An event value-driven door lock recording method and system

Abstract

The application discloses a door lock video recording method and system based on event value driving, and the method comprises the following steps: acquiring the video stream address of a second door lock through door relationship information according to the event that a first door lock detects the appearance of a person, and synchronously buffering the video streams of the two door locks; judging the event type according to the video analysis result, saving the complete video streams of the two door locks for important events, and saving only panoramic images for non-important events; according to the time difference of face detection of the two door locks and the remaining storage space, negotiating and determining the single door lock to store the video streams of the two door locks; and according to the user query request, directly responding locally or acquiring video data through cooperation with the door to complete the query display. By using the embodiment of the application, efficient video acquisition and storage can be realized through event driving and cooperation with the door lock, the completeness of the security video recording is improved, the storage space is saved, and more efficient and reliable intelligent query response is supported.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of intelligent security and protection, and particularly relates to a door lock video recording method and system based on event value driving. BACKGROUND

[0002] In recent years, intelligent door lock systems have become increasingly popular, and door lock devices with video recording functions are widely used in home and community security and protection. In the prior art, door locks usually work independently, and only record videos locally based on their own triggering events, lacking a cooperative mechanism with doors or surrounding devices, resulting in fragmented event records and single perspectives. Especially in scenes such as personnel access and abnormal detection, a single device cannot fully capture the entire process of an event, and there are problems such as missing key information, storage redundancy, and low query efficiency. In addition, the existing solutions lack the ability to distinguish the importance of events, and cannot adaptively adjust the storage strategy according to the event value, resulting in waste of storage resources and insufficient preservation of effective information. SUMMARY

[0003] The purpose of the present application is to provide a door lock video recording method and system based on event value driving, to solve the problems in the prior art, and to realize efficient video acquisition and storage through event driving and cooperation with door locks, improve the completeness of security and protection video recording, save storage space, and support more efficient and reliable intelligent query response.

[0004] One embodiment of the present application provides a door lock video recording method based on event value driving, which comprises:

[0005] Cooperative video stream acquisition: according to the event of the first door lock detecting the presence of a person, the video stream address of the second door lock is obtained through the door relationship information, and the video streams of both door locks are synchronously cached;

[0006] Event value evaluation and processing: according to the video analysis result, the event type is judged, the complete video stream of both parties is saved for important events, and only the panoramic image is saved for non-important events;

[0007] Storage responsibility negotiation: according to the face detection time difference and the remaining storage space of both door locks, the storage of the video stream of both parties by a single door lock is determined through negotiation;

[0008] Intelligent query response: according to the user query request, the video data is obtained directly locally or through door cooperation to complete the query display.

[0009] Optionally, the cooperative video stream acquisition comprises:

[0010] Personnel detection trigger: according to the event of the first door lock video analysis module detecting the presence of a person, a door lock event record is created and the starting time is recorded;

[0011] Video request to the door: according to the pre-stored door relationship information, send a video stream request message to the second door lock, and obtain the video stream address of the second door lock;

[0012] Video stream synchronization buffer: according to the obtained video stream address, synchronously capture and buffer the video stream data of the first door lock and the second door lock;

[0013] Event state maintenance: according to the personnel activity state, real-time update the event type field of the door lock event record.

[0014] Optionally, the event value evaluation and processing includes:

[0015] Event type judgment: according to the video analysis result, the event type is judged to be an important event or a non-important event;

[0016] Important event processing: for the important events of unlocking and opening the door or personnel wandering, the complete video stream data of the two door locks is saved;

[0017] Non-important event processing: for the non-important events that the door lock video scene changes or the personnel quickly passes through, when the duration is less than the threshold, the video stream is deleted, and only the panoramic image is saved;

[0018] Panoramic image extraction: from the video stream of the first door lock and the second door lock in the last event record, one panoramic image is extracted from the time when the personnel leaves, which is used for scene recording in non-important period.

[0019] Optionally, the storage responsibility negotiation includes:

[0020] Face detection time record: according to the face detection module of the two door locks, the time stamp of the first detected face is recorded;

[0021] Time difference comparison: calculate the time difference of face detection of the two door locks, and determine the priority storage side according to the positive and negative of the time difference;

[0022] Storage space comparison: when the time difference is zero, compare the remaining storage space of the two door locks to determine the storage responsibility side;

[0023] Storage negotiation execution: according to the negotiation result, the video stream of the two sides is stored by the determined responsibility side, and the other side only stores the event index.

[0024] Optionally, the intelligent query response includes:

[0025] Local query processing: for the events that store complete video locally, the video stream is directly extracted to respond to the query request;

[0026] Remote query processing: for the events that only store event index, send a video request message to the door lock opposite to the door;

[0027] Video stream acquisition and display: receiving video stream data returned by the door lock, locally caching and displaying to the user;

[0028] Cache optimization processing: determining whether to locally cache the video stream returned by the door lock according to the query frequency, to optimize the subsequent query efficiency.

