CN120825613A

CN120825613A - Surveillance video processing method, device and storage medium

Info

Publication number: CN120825613A
Application number: CN202511301205.8A
Authority: CN
Inventors: 陈迅; 何梓轩; 朱剑文
Original assignee: Shenzhen Jetio Technology Co ltd
Current assignee: Shenzhen Jetio Technology Co ltd
Priority date: 2025-09-12
Filing date: 2025-09-12
Publication date: 2025-10-21

Abstract

The application discloses a processing method, equipment and a storage medium of a monitoring video, and relates to the technical field of electronic digital data processing, wherein the method comprises the steps of determining the storage medium of the monitoring video when a playback request of the monitoring video is received; when the storage medium is a magnetic tape, the monitoring video corresponding to the playback request is read from the magnetic tape, the monitoring video is migrated to a magnetic disk, and the monitoring video corresponding to the playback request is read and played in the magnetic disk. When the monitoring video is played back, the application transfers the cold data in the magnetic tape to the magnetic disk, thereby not only retaining the advantages of low cost and large capacity of long-term archiving of the magnetic tape, but also obviously shortening the data positioning and loading time through the magnetic disk reading and realizing the quick playback response.

Description

Method, device and storage medium for processing surveillance video

Technical Field

The present application relates to the field of electronic digital data processing technologies, and in particular, to a method and apparatus for processing a surveillance video, and a storage medium.

Background

In a scene with dominant massive cold data such as video monitoring, a tape mode is generally adopted to store massive videos, however, the tape is limited by sequential reading and writing and mechanical positioning by using a tape library, the retrieval time is different from a minute level to an hour level, the instant retrieval requirement in an emergency can not be met, and the video data retrieval efficiency is low.

Disclosure of Invention

The application mainly aims to provide a processing method, equipment and storage medium of a monitoring video, and aims to solve the technical problem of low video data retrieval efficiency.

In order to achieve the above object, the present application provides a method for processing a surveillance video, where the method for processing a surveillance video includes:

When a playback request of a monitoring video is received, determining a storage medium in which the monitoring video is located;

when the storage medium is a magnetic tape, reading the monitoring video corresponding to the playback request from the magnetic tape, and migrating the monitoring video to a magnetic disk;

And reading and playing the monitoring video corresponding to the playback request in the magnetic disk.

In one embodiment, the method comprises:

Grouping the monitoring videos according to at least one of a spatial logic relationship, a time sequence logic relationship and an event logic relationship of the cameras of the monitoring videos, wherein the spatial logic relationship comprises physical coordinates, visual angle coverage and direction association of the cameras;

and storing the monitoring video in the magnetic tape according to the grouping.

In one embodiment, the step of reading the surveillance video corresponding to the playback request from the magnetic tape and migrating the surveillance video to the magnetic disk includes:

Determining the bandwidth adopted by data migration according to the packets corresponding to the playback requests;

And migrating the monitoring video from the magnetic tape to a buffer area of the magnetic disk according to the bandwidth.

In one embodiment, the step of determining the bandwidth adopted by the data migration according to the packet corresponding to the playback request includes:

Determining the priority of the playback request according to the user permission and/or the emergency degree of the event;

And determining the bandwidth adopted by data migration according to the data quantity of the packets corresponding to the playback request and the priority.

Reading the monitoring video corresponding to the playback request from the magnetic tape, and caching the read monitoring video to a disk buffer area;

Determining the play rate of the monitoring video according to the playback request, selecting a key frame from the monitoring video in the disk buffer according to the play rate, and discarding video frames except the key frame;

And generating the processed monitoring video according to the key frame, and storing the processed monitoring video to the magnetic disk.

In one embodiment, the method comprises:

When the monitoring video is stored in a magnetic tape, acquiring video features corresponding to the monitoring video, wherein the video features comprise at least one of basic features, content features and historical features, the basic features comprise camera marks and/or labels, the content features comprise content importance scores and/or key event types, and the historical features comprise historical access frequencies of the monitoring video and/or average retrieval time of similar videos;

Inputting the video features into a prediction model to obtain the predicted migration time of the monitoring video output by the prediction model, wherein the prediction model is obtained by training based on video training features and historical migration records corresponding to video training data;

And reading the monitoring video from the magnetic tape at the predicted migration time, and migrating the monitoring video to a magnetic disk.

