruction is sent to the path door locks, so that convenient, efficient and intelligent door lock guidance can be carried out, and accurate guidance and support for the elderly can be realized.

[0060] An embodiment of the present invention further provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0061] Specifically, the above electronic device may further include a transmission device and an input / output device. Among them, the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0062] Specifically, in this embodiment, the above processor may be configured to execute the following steps through a computer program: S201, lost behavior determination: Based on the target door lock A detecting that the elderly person fails to verify continuously for multiple times and stays in front of the door for an excessive time, the lost score value is calculated through the lost score model. When the score value exceeds the threshold, it is determined as the lost state; among them, the lost score model fuses the number of verification failures, the stay duration and the radar trajectory confidence to obtain the lost determination result; S202, cross-room identity matching: According to the door lock A, the biometric features of the elderly person are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received; among them, the biometric features include a face feature vector and a fingerprint hash value, and the target room positioning result is obtained; S203, dynamic path planning: According to the floor topology data of the door lock A and the door lock B, the optimal guiding path is generated; among them, the path on the same floor is directly planned, and the path across floors is transferred through the connecting point door lock to obtain the path data including the sequence of passing door locks; S204, multi-modal guidance execution: According to the path data, the strip lights along the way are controlled to be lit in a preset mode, and the directional speaker is coordinated to play voice guidance; among them, the strip lights display direction arrows and the non-path segments remain extinguished to obtain the real-time guidance effect; S205, Endpoint Confirmation and Reset: Send a reset command to the path door lock based on the detection by door lock B that the elderly person has successfully unlocked or arrived; wherein, the command triggers the light strip to turn off and the log to be uploaded, completing the guidance closed-loop.

[0063] It can be seen that based on the detection by target door lock A that the elderly person has failed to verify continuously for multiple times and stayed in front of the door for an excessive time, the lost score value is calculated through the lost score model. When the score value exceeds the threshold, it is determined to be in a lost state, and the lost determination result is obtained; according to door lock A, the biometric features of the elderly person are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received to obtain the target room positioning result; according to the floor topology data of door lock A and door lock B, the optimal guidance path is generated to obtain the path data including the sequence of passing door locks; according to the path data, control the light strips along the way to light up according to the preset mode, and cooperate with the directional speaker to play voice guidance; according to the detection by door lock B that the elderly person has successfully unlocked or arrived, send a reset command to the path door lock, so that convenient, efficient and intelligent door lock guidance can be carried out, realizing precise guidance and support for the elderly.

[0064] The structure, features and effects of the present invention have been described in detail based on the embodiments shown in the drawings above. The above are only the preferred embodiments of the present invention, but the present invention is not limited to the scope of implementation shown in the drawings. Any changes made according to the concept of the present invention, or modified into equivalent embodiments with equivalent changes, still within the spirit covered by the specification and the drawings, should be within the protection scope of the present invention.

Claims

1. A method for a door lock to guide the elderly to go home, characterized in that, The method includes: Lost behavior determination: Based on the target door lock A detecting that the elderly person has failed to verify continuously for multiple times and has stayed in front of the door for an overtime, calculate the lost score value through a lost score model. When the score value exceeds the threshold, it is determined to be in a lost state; among them, the lost score model fuses the number of verification failures, the stay duration, and the radar trajectory confidence to obtain the lost determination result; Cross-room identity matching: Encrypt and broadcast the biometric characteristics of the elderly person to all door locks according to door lock A, and receive the confirmation information returned by the successfully matched door lock B; among them, the biometric characteristics include a face feature vector and a fingerprint hash value to obtain the target room positioning result; Dynamic path planning: Generate an optimal guiding path according to the floor topology data of door lock A and door lock B; among them, the path on the same floor is directly planned, and the cross-floor path is transferred through the connecting point door lock to obtain the path data including the sequence of passing door locks; Multi-modal guidance execution: According to the path data, control the strip lights along the way to light up in a preset mode, and cooperate with the directional speaker to play voice guidance; among them, the strip lights display direction arrows and the non-path segments remain extinguished to obtain the real-time guidance effect; Endpoint confirmation and reset: According to door lock B detecting that the elderly person has successfully unlocked or arrived, send a reset instruction to the path door locks; among them, the instruction triggers the strip lights to go out and the log to be uploaded to complete the guidance closed-loop.

2. The method according to claim 1, wherein The lost behavior determination includes: Multi-modal verification failure detection: Based on door lock A detecting that the elderly person has failed in face / fingerprint / password verification continuously for 3 times, record the number of failure times N; among them, the interval between each verification is less than 5 seconds to obtain the verification failure record; Behavior feature assisted determination: According to the millimeter-wave radar monitoring that the elderly person has stayed in front of the door for more than 30 seconds without leaving, record the stay duration T; among them, the radar synchronously generates the trajectory confidence S to obtain the behavior monitoring data; Lost score calculation: Calculate the lost score value according to the formula P =α⋅N+β⋅T+γ⋅S; among them, α, β, and γ are weight coefficients, N is the number of verification failures, T is the stay duration, S is the radar trajectory confidence. When P>1, trigger the lost determination to obtain the lost state confirmation signal.

3. The method according to claim 2, wherein The cross-room identity matching includes: Feature broadcast: Encrypt and broadcast a packet containing a face feature vector and a fingerprint hash value according to door lock A, and send it to all door locks through the local area network; among them, the broadcast packet is appended with the ID of door lock A to obtain the feature diffusion result; Feature comparison: According to door lock B receiving the broadcast packet, compare its face feature with the local registered feature library; among them, the cosine similarity algorithm is used, and when the similarity exceeds 95%, it is determined to be a match to obtain the matching result; Target confirmation: According to door lock B, unicast a confirmation message containing its own ID to door lock A; among them, the message is encrypted by AES to obtain the precise positioning of the target room.