[0029] Another embodiment of the present application provides a door lock video recording system based on event value driving, the system comprising:

[0030] The acquisition module is used for video stream acquisition: according to the event of detecting the presence of personnel by the first door lock, obtaining the video stream address of the second door lock through the door relationship information, and synchronously caching the video streams of the two door locks;

[0031] The evaluation module is used for event value evaluation and processing: judging the event type according to the video analysis result, saving the complete video streams of both sides for important events, and saving only the panoramic image for non-important events;

[0032] The negotiation module is used for storage responsibility negotiation: determining the storage of the video streams of both sides by a single door lock according to the time difference of face detection of the two door locks and the remaining storage space;

[0033] The response module is used for intelligent query response: directly responding locally or obtaining video data through door cooperation according to the user query request, and completing the query display.

[0034] Another embodiment of the present application provides a storage medium, which stores a computer program, wherein the computer program is set to execute the method described in any of the above embodiments when running.

[0035] Another embodiment of the present application provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is set to run the computer program to execute the method described in any of the above embodiments.

[0036] Compared with the prior art, the door lock video recording method based on event value driving provided by the present application can realize efficient video acquisition and storage through event driving and door lock cooperation, improve the completeness of security video recording, save storage space, and support more efficient and reliable intelligent query response. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 A hardware structure block diagram of a computer terminal of an event value driving-based door lock video recording method provided for an embodiment of the present application is shown in the figure.

[0038] Figure 2 A flowchart of an event value driving-based door lock video recording method provided for an embodiment of the present application is shown in the figure.

[0039] Figure 3 A structure diagram of an event value driving-based door lock video recording system provided for an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0040] The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and cannot be explained as a limitation of the present application.

[0041] The present application belongs to the technical field of intelligent security and protection, and particularly relates to a cooperative video recording method and system of an intelligent door lock, and specifically relates to a technical scheme in which a door-to-door intelligent door lock directly transmits a video stream, dynamically adjusts a storage strategy, and distributes storage responsibilities based on face detection.

[0042] 1. Prior art and defects:

[0043] With the development of smart homes, the camera function of intelligent door locks has become an important means of home security, but the existing technology has the following key defects:

[0044] (1) Video transmission relies on the cloud, and privacy and efficiency are insufficient: in the existing technology (such as CN120091097A), the video of a neighbor's door lock needs to be transferred through the cloud, and the video data is stored in the cloud, which poses a risk of privacy leakage; at the same time, the cloud transfer increases the data transmission cost and complexity.

[0045] (2) Storage strategy is fixed, and resources are wasted seriously: existing door locks use full-recording or fixed-interval recording (such as CN119893017A) for all events (such as abnormal wandering and personnel passing by), and low-value events (such as multiple delivery personnel passing by every day) occupy more than 70% of the storage space, resulting in insufficient utilization of storage resources.

[0046] (3) Lack of neighbor cooperation, and storage responsibilities are not clear: in a door-to-door living scenario, a single door lock camera has a limited field of view (such as narrow areas in a corridor that can form blind spots), but the existing technology has no direct video sharing mechanism; and the same event is repeatedly stored by multiple door locks, and the responsibilities are not distributed based on face detection, resulting in incomplete storage of key information (such as facial features).

[0047] 2. Necessity of the invention:

[0048] In view of the above defects, there is an urgent need for a technical solution that can realize door-to-door video direct transmission, dynamic storage and clear storage responsibility, so as to improve storage efficiency, strengthen neighborhood cooperation and protect user privacy.

[0049] The embodiment of the application first provides an event value driven door lock video recording method, which can be applied to an electronic device such as a computer terminal, specifically a general computer, etc.

[0050] The following will be described in detail taking a computer terminal as an example. Figure 1 A hardware structure block diagram of a computer terminal of an event value driven door lock video recording method provided by the embodiment of the application is shown in FIG. Figure 1 As shown in the figure, the computer device includes a processor, a memory and a network interface connected through a system bus, wherein the memory can include a non-volatile storage medium and an internal memory.

[0051] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions which, when executed, can cause the processor to execute any event value driven door lock video recording method.

[0052] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0053] The internal memory provides an environment for the running of the computer program in the non-volatile storage medium, which, when executed by the processor, can cause the processor to execute any event value driven door lock video recording method.

[0054] 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 the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0055] It should be appreciated that the processor can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0056] Referring to Figure 2 Embodiments of the present application provide a door lock video recording method based on event value driving, which can include the following steps:

[0057] S201, cooperative video stream acquisition: according to the event of detecting the appearance of personnel by the first door lock, the video stream address of the second door lock is obtained by acquiring the door relationship information, and the video streams of the two door locks are synchronously cached; specifically, the cooperative video stream acquisition includes:

[0058] Personnel detection trigger: according to the event of detecting the appearance of personnel by the first door lock video analysis module, a door lock event record is created and the starting time is recorded;

[0059] Door video request: according to the pre-stored door relationship information, a video stream request message is sent to the second door lock to obtain the video stream address of the second door lock;

[0060] Video stream synchronous caching: according to the obtained video stream address, the video stream data of the first door lock and the second door lock is synchronously captured and cached;

[0061] Event state maintenance: according to the personnel activity state, the event type field of the door lock event record is updated in real time.