In an embodiment, before the step of determining the storage medium on which the monitoring video is located when the playback request of the monitoring video is received, the method further includes:

Determining a camera identifier and at least one tag corresponding to a video clip of each monitoring video;

storing video clips of the monitoring video in the magnetic tape according to the camera identification and the at least one tag;

the step of reading the monitoring video corresponding to the playback request from the magnetic tape and migrating the monitoring video to a magnetic disk comprises the following steps:

Determining a label corresponding to the monitoring video according to the playback request;

Determining the address of the monitoring video corresponding to the tag according to the corresponding relation between the tag and the address of the monitoring video, wherein the address comprises a start byte offset and an end byte offset of each video segment of the monitoring video;

And reading the monitoring video from the magnetic tape according to the initial byte offset and the ending byte offset, and migrating the monitoring video to a magnetic disk.

In an embodiment, the method further comprises:

Scanning the monitoring video and the corresponding label stored in the magnetic disk;

determining the invoking heat of the monitoring video according to the type of the tag of the monitoring video;

Determining a storage time length corresponding to the monitoring video according to the invoking heat, and transferring the monitoring video from the magnetic disk to the magnetic tape according to the storage time length;

The higher the calling heat is, the longer the storage time of the monitoring video in the magnetic disk is, the lower the calling heat is, and the shorter the storage time of the monitoring video in the magnetic disk is.

In addition, in order to achieve the above object, the application also provides a processing device of the monitoring video, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the computer program is configured to realize the steps of the processing method of the monitoring video.

In addition, in order to achieve the above object, the present application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, which when being executed by a processor, implements the steps of the method for processing surveillance video as described above.

Furthermore, to achieve the above object, the present application provides a computer program product comprising a computer program which, when being executed by a processor, implements the steps of the method for processing surveillance video as described above.

One or more technical schemes provided by the application have at least the following technical effects:

When receiving the playback request, firstly judging whether the monitoring video is stored on the magnetic tape, if yes, automatically transferring the monitoring video to the magnetic disk, and then reading and playing the monitoring video from the magnetic disk. When the monitoring video is played back, cold data in the magnetic tape is migrated to the magnetic disk, so that the advantages of low cost and large capacity of long-term archiving of the magnetic tape are reserved, the data positioning and loading time is obviously shortened through magnetic disk reading, quick playback response is realized, and user experience and system usability are improved.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the application and together with the description, serve to explain the principles of the application.

In order to more clearly illustrate the embodiments of the application or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, and it will be obvious to a person skilled in the art that other drawings can be obtained from these drawings without inventive effort.

FIG. 1 is a flowchart of a first embodiment of a method for processing surveillance video according to the present application;

FIG. 2 is a schematic flow chart of a first embodiment of a method for processing surveillance video according to the present application;

FIG. 3 is a schematic flow chart of a first embodiment of a method for processing surveillance video according to the present application;

Fig. 4 is a schematic flow chart of a second embodiment of a method for processing a surveillance video according to the present application;

fig. 5 is a schematic structural diagram of a second embodiment of a method for processing a surveillance video according to the present application;

Fig. 6 is a schematic flow chart of a third embodiment of a method for processing a surveillance video according to the present application;

FIG. 7 is a schematic flow chart of a fourth embodiment of a method for processing surveillance video according to the present application;

fig. 8 is a schematic flow chart of a fifth embodiment of a method for processing a surveillance video according to the present application;

Fig. 9 is a schematic device structure diagram of a hardware operating environment related to a method for processing a surveillance video according to an embodiment of the present application.

The achievement of the objects, functional features and advantages of the present application will be further described with reference to the accompanying drawings, in conjunction with the embodiments.

Detailed Description

It should be understood that the specific embodiments described herein are merely illustrative of the technical solution of the present application and are not intended to limit the present application.

For a better understanding of the technical solution of the present application, the following detailed description will be given with reference to the drawings and the specific embodiments.

The method and the device mainly solve the problems that when a playback request of a monitoring video is received, a storage medium where the monitoring video is located is determined, when the storage medium is a magnetic tape, the monitoring video corresponding to the playback request is read from the magnetic tape, the monitoring video is migrated to a magnetic disk, and the monitoring video corresponding to the playback request is read and played in the magnetic disk.

In this embodiment, for convenience of description, the following description will be made with the processing apparatus for monitoring video as an execution subject.