4. The method according to claim 3, characterized in that, The dynamic path planning includes: Floor topology analysis: Retrieve the locally stored floor topology map according to door lock A to determine whether door lock B is on the same floor; among them, the topology map includes door lock positions and connection point information to obtain the floor relationship determination; Path generation: Based on the same - floor determination result, if on the same floor, a straight - line path is planned; if across floors, stair / elevator connection - point door locks are added. Among them, when the floor interval is 1 floor, the path priority of the stairs > the path priority of the elevator; when the floor interval is greater than or equal to 2 floors, the path priority of the elevator > the path priority of the stairs, obtaining an optimized path sequence. Path binding: Based on the generated path, bind the elderly's face features with the passing door - lock IDs. Among them, the bound data is saved to the cache of each door lock, obtaining the path persistence result.

5. The method according to claim 4, wherein The multi - modal guidance execution includes: Sending a lighting instruction to the corresponding light strip according to the path door lock. Among them, the instruction includes color and arrow direction, and the response delay of the light strip < 200ms, obtaining a visual guidance effect. Voice collaboration: Triggering the directional speaker to play guiding voices according to the connection - point door lock. Among them, the voice content includes the elderly's name and turning instructions, and the broadcast interval is 10 seconds, obtaining an auditory guidance effect. Exception handling: When the camera detects that the elderly goes in the wrong direction, controlling the light strip in the opposite direction to turn red and playing error - correcting voices. Among them, the inner tube of the light strip on the correct path is synchronously activated, obtaining a deviation - correction guidance effect.

6. The method according to claim 5, wherein The end - point confirmation and reset include: Successful arrival determination: When door lock B detects that the elderly passes the verification or the camera confirms that they enter the room, generating an unlocking - successful event. Among them, the event includes the elderly's face features, obtaining an end - point confirmation signal. System reset: Broadcasting a reset instruction to the path door lock according to door lock B. Among them, the instruction triggers the light strip to go out, clears the cached feature data, and uploads the log, obtaining a system - reset result. Emergency handling: When the event of not arriving within the time limit occurs, door lock B activates a red - flashing light strip and notifies the caregiver. Among them, continuous guidance is provided until manual intervention, obtaining an emergency - response result.

7. The method according to claim 6, characterized in that The method further includes: When door lock A detects that the elderly X gets lost and generates a first path plan, broadcasting path data including the first light - strip identifier to the passing door locks, obtaining a first activated path. When door lock B detects that the elderly Y gets lost and broadcasts path - planning information to the path door locks, and door lock L is on the first activated path and receives the path - planning from door lock B, feedbacking a path - conflict notification to door lock B, obtaining a conflict - detection result. Based on the conflict - detection result, door lock B selects a second light - strip identifier to re - broadcast the path data, and the passing door locks synchronously update the light - strip display mode, obtaining a differentiated path. When door lock C detects that the elderly Z gets lost and receives the conflict notification from door lock L, and after planning a third light - strip path, before door lock J prepares to plan a path, it receives a conflict notification including three light - strip identifiers, and waits for any path to end and then activates the path with the third light - strip identifier, obtaining a time - division multiplexing guidance effect.

8. The method according to claim 7, wherein The method further includes: When non - path door lock E detects through the camera that the elderly X walks in the reverse direction, broadcasting a path - acquisition request including their face features, obtaining a reverse - walking event. When the initiating door lock A responds to the request and returns the original path data, door lock E combines the floor topology to generate a supplementary path including the error section, obtaining enhanced path information. When door lock E synchronously detects that the elderly Y walks in the wrong direction in the same direction, activating the correct - direction indication corresponding to the second light - strip identifier in the supplementary path, obtaining a dual - path deviation - correction interface. Based on the light strip system that simultaneously displays a red error warning and a two-color correct path identifier, combined with voice differentiation prompts, a multi-person collaborative deviation correction effect is obtained.

9. A system for a door lock to guide the elderly home, characterized in that, The system includes: A determination module for determining lost behavior: According to the target door lock A detecting that the elderly person fails to verify continuously for multiple times and stays in front of the door for an overtime, the lost score value is calculated through a lost score model. When the score value exceeds the threshold, it is determined to be in a lost state; among them, the lost score model fuses the number of verification failures, the stay duration, and the radar trajectory confidence to obtain a lost determination result; A matching module for cross-room identity matching: According to door lock A, the biometric characteristics of the elderly are encrypted and broadcast to all door locks, and the confirmation information returned by the successfully matched door lock B is received; among them, the biometric characteristics include a face feature vector and a fingerprint hash value to obtain the target room positioning result; A planning module for dynamic path planning: According to the floor topology data of door lock A and door lock B, an optimal guiding path is generated; among them, the same-floor path is directly planned, and the cross-floor path is transferred through the connecting point door lock to obtain path data including the sequence of passing door locks; An execution module for multi-modal guidance execution: According to the path data, control the light strips along the way to be lit in a preset mode, and cooperate with the directional speaker to play voice guidance; among them, the light strip displays a direction arrow and the non-path segments remain extinguished to obtain a real-time guidance effect; A reset module for end point confirmation and reset: According to door lock B detecting that the elderly person has successfully unlocked or arrived, send a reset instruction to the path door locks; among them, the instruction triggers the light strip to go out and the log to be uploaded to complete the guidance closed-loop.

10. An electronic device, comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is set to run the computer program to execute the method according to any one of claims 1-8.

Images