[0062] S202, event value evaluation and processing: according to the video analysis result, the event type is judged, the complete video stream of both sides is saved for important events, and only panoramic images are saved for non-important events; specifically, the event value evaluation and processing includes:

[0063] Event type judgment: according to the video analysis result, the event type is judged as important event or non-important event;

[0064] Important event processing: for important events such as unlocking and opening the door or personnel wandering, the complete video stream data of both door locks is saved;

[0065] Non-important event processing: for the non-important event that the door lock video scene changes (due to the change of light, etc.) or the person quickly passes, when the duration is less than the threshold, delete the video stream, and only save the panoramic image;

[0066] Panoramic image extraction: extract the key frame from the first door lock and the second door lock video stream after the person leaves the time in the last event record to generate each 1 panoramic image, which is used for non-important period scene record.

[0067] S203, storage responsibility negotiation: according to the time difference of face detection of both door locks and the remaining storage space, the storage responsibility of single door lock for storing video streams of both door locks is determined; specifically, the storage responsibility negotiation includes:

[0068] Face detection time record: according to the face detection module of both door locks, the time stamp of the first detection of the face is recorded;

[0069] Time difference comparison: calculate the time difference of face detection of both door locks, and determine the priority storage side according to the positive and negative of the time difference;

[0070] Storage space comparison: when the time difference is zero, compare the remaining storage space of both door locks to determine the storage responsibility side;

[0071] Storage negotiation execution: according to the negotiation result, the video streams of both door locks are stored by the determined responsibility side, and the other side only stores the event index.

[0072] S204, intelligent query response: according to the user query request, the local directly responds or obtains video data through the door cooperation to complete the query display. Specifically, the intelligent query response includes:

[0073] Local query processing: for the event that the complete video is stored locally, the video stream is directly extracted to respond to the query request;

[0074] Remote query processing: for the event that only stores the event index, a video request message is sent to the door lock;

[0075] Video stream acquisition and display: receive the video stream data returned by the door lock, locally cache and display to the user;

[0076] Cache optimization processing: according to the query frequency, it is determined whether to locally cache the video stream returned by the door lock to optimize the subsequent query efficiency.

[0077] In practical application, the present application aims to solve the three problems of the prior art: video transmission depends on the cloud; low-value event resource waste caused by fixed storage strategy; monitoring blind spot and incomplete information caused by lack of neighborhood cooperation and unclear storage responsibility. A specific technical solution is as follows:

[0078] 1. Core idea:

[0079] The door lock A and the door lock B belong to the door locks of two big doors of a door, and all the community door locks are networked, and the door locks of the door have the ID and the IP address of each other.

[0080] When the door lock A detects a person entering or leaving or a person lingering at the door through a video, a message is sent to the door lock B to immediately start sending a video stream to the door lock A, and the door lock A starts to synchronously store the video images shot by itself and the video images shot by the door lock B; when the door lock A detects that the person leaves through the video, a message is sent to the door lock B to end the sending of the video stream to itself, and the door lock A also ends the storage of the video images, so as to realize the all-around recording of important events.

[0081] In order to ensure the all-weather scene recording, the door lock also needs to record the non-important scenes, and in order to save the storage space, the application proposes a recording method based on the value event, that is, although the scene outside the door lock changes due to the change of light and the like, the non-important event is ignored; therefore, the door lock A records a picture at the beginning (for example, 9:00), and no important event occurs from 9:00 to 12:00, and only the picture changes due to the change of light, so no video image is recorded in this period of time, when the user clicks the picture at 11:00 from the client, the door lock feeds back the picture at 9:00 to the client, so as to greatly save the storage space. For the non-important event, for example, a person quickly passes by at 10:30, the door lock detects the person, and then starts to record and informs the door to send the video to record, when the person leaves, the recording is stopped, and if the time is shorter than a threshold (for example, 2 minutes), the door lock actively deletes the video recording in this period of time, when the user clicks the picture at 10:31 from the client, the picture at 9:00 is still fed back, so as to greatly save the storage space.

[0082] If the storage space is tight, and the client needs to further save the space, the door lock A and the door lock B only keep one copy of the video. When the door lock A detects a person appearing outside, a message is sent to the door lock B to negotiate, and the party that detects the face first is responsible for the storage, if neither detects the face, the party with more balance of the storage space is responsible for the storage; when the user wants to play back the video of the door lock A, the two door locks cooperate to play back.

[0083] 2, the complete flow of technical implementation is as follows:

[0084] (1) preset conditions:

[0085] The smart door lock is installed on the entrance door of a household, and the door lock has an independent door lock ID number and door lock IP address; all the door locks in the community are networked; the door lock saves "door relationship information" including door lock ID number and door lock IP address of the opposite door; the door lock stores "authorized personnel information list" including personnel ID number, personnel name, face image / feature vector, fingerprint, password and the like; the door lock supports unlocking and opening the door through comparison of face recognition, password and fingerprint and the like; and the door lock supports video intelligent detection based on face detection and personnel loitering of video images.