The application provides a solution, when receiving a playback request, firstly judging whether the monitoring video is stored on a magnetic tape, if the monitoring video is confirmed to be stored on the magnetic tape, automatically transferring the monitoring video to a magnetic disk, and then reading and playing the monitoring video from the magnetic disk. When the monitoring video is played back, cold data in the magnetic tape is migrated to the magnetic disk, so that the advantages of low cost and large capacity of long-term archiving of the magnetic tape are reserved, the data positioning and loading time is obviously shortened through magnetic disk reading, quick playback response is realized, and user experience and system usability are improved.

It should be noted that, the execution body of the embodiment may be a computing service device with functions of data processing, network communication and program running, such as a tablet computer, a personal computer, a mobile phone, or an electronic device, a processing device for monitoring video, or the like, which can implement the above functions. The present embodiment and the following embodiments will be described below by taking a processing device for monitoring video as an example.

Based on this, an embodiment of the present application provides a method for processing a surveillance video, and referring to fig. 1, fig. 1 is a schematic flow chart of a first embodiment of the method for processing a surveillance video according to the present application.

In this embodiment, the method for processing the surveillance video includes steps S10 to S30:

step S10, when a playback request of the monitoring video is received, a storage medium where the monitoring video is located is determined.

In this embodiment, the storage medium of the monitoring video may be a magnetic disk or a magnetic tape, or the like. Wherein the magnetic disk is used for real-time writing and high-frequency playback of data, and the magnetic tape is used for long-time storage of data.

In an embodiment, the playback request of the surveillance video includes a play requirement of the surveillance video, where the play requirement includes a play rate and/or a play definition, and the play definition requirement corresponds to a requirement of resolution, a code rate, a coding standard, and the like.

Optionally, the playback request of the surveillance video further includes an identification of the surveillance video, a surveillance time period, a surveillance target, etc., a type of alarm event, etc.

In one embodiment, before step S10, the method further comprises the steps of writing the monitoring video stream into a disk as video data in time slices and generating a storage medium mark of metadata of the video data, updating the storage medium mark after the video data are migrated from the disk to the magnetic tape to locate the storage medium where the video data are located, and step S10 comprises the steps of determining the storage medium where the monitoring video is located according to the storage medium mark in the metadata and achieving medium non-sensing access through the metadata index.

Optionally, after step S10, the method further includes, when the storage medium on which the monitoring video is located is a magnetic disk, reading the monitoring video corresponding to the playback request from the magnetic disk, and when the storage medium is a magnetic tape, reading the monitoring video corresponding to the playback request from the magnetic tape, and migrating the monitoring video to the magnetic disk. Illustratively, referring to FIG. 2, when the service system reads data, the data is directly retrieved from the disk storage subsystem when the data is stored in the disk storage subsystem, and when the data is stored in the tape library subsystem, the data is migrated back to the disk storage subsystem and retrieved from the disk storage subsystem.

Optionally, a metadata index library is constructed, the time stamp or storage location of the video in the tape or disk is recorded, and the positioning of the surveillance video is accelerated through the time stamp or storage location.

In an alternative embodiment, the method further comprises the steps of scanning video data stored in the magnetic disk based on a preset period, migrating the video data from the magnetic disk to the magnetic tape when the storage time of the video data is longer than a preset time period threshold value and/or the access frequency is smaller than a preset frequency threshold value, and deleting the video data in the magnetic tape when the storage time of the video data is longer than a preset time period upper limit value before the step S10. Optionally, when the storage duration of the video data is less than or equal to a preset duration threshold and the access frequency is greater than a preset frequency threshold, the video data is retained in the tape. According to the double threshold values of the storage duration and the access frequency, cold data is automatically migrated from a disk to a low-cost magnetic tape, and is thoroughly deleted after the expiration, so that not only is the fact that a recently high-access video always resides in the disk and can be played back in second level ensured, but also the space of an expensive disk is released to the maximum extent, the long-term archiving cost is reduced, and the dynamic balance of performance and cost is realized.

In one embodiment, referring to FIG. 3, data is scanned on a regular basis, data migration is initiated when the data storage period exceeds the hot data storage period, data is migrated from the disk storage subsystem to the tape library storage subsystem, and data is deleted when the data storage period exceeds the hot data storage period.

In one embodiment, video data is migrated from disk to tape based on the frequency of access of the video data. Optionally, the video data is migrated from the disk to the tape when the access frequency is less than a preset frequency threshold. When the access frequency is greater than or equal to a preset frequency threshold, video data is retained in the disk.