[0086] Door relationship explanation: when the door lock is first networked, the door lock is initialized and configured (for example, remote configuration through a remote management terminal of the door lock, or on-site initialization configuration using a door lock management terminal on the door lock) by a platform management personnel according to the floor plan of the position of each door lock, the room ID number, door lock ID number, door lock IP and the like of each door lock stored on the platform, the door relationship of the door lock, the door lock ID number and IP address of the opposite door and the like.

[0087] (2) When a person is detected in front of the door lock, the video streams of both sides are acquired and cached: when door lock A detects a person in front of the door lock through the video (A) of the door lock, the video (B) of door lock B is acquired through the IP address of the opposite door, and the video streams of both sides are cached in the door lock event.

[0088] At t1, when door lock A detects a person in front of the door lock through video intelligent analysis (at this time, video intelligent analysis is performed through the camera of door lock A to shoot video (A)), it immediately starts caching the video stream (A), establishes a door lock event record, and includes door lock event ID (the numbering method can adopt: door lock ID (A) + time stamp, recorded as door lock event ID (A+t1)), event type / empty (the types include unlocking and opening the door, personnel loitering, personnel passing by, and no one, at this time, it is empty because the event type has not been detected and judged), video clip (A) / empty, video clip (B) / empty, start time / t1, end time / empty and the like. The door lock IP (B) of door lock B is immediately queried from the "door relationship information" and a "request video" message is sent to door lock B, including door lock event ID, door lock ID (A) and the like. Door lock B receives and replies to the "request video" message, including video (B) address (for example, taking the streaming media protocol RTSP as an example, the current video stream address of door lock B is: rtsp: / / {IP(B)}:8554 / streamB) and the like. After receiving the video (B) address of B, door lock A captures and caches video (B) according to the video address.

[0089] According to the same method above, at t1 (at this time, since the door locks A and B are in a pair relationship, it is assumed that the appearance of the same person can be detected at the same time), the door lock B starts to cache the video (B), acquires and starts to cache the video (A), and records the door lock event ID (B+t1).

[0090] (3) When unlocking the door or the person at the door is detected to be loitering, the door lock event video is recorded in all directions: when the door lock A detects that the person unlocks the door or that the person is loitering in front of the door lock through the video (A), the cached two-way video is converted into the door lock event and saved, realizing all-round recording of important events.

[0091] At t2, the door lock A detects that the person at the door unlocks the door (such as face recognition, etc.) and enters the house and closes the door lock, or detects that the person is loitering (such as through intelligent video analysis), and the person disappears from the video, then at t2+30 seconds (the delay of 30 seconds is to intercept and save the panoramic view after the disappearance of the person in the scene) the door lock A immediately stops the caching of the video (A) and the video (B), and saves the cached video clips (A) and video clips (B) to the door lock event, including the door lock event ID (A+t1), event type / (unlocking the door / person loitering), video clip (A), video clip (B), start time / t1, end time / t2+30 seconds, etc., and the door lock event is automatically deleted after the expiration of the storage period (the storage period can be set as needed, such as 15 days). After stopping the caching of the video (B), the door lock A immediately sends a "stop video" message to B, including the door lock event ID (A+t1), the door lock ID (A), and the closing instruction (such as TEARDOWN). After receiving the "stop video", the door lock B stops sending the video stream (B).

[0092] According to the same method above, at t2 (at this time, since the door locks A and B are in a pair relationship, it is assumed that the departure of the same person can be detected at the same time), the door lock B notifies the door lock A to stop sending the video stream (A), and records the door lock event, including the door lock event ID (B+t1), event type / (unlocking the door / person loitering), video clip (A), video clip (B), start time / t1, end time / t2+30 seconds, etc. After receiving the "stop video", the door lock A stops sending the video stream (A).

[0093] Special case: if only the door lock A has an unlocking door during this period (at this time, the door lock B does not have an unlocking door), then the door lock A records the door lock event according to the method above (at this time, the door lock event ID (A+t1)); and the door lock B does not detect an unlocking door or a person loitering, then the door lock B is processed according to the method of step (4).

[0094] Description: There are several cases as follows: If the person in front of the door lock unlocks the door at this time, the door lock A saves the video clips of both parties; If the person in front of the door lock unlocks the door at this time, the door lock A, the door lock B (at this time, there may be many people, both entering the door lock A and entering the door lock B), the door lock A, the door lock B saves the video clips of both parties; If the door lock A has unlocked the door, but the door lock B has not unlocked the door, and the video intelligent analysis of the door lock B has not judged as abnormal behavior such as personnel wandering, the door lock B records this event as a non-important event according to the method of step (4).

[0095] (4) When the door lock detects that the person in front of the door lock leaves, extract and save the initial panoramic image: When the door lock A detects that the person leaves the video scene, extract one panoramic image (A), panoramic image (B) from the video clip (A), video clip (B) respectively, and save it to the newly established door lock event record, until the detection of the appearance of the person, realize the all-weather scene recording of the non-important scene in the form of panoramic image, can greatly save the storage space.