In one embodiment, video data is migrated to a tape based on a storage duration of the video data. And when the storage time length is greater than a preset time length threshold value, migrating the video data from the disk to the magnetic tape. And when the storage time length is less than or equal to a preset time length threshold value, video data are reserved in the disk.

In one embodiment, video data is migrated from disk to tape based on access frequency and access period of the video data. Optionally, the video data is migrated from the disk to the tape when the access frequency is less than a preset frequency threshold and outside of the access period. When the access frequency is greater than or equal to a preset frequency threshold or within an access period, video data is reserved in the disk.

And step S20, when the storage medium is a magnetic tape, reading the monitoring video corresponding to the playback request from the magnetic tape, and migrating the monitoring video to a magnetic disk.

In the monitoring system, if the monitoring video stored in the tape needs to be played back, the video data in the tape should be read sequentially and completely migrated to a cache buffer, such as an SSD (Solid STATE DRIVE) buffer pool. Then, the video data is read from the disk buffer at random according to the time stamp, and is restored to a playable format through demultiplexing and decoding, and finally output to the display terminal. Where the disk buffer is temporary storage, such as a dedicated cache disk, rather than long-term storage, it may persist to disk after the data has migrated. Optionally, a ring buffer is used to automatically cover the outdated data, avoiding overflow. The method comprises the steps of storing according to a time or camera identification catalog, assisting in recording paths of a database, reserving copies of a buffer area, supporting concurrent access, and deleting after migration.

If the tape data is a complete file, such as MP4 fragment, it needs to be completely migrated to disk before decoding and playing. In the case of streaming data, such as TS (Transport Stream), decoding can be performed while migration, but the characteristic of tape sequential reading may cause delay fluctuations. If the tape stores the original encoded stream, the video stream is separated by a demultiplexer and then decoded. If the tape is stored in a private format, the tape is converted into a standard format and then decoded for general use.

When the monitoring video is read in the magnetic tape, the magnetic tape is mechanically positioned and time-consuming, and can be preloaded with adjacent time period data, such as requests of 10:00-10:30 and additional loads of 9:50-10:40, and the magnetic tape library is adopted for parallel reading.

Optionally, the periodicity of the surveillance video is determined according to the type of the surveillance video, and the predicted migration time of the surveillance video is determined according to the periodicity. And reading the monitoring video corresponding to the playback request from the magnetic tape at the predicted migration time, and migrating the monitoring video to the magnetic disk.

In one embodiment, when a monitoring video is stored in a magnetic tape, a camera identification and a label corresponding to the monitoring video are acquired, the camera identification and the label are input into a prediction model, the predicted migration time of the monitoring video output by the prediction model is acquired, wherein the prediction model is obtained by training based on the camera identification, the label and a historical migration record corresponding to video training data, and the monitoring video is read from the magnetic tape and migrated to a magnetic disk at the predicted migration time. And storing the video segments of the monitoring videos in the magnetic tape according to the camera identifications and the at least one label.

In one embodiment, a playback request is determined according to a playback request, the playback request includes a resolution and/or a code rate, a monitoring video corresponding to the playback request is read from a tape, the monitoring video is cached in a disk buffer, the monitoring video in the disk buffer is processed according to the resolution and/or the code rate, and the processed monitoring video is migrated to a disk.

In one embodiment, the monitoring video in the disk buffer is processed according to the resolution, and the processed monitoring video is migrated to the disk. Optionally, the monitoring video in the disk buffer is coded according to the code rate, and the coded monitoring video is migrated to the disk. Optionally, the monitoring video in the disk buffer is processed according to the resolution, the processed monitoring video is encoded according to the code rate, and the encoded monitoring video is migrated to the disk.

In one embodiment, the monitoring video in the disk buffer is encoded according to the code rate, and the encoded monitoring video is migrated to the disk. The original monitoring video in the disk buffer area is dynamically recoded according to the adjustment code rate, the file volume can be compressed within the acceptable range of the image quality, precious online disk space is immediately vacated, the recoded low-code stream file still falls off first, the second-level retrieval of emergency is ensured, and the batch migration is carried out to the long-term archiving of the magnetic tape after the recoded low-code stream file is thoroughly cooled, so that the cost of expanding capacity, depreciation and offline capacity of the magnetic tape of the magnetic disk is obviously reduced on the premise of not sacrificing instant usability. Optionally, the code rate is dynamically adjusted according to the scene complexity, wherein a high code rate is adopted in a motion state, and a low code rate is adopted in a static state.