[0096] At t2, when the person leaves, the door lock A extracts one video screenshot panoramic image (A), panoramic image (B) at t2+30 seconds from the video clip (A), video clip (B) respectively and saves it to the newly established door lock event record, containing door lock event ID (A+t2), event type / no one, panoramic image (A), panoramic image (B), start time / t2, end time / empty, etc.

[0097] The door lock A continues to carry out video intelligent analysis based on the video stream (A) until it detects the appearance of the person at t3. The door lock A calculates and checks whether the length from the appearance of the person to the departure of the person (assuming that the person appears at 10:30 and the person leaves at 10:31, the calculation length is 1 minute) exceeds the threshold value (for example, the threshold value can be set to 2 minutes), if it does not exceed the threshold value, delete the cached door lock video of both parties, do not record the door lock event of the appearance process of the person. If it exceeds the threshold value, the door lock A updates the saved door lock event ID (A+t2), containing door lock event ID (A+t2), event type / no one, panoramic image (A), panoramic image (B), start time / t2, end time / t3, etc.

[0098] According to the same method above, the door lock B records the door lock event of the no one period from t2 to t3, containing door lock event ID (B+t2), event type / no one, panoramic image (A), panoramic image (B), start time / t2, end time / t3, etc.

[0099] Recording and query instructions: When the last door lock event ends (person leaves), if it is 9:00, capture and store 1 panoramic picture of both door locks at this time, which are panoramic picture (A) and panoramic picture (B) respectively. Subsequently, until 12:00, no important event (i.e. no one appears) occurs, only panoramic pictures are saved, and no additional video images are recorded. When the user queries the door lock event through the APP or directly on the door lock display screen, if the user clicks to query the record at 11:00, the panoramic picture (A) and panoramic picture (B) at 9:00 are displayed to the user.

[0100] Video intelligent analysis instructions for non-important events: There are two cases: the first is when the video scene changes due to changes in lighting and other reasons, but no important event occurs (i.e. no one appears), so the video scene change can be ignored; the second is when a person quickly passes by, such as at 10:30, someone quickly passes by the door lock, the door lock detects the appearance of the person according to step (2), obtains the video of the opposite door lock, and records the cache of the video of both door locks. If the person leaves until 10:31 (during which no one is detected to linger or unlock the door event), the cache of the video of both door locks is stopped, the door lock performs importance judgment, checks whether the time from the appearance of the person to the departure of the person (at this time, the person appears at 10:30 and the person leaves at 10:31, the time is 1 minute) exceeds the threshold value (for example, the threshold value can be set to 2 minutes), if it does not exceed the threshold value, the cached video of both door locks is deleted, and the door lock event of the person appearing process is not recorded. If it exceeds the threshold value, the door lock event is recorded. If the user queries the door lock event through the APP or directly on the door lock display screen at this time, if the user clicks to query the record at 10:31, the panoramic picture (A) and panoramic picture (B) at 9:00 are displayed to the user.

[0101] (5) Further improvement, to solve the problem of tight storage space of the door lock, the door locks A and B negotiate to keep the video of only one party that first detects the face: for step (3), the door lock A first detects the face, sends a negotiation message to the door lock B, and determines by negotiation of both parties that the door lock A saves the door lock event containing the video of both parties.

[0102] In step (3), if the door locks A and B record the video clips of both parties for the same person in front of the door lock, it is a repeated record when viewed globally, especially when the remaining storage space of the door lock is insufficient, which will affect the storage of subsequent door lock events. In order to solve this problem, the optimization is as follows:

[0103] Scenario backtracking: for the scenario of step (2), (3), at t1, both door locks A and B detect the presence of a person, request and obtain the video stream of the other, and start the caching of video (A) and video (B). At t2, the person in front of the door lock disappears from the video (at this time, an important event occurs: the single lock or double lock is unlocked and opened, or the person lingers, or the person passes by for a long time exceeding the threshold).

[0104] At t1, after door lock A detects the presence of a person in front of the door lock, it simultaneously starts face detection (for example, using a deep learning-based video face detection algorithm). When at t1+δ1 (at this time, t1+δ1 is a certain time between t1 and t2), door lock A first detects the face of the person in front of the door lock, caches the "detected face record" containing the door lock event ID (A+t1), face detection time / (t1+δ1), etc. At the same time, door lock A immediately sends a "detected face" negotiation message to B, containing the door lock event ID (A+t1), face detection time / (t1+δ1), etc. Door lock B receives and caches the "detected face" negotiation message sent by A.

[0105] At t1+δ2 (at this time, t1+δ2 is a certain time between t1 and t2), if door lock B first detects the face of the person in front of the door lock, it caches the detected face record containing the door lock event ID (B+t1), face detection time / (t1+δ2), etc. And sends a "detected face" negotiation message to A, containing the door lock event ID (A+t1), face detection time / (t1+δ2), etc. Door lock A receives and caches the "detected face" negotiation message sent by B.

[0106] At t2, when door lock A detects that the person unlocks the door, or the person lingers, or the person passes by for a long time exceeding the threshold, door lock A delays for 30 seconds and then starts comparing the difference Δt between the "detected face record" in the cache and the face detection time in the "detected face" negotiation message sent by B (at this time, Δt=(t1+δ2)-(t1+δ1)). According to the value of Δt, 3 different processes are performed:

[0107] 1) If At > 0 (at this time, it means that the time when the door lock B detects the face for the first time lags behind the time when the door lock A detects the face), the door lock A records the complete door lock event, including storing the video clips of both door locks (video clip (A), video clip (B)); the door lock A sends the message "I detect the face first, I will store" to B, including the door lock event ID (A + t1), the start time / t1, the end time / t2, etc.; at t2, the door lock B also calculates the result At > 0 according to the same method, then delays for 30 seconds (i.e. t2 + 30 seconds) to wait for receiving the negotiation message of A, when receiving the message "I detect the face first, I will store" sent by the door lock A to B, replies "OK", and stores the door lock event ID (B + t1), the start time / t1, the end time / t2, etc. in the door lock event ID (B + t1) of the door lock event of the door lock B (at this time, the door lock event message recorded in the door lock event ID (B + t1) of the door lock B contains the index of the door lock event of the door lock A).

[0108] Explanation: The mechanism of sending negotiation message adopts that who detects the face first initiatively sends the negotiation message, for example, the door lock A detects the face first, then A sends the negotiation message, and the door lock B delays for 30 seconds to wait for the negotiation message of B after calculating and judging At.

[0109] 2) If At < 0 (at this time, it means that the time when the door lock B detects the face for the first time lags behind the time when the door lock A detects the face), the door lock A delays for 30 seconds (i.e. t2 + 30 seconds) to wait for receiving the negotiation message of B. At t2, the door lock B also calculates the result At < 0 according to the same method, then the door lock B records the complete door lock event, including storing the video clips of both door locks (video clip (A), video clip (B)). At the same time, the door lock B sends the message "I detect the face first, I will store" to A, including the door lock event ID (B + t1), the start time / t1, the end time / t2, etc.; after receiving the message "I detect the face first, I will store" sent by the door lock B to A, the door lock A replies "OK", and stores the door lock event ID (B + t1), the start time / t1, the end time / t2, etc. in the door lock event ID (A + t1) of the door lock event of the door lock A.

[0110] 3) If△t=0 (at this time, the door lock B detects the face for the first time, the same as the door lock A, or both do not detect the face), the door lock A extracts the current remaining storage capacity of the lock and immediately sends the "storage negotiation" message to the door lock B, containing the door lock event ID (A+ t1), the remaining storage capacity / △C_A, and then waits to receive the "storage negotiation" message sent by B; When receiving the "storage negotiation" message sent by B (at this time, according to the same method, the door lock B completes the extraction of the current remaining storage capacity of the lock and immediately sends the "storage negotiation" message to the door lock A, containing the door lock event ID (B+ t1), the remaining storage capacity / △C_B, and then waits to receive the "storage negotiation" message sent by A), calculate the difference△C between△C_A and△C_B (that is,△C=△C_A -△C_B), when△C≥0 (at this time, it means that the remaining storage space of the door lock A is larger than that of B), the door lock A performs the storage of the door lock event ID (A+ t1) containing the video clips of the two door locks (video clip (A), video clip (B)), and sends the "I will store" message, containing the door lock event ID (A+ t1), the start time / t1, the end time / t2. The door lock B receives the "I will store" message sent by A, replies OK and saves the door lock event ID (B+ t1) in the door lock event ID (A+ t1) as an index; When△C<0, the door lock A sends the "you store" message to B, containing the door lock event ID (A+ t1), etc., and records the door lock event ID (B+ t1) number in the "I will store" message sent by the door lock B in the door lock event ID (A+ t1).

[0111] According to who first detects the face and the size of the remaining storage space, the door lock is judged according to the following judgment rules, which are as follows: first, the party who first detects the face is responsible for storage; if the detection time of the face is the same / or both are not seen, then the party with more remaining storage space is responsible for storage; if the number of detected faces is the same / or both are not seen, and the remaining storage space is also the same, then the door lock A is forced to store. The specific criteria are summarized in Table 1 as follows:

[0112] Table 1

[0113]

[0114] (6) Door lock event query: directly query when the lock stores complete door lock event; request assistance query when the lock does not store complete door lock event.

[0115] The present process takes the following scenario as an example: door lock A saves door lock event ID (A+t1) information containing both door lock video clips (video clip (A), video clip (B)); door lock B only saves door lock event ID (B+t1) containing door lock event ID (A+t1) number, which does not contain both door lock video clips.

[0116] When the owner of door lock A inquires about the door lock event through the mobile phone APP (the mobile phone is installed with the door lock APP associated with door lock A) or directly on the display screen of door lock A (such as the indoor display panel of the door lock), the "door lock event list" given by the interface is inquired according to the time period, and the attached video clip (such as video clip (A) or video clip (B)) of the time is directly clicked, and the door lock extracts the corresponding video and sends it to the mobile phone APP or directly displays it on the door lock display screen.