In an alternative embodiment, before step S10, the method further comprises the steps of determining a camera identifier and at least one tag corresponding to the video segment of each monitoring video, and storing the video segment of the monitoring video in the magnetic tape according to the camera identifier and the at least one tag. The step S20 comprises the steps of determining a label corresponding to the monitoring video according to a playback request, determining the address of the monitoring video corresponding to the label according to the corresponding relation between the label and the address of the monitoring video, wherein the address comprises a start byte offset and an end byte offset of each video segment of the monitoring video, reading the monitoring video from a magnetic tape according to the start byte offset and the end byte offset, and migrating the monitoring video to a magnetic disk.

The method includes that a section of monitoring video is divided into at least two video segments, each video segment is marked with a label, for example, the label is a person, when the monitoring video is stored, the video segments of the monitoring video are sequentially stored in a tape according to a camera identification and the label, long monitoring videos are conveniently read in sequence during playback, and reading efficiency is improved. The video is divided into segments, each segment having a well-defined byte offset, which is particularly suitable for linear storage media such as magnetic tape. The magnetic tape is a sequential access device that does not support random reading and writing, so that the data location can be precisely located by an offset, reducing the read time. The fragmentation design is also convenient for processing large video files, allows partial reading and improves efficiency.

A mapping relationship, such as a database table or configuration file, is maintained to associate tags with the physical storage addresses of the video. The address includes a start byte offset and an end byte offset for each video segment. This step is logically coherent, with the tag as an intermediary, connecting abstract requests and concrete storage locations, avoiding the complexity of directly manipulating the underlying storage.

And step S30, reading and playing the monitoring video corresponding to the playback request in the magnetic disk.

In this embodiment, the data stream of the monitoring video is read, video decoding is performed on the data stream, and the decoded video frame sequence is rendered and output according to the playing time sequence.

In an alternative embodiment, because the monitor video may be discontinuous physical addresses when being written into the disk, the storage position information of the monitor video in the disk is obtained, the storage position information includes a start byte offset and an end byte offset of each video segment of the monitor video, and the monitor video is obtained and decoded according to the start byte offset and the end byte offset, and the decoded monitor video is played. The physical fragments of a file system or RAID (Redundant Array of INDEPENDENT DISKS ) are abstracted into a section of logically continuous data stream through the initial byte offset and the final byte offset, and the reading can be completed only through one-time I/O (Input/Output) scheduling, thereby greatly reducing the magnetic head movement or flash random read amplification, and reducing the delay and the occupation of a CPU (Central Processing Unit, a central processing unit). The offset directly corresponds to the video time point, can achieve millisecond jump, compared with the traditional two-stage index of searching for files first and then searching for GOP (Group of Pictures ), the overhead of analyzing an index table at a container layer is saved, and the method is suitable for random playback and alarm backtracking scenes of massive monitoring.

In one embodiment, when the playback request of the monitoring video includes the monitoring video requesting the preset target, determining storage media in which at least two monitoring videos of the preset target are respectively located, when the storage media includes a tape, transferring the monitoring video from the tape to a disk, reading the monitoring video corresponding to the playback request in the disk, when the storage media includes the disk, reading the monitoring video corresponding to the playback request in the disk, and finally merging and outputting the at least two monitoring videos of the preset target. Optionally, acquiring a face area and/or a scene area selected by a user, and determining a preset target corresponding to the playback request according to the face area and/or the scene area.

In one embodiment, when the playback request of the monitoring video includes the type of the alarm event, determining the monitoring video associated with the type of the alarm event according to the type of the alarm event, determining a storage medium in which the associated monitoring video is located, when the storage medium is a magnetic tape, reading the monitoring video corresponding to the playback request from the magnetic tape, and transferring the monitoring video to a magnetic disk, and reading and playing the monitoring video corresponding to the playback request in the magnetic disk. And when the storage medium comprises a disk, reading the monitoring video corresponding to the playback request from the disk, and finally merging and outputting the associated monitoring video.

In the technical scheme of the embodiment, when a playback request is received, whether the monitoring video is stored on the magnetic tape is firstly judged, if the monitoring video is confirmed to be stored on the magnetic tape, the monitoring video is automatically migrated to the magnetic disk, and then the monitoring video is read and played from the magnetic disk. When the monitoring video is played back, cold data in the magnetic tape is migrated to the magnetic disk, so that the advantages of low cost and large capacity of long-term archiving of the magnetic tape are reserved, the data positioning and loading time is obviously shortened through magnetic disk reading, quick playback response is realized, and user experience and system usability are improved.

In the second embodiment of the present application, the same or similar content as the above-described embodiment can be referred to the above description, and the description thereof will not be repeated. On this basis, referring to fig. 4, the method further includes:

Step S40, grouping the monitoring videos according to at least one of a spatial logic relationship, a time sequence logic relationship and an event logic relationship of the cameras of the monitoring videos, wherein the spatial logic relationship comprises physical coordinates, visual angle coverage and direction association of the cameras;

And step S50, storing the monitoring video in the magnetic tape according to the grouping.

Referring to fig. 5, the magnetic tape may be plural, and the magnetic disk may be plural, and video data may be retrieved from different magnetic tapes to a disk buffer based on a playback request, processed in the disk buffer, and stored in the magnetic disk. Optionally, video data is retrieved from different tapes to a disk buffer based on different playback requests, and the video data is processed in the disk buffer, and the processed video data is stored in the disk.

In this embodiment, the spatial logical relationship includes physical coordinates of the camera, view angle coverage, and direction association. Wherein the physical coordinates include the absolute or relative position of the camera in three-dimensional space, such as building coordinates or relative spacing. The visual angle coverage comprises a visual field overlapping area so as to form three-dimensional monitoring, the visual angle coverage comprises blind area complementation, and the A is invisible and is complemented by the B. The direction association, for example camera a is towards the east photographing entrance and camera B is towards the west photographing exit, both of which form an in-out relationship. The time logic relationship includes trigger sequence and time zone synchronization. And triggering a sequence, namely triggering a gate camera B to take a candid photograph of a driver after the parking lot camera A detects a license plate. Time zone synchronization is because the time zone cameras need to be uniformly time stamped, otherwise the event sequence is disordered. The event logic relationship includes a behavioral association, for example, if camera a recognizes that a person is missing suspicious spam, and camera B tracks the person within 5 seconds.

In an alternative embodiment, the steps of reading the monitoring video corresponding to the playback request from the magnetic tape and migrating the monitoring video to the magnetic disk comprise determining a bandwidth adopted by data migration according to the packet corresponding to the playback request, and migrating the monitoring video from the magnetic tape to a buffer area of the magnetic disk according to the bandwidth. The larger the data amount of the packet corresponding to the playback request, the larger the required bandwidth, and the smaller the data amount of the packet, the smaller the required bandwidth.

In an alternative embodiment, the step of determining the bandwidth adopted by the data migration according to the packets corresponding to the playback request comprises determining the priority of the playback request according to the user authority and/or the emergency degree of the event, and determining the bandwidth adopted by the data migration according to the data quantity and the priority of the packets corresponding to the playback request.

In this embodiment, the amount of data consumed is the total amount of video data that needs to be read and decoded from the data source per second in the video play and migration scenario.

Optionally, the user authority is an advanced management authority, the priority of the playback request is higher, the user authority is a common management authority, and the priority of the playback request is lower. Alternatively, the higher the event urgency, the higher the priority of the playback request, and conversely, the lower the event urgency, the lower the priority of the playback request. Optionally, the first priority of the playback request is determined according to the user authority, the second priority of the playback request is determined according to the time emergency degree, and the priority of the playback request is determined according to the preset weight parameter, the first priority and the second priority.

Optionally, determining that the data migration employs a first bandwidth when the data amount is greater than a preset data amount threshold and the priority is greater than a preset priority threshold, determining that the data migration employs a second bandwidth when the data amount is greater than a preset data amount threshold or the priority is greater than a preset priority threshold, and determining that the data migration employs a third bandwidth when the data amount is less than or equal to the preset data amount threshold and the priority is less than or equal to the preset priority threshold, wherein the first bandwidth is greater than the second bandwidth and the second bandwidth is greater than the third bandwidth.

In the technical scheme of the embodiment, the monitoring video is intelligently grouped according to the space, time sequence or event logic relation among cameras and then written into the magnetic tape, so that the data can carry the index information of scenes, behaviors and time at a physical layer, and the required fragments can be quickly positioned according to the logic relation during retrieval without sequentially scanning the whole magnetic tape, thereby obviously shortening the searching time, reducing the reading expenditure, reducing the return of irrelevant data and improving the storage utilization rate and the integral response efficiency of the system.