[0117] When the owner of door lock B inquires about the door lock event through the mobile phone APP (the door lock APP associated with door lock B) or directly on the display screen of door lock B, the "door lock event list" given by the interface is inquired according to the time period, and when the attached video clip (at this time, video clip (A) or video clip (B) does not exist) of the time is clicked, door lock B indexes the door lock event ID (A+t1) under the door lock event ID number (at this time, the door lock event stored by door lock B is door lock event ID (B+t1)), and queries the door lock IP (A) of the opposite door lock A in the "door relationship information" and sends a "request door lock event" message to door lock A, containing door lock ID (B), door lock event ID (A+t1) and the like, and door lock A receives the "request door lock event" message and extracts video clip (A) or video clip (B) according to door lock event ID (A+t1) to reply to door lock B, and door lock B receives and caches the A-replied video (if necessary, video clip (A) or video clip (B) can also be saved to door lock event ID (B+t1), which can ensure that subsequent inquiries do not need to send a request message to door lock A, but can be directly inquired locally to improve the inquiry response efficiency, but this method may bring storage overhead), and sends the video to the mobile phone APP or the door lock display screen, at this time the owner of door lock B can view the video on his own mobile phone APP or the display screen of door lock B.

[0118] Beneficial effects: through the storage method of the present scheme, video images can be saved comprehensively when important door lock events occur, and when the storage space is tight, the single-side saving method can be further used to save storage space.

[0119] (1) When the door lock detects that the person in front of the door lock leaves, the initial panoramic image is extracted and saved: when the door lock A detects that the person leaves the video scene, the panoramic image (A) and the panoramic image (B) are extracted after a delay of 2 minutes, a new door lock event is cached, and the detection of the person is detected.

[0120] (2) When unlocking the door or the person in the doorway lingers, record the door lock event video in all directions: when the door lock A detects that the person unlocks the door, or through the video (A) detects that the person in front of the door lock lingers, the door lock event containing the two videos is saved.

[0121] (3) To solve the problem of tight storage space of the door lock, the door locks A and B negotiate to keep only the video of the party that first detects the face: the door lock A first detects the face, sends a negotiation message to the door lock B, and determines by negotiation that the door lock A saves the door lock event containing the two videos.

[0122] (4) When detecting that a person appears in front of the door lock, the two video streams are obtained and cached, and after the person leaves, the two panoramic images are collected after a delay of 2 minutes: when the door lock A detects that a person appears in front of the door lock through the lock video (A), the video (B) of the door lock B is obtained through the door lock IP address, and the two videos are cached in the door lock event.

[0123] It can be seen that according to the event of detecting the appearance of the person by the first door lock, the video stream address of the second door lock is obtained through the door relationship information, and the video streams of the two door locks are cached synchronously; according to the video analysis result, the event type is judged, the complete video stream of the two parties is saved for important events, and only the panoramic image is saved for non-important events; according to the time difference of the face detection of the two door locks and the remaining storage space, the single door lock is determined to store the video stream of the two parties through negotiation; according to the user query request, the video data is obtained locally or through the door cooperation to complete the query display, so that efficient video acquisition and storage can be realized through event driving and door lock cooperation, the completeness of the security video recording is improved, the storage space is saved, and more efficient and reliable intelligent query response is supported.

[0124] Another embodiment of the application provides a door lock video recording system based on event value driving, referring to Figure 3 , the system can include:

[0125] The acquisition module 301 is used for cooperative video stream acquisition: according to the event of detecting the appearance of the person by the first door lock, the video stream address of the second door lock is obtained through the door relationship information, and the video streams of the two door locks are cached synchronously;

[0126] The evaluation module 302 is used for event value evaluation and processing: according to the video analysis result, the event type is judged, the complete video stream of the two parties is saved for important events, and only the panoramic image is saved for non-important events;

[0127] The negotiation module 303 is configured to store responsibility negotiation, and the single door lock stores video streams of the two parties according to the time difference of face detection of the two parties and the remaining storage space;

[0128] The response module 304 is configured to store intelligent query response, and the query display is completed according to the user query request, local direct response or video data obtained through cooperation with the door.

[0129] The embodiment of the present application further provides a storage medium, and the storage medium has a computer program stored therein, wherein the computer program is arranged to execute the steps in any one of the method embodiments when running.

[0130] The embodiment of the present application further provides an electronic device, which comprises a memory and a processor, the memory has a computer program stored therein, and the processor is arranged to execute the computer program to execute the steps in any one of the method embodiments.

[0131] Specifically, the electronic device can further comprise a transmission device and an input / output device, wherein the transmission device is connected with the processor, and the input / output device is connected with the processor.

[0132] The above embodiment according to the drawings illustrates the structure, features and effects of the present application, and the above description is only the preferred embodiment of the present application, but the present application is not limited to the drawings, any change or modification within the scope of the present application should be within the protection scope of the present application.

Claims

1. A door lock recording method based on event value-driven principles, characterized in that, The method includes: Collaborative video stream acquisition: Based on the first door lock detecting the presence of a person, the first door lock obtains the video stream address of the second door lock through the door relationship information and synchronously caches the video streams of both door locks; at the same time, based on the second door lock detecting the presence of the same person, the second door lock obtains the video stream address of the first door lock through the door relationship information and synchronously caches the video streams of both door locks. Event value assessment and handling: Based on the video analysis results, the event type is determined. For important events, the complete video streams of both parties are saved, while for non-important events, only panoramic images are saved. Storage responsibility negotiation: Based on the time difference of face detection on both door locks and the remaining storage space, the two door locks negotiated and determined that a single door lock would store the video streams of both parties; Intelligent query response: Based on the user's query request, the system can respond directly locally or obtain video data through collaboration with other systems to complete the query and display.

2. The method according to claim 1, characterized in that, The collaborative video stream acquisition includes: Personnel detection trigger: Based on the event of personnel presence detected by the first door lock video analysis module, a door lock event record is created and the start time is recorded; Video request to the second door lock: Based on the pre-stored door relationship information, send a video stream request message to the second door lock to obtain the video stream address of the second door lock; Video stream synchronization caching: Based on the obtained video stream address, synchronously capture and cache the video stream data of the first and second locks; Event status maintenance: Update the event type field of the door lock event record in real time based on the personnel activity status.

3. The method according to claim 2, characterized in that, The event value assessment and handling include: Event type determination: Based on the video analysis results, determine whether the event type is important or unimportant; Important event handling: For important events such as unlocking or people loitering, save the complete video stream data of both door locks; Non-critical event handling: For non-critical events such as changes in the door lock video scene or people passing by quickly, delete the video stream and save only the panoramic image when the duration is below the threshold; Panoramic image extraction: Extract keyframes from the video streams of the first and second locks in the previous event record after the person leaves to generate one panoramic image for scene recording during non-critical periods.

4. The method according to claim 3, characterized in that, The storage responsibility negotiation includes: Face detection time recording: Based on the face detection modules of both door locks, record the timestamp of the first face detection; Time difference comparison: Calculate the time difference between face detection at both door locks, and determine the priority storage party based on the sign of the time difference; Storage space comparison: When the time difference is zero, compare the remaining storage space of both door locks to determine the responsible party for storage; Storage negotiation execution: Based on the negotiation results, the party responsible for storage will store the video streams of both parties, while the other party will only store the event index.

5. The method according to claim 4, characterized in that, The intelligent query response includes: Local query processing: For events where the complete video is stored locally, the video stream is directly extracted to respond to the query request; Remote query processing: For events that only store event indexes, send a video request message to the door lock opposite the door; Video stream acquisition and display: Receive video stream data returned by the door lock, cache it locally, and display it to the user; Cache optimization: Decide whether to cache the video stream returned by the door lock locally based on the query frequency to optimize the efficiency of subsequent queries.

6. A door lock recording system based on event value-driven principles, characterized in that, The system includes: The acquisition module is used for collaborative video stream acquisition: based on the first door lock detecting the appearance of a person, the first door lock obtains the video stream address of the second door lock through the door relationship information and synchronously caches the video streams of both door locks; at the same time, based on the second door lock detecting the appearance of the same person, the second door lock obtains the video stream address of the first door lock through the door relationship information and synchronously caches the video streams of both door locks. The evaluation module is used for event value assessment and processing: it determines the event type based on video analysis results, saves the complete video streams of both parties for important events, and saves only the panoramic image for non-important events; The negotiation module is used for storing responsibility negotiation: based on the time difference of face detection between the two door locks and the remaining storage space, the two door locks negotiate to determine the storage of the video streams of both parties in a single door lock; The response module is used for intelligent query response: based on the user's query request, it can respond directly locally or obtain video data through collaboration with other departments to complete the query display.

7. The system according to claim 6, characterized in that, The acquisition module is specifically used for: Personnel detection trigger: Based on the event of personnel presence detected by the first door lock video analysis module, a door lock event record is created and the start time is recorded; Video request to the second door lock: Based on the pre-stored door relationship information, send a video stream request message to the second door lock to obtain the video stream address of the second door lock; Video stream synchronization caching: Based on the obtained video stream address, synchronously capture and cache the video stream data of the first and second locks; Event status maintenance: Update the event type field of the door lock event record in real time based on the personnel activity status.

8. The system according to claim 7, characterized in that, The evaluation module is specifically used for: Event type determination: Based on the video analysis results, determine whether the event type is important or unimportant; Important event handling: For important events such as unlocking or people loitering, save the complete video stream data of both door locks; Non-critical event handling: For non-critical events such as changes in the door lock video scene or people passing by quickly, delete the video stream and save only the panoramic image when the duration is below the threshold; Panoramic image extraction: Extract keyframes from the video streams of the first and second locks in the previous event record after the person leaves to generate one panoramic image for scene recording during non-critical periods.

9. A storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is executed by a processor to perform the method of any one of claims 1-5.

10. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to perform the method of any one of claims 1-5.

Images