In the third embodiment of the present application, the same or similar contents as those of the above-described embodiments can be referred to the above description, and the description thereof will not be repeated. On this basis, referring to fig. 6, step S20 includes:

Step S21, reading the monitoring video corresponding to the playback request from the magnetic tape, and caching the read monitoring video to a disk buffer area;

step S22, determining the play rate of the monitoring video according to the playback request, selecting a key frame from the monitoring video in the disk buffer according to the play rate, and discarding video frames except the key frame;

step S23, generating the processed monitoring video according to the key frame, and storing the processed monitoring video to the magnetic disk.

Optionally, in a scene of high-speed playing, only the I frame, i.e. the key frame, is extracted, reducing the data volume. Under the scene of low-speed playing, an intermediate frame is generated by interpolation, and the motion blurring processing is increased to avoid the jamming.

In one embodiment, a key frame is selected from the monitor video in the disk buffer according to the play rate, and the monitor video after the frame extraction processing is generated according to the key frame. By precisely selecting the key frames in the disk buffer area according to the playing speed and generating the frame extraction video, the data volume can be compressed on the premise of not losing the overall view of the event, the expensive disk space is immediately released, meanwhile, the key frame sequence still keeps the complete time sequence, and the disk end can continue to provide second-level quick browsing.

Optionally, the monitoring video is uniformly frame-extracted or key frame-extracted. Optionally, during high-speed playing, a key frame or an I frame is extracted, during low-speed playing, a difference value complements the frame, and during normal speed, a complete frame is reserved.

And performing key frame extraction on the monitoring video so as to remove redundant frames while keeping key information, thereby improving storage efficiency and subsequent analysis efficiency. Wherein the key information comprises dynamic target behavior, abnormal events, important state changes and the like.

In one embodiment of extracting the key frame, a target detection model is adopted to identify targets in the video, such as people, vehicles and objects in real time, and then semantic behaviors of the targets, such as people falling, vehicle emergency braking and abnormal movement of the objects, are analyzed through a behavior identification model. Setting a semantic key event library, such as collision of personnel aggregation, danger of vehicle retrograde, suspicious article leaving and the like, automatically triggering key frame extraction when an event in the semantic key event library is detected, and marking event types, such as personnel falling at 15 points.

Optionally, semantic filtering is performed on nonsensical motions, leaving only frames relevant to the target key behavior as key frames. For example, in campus monitoring, the motion of wind blowing the branches is filtered, and the action of a student crossing the fence is determined as a key event, and the video frame at that time is extracted as a key frame.

In an embodiment for extracting the key frames, the key frames are extracted by fusing multi-mode information outside the video, wherein the multi-mode information comprises audio, sensors, environmental data and the like, the information limitation of relying on the video frames only is broken through, and key events which are difficult to find in a single mode are captured.

Optionally, in a monitoring scenario, audio signals such as glass breaking sounds, distress sounds, abnormal sounds are often strongly correlated with critical events such as theft, collision. Abnormal audio is identified through the audio event detection model, such as screaming means that an emergency event occurs, continuous striking means destruction, and video frame extraction at corresponding time is synchronously triggered.

Optionally, the environmental sensor data are fused to extract a key frame, for example, the environmental sensor data comprise illumination intensity, temperature and humidity, vibration sensors and the like, when the light is changed suddenly in rainy days or at night, the video motion sensitivity is reduced, and the situation that the frame where the raindrops or the light changes is misjudged to be the key frame is avoided. When the vibration sensor detects severe vibration such as wall knocking, the sensitivity of the video to micro motions is improved, and the anomalies near the seismic source are captured.

In one embodiment of extracting key frames, based on video timing continuity, complete event logic is preserved with fewer frames, avoiding repeated extraction of consecutive similar frames, and improving storage efficiency.

Alternatively, a transition or LSTM (Long Short-Term Memory network) model is used to learn the time sequence association between frames, such as continuous action logic of walking, turning and stopping of a person, and identify information inflection points in the time sequence, i.e. frames representing a continuous action core logic, such as a key transition of walking, stopping of a person. By way of example, a section of vehicle is driven to stop from a video, the time sequence model only extracts 3 key frames of driving into an intersection, turning on a turn signal lamp and stopping, and the complete process can be restored through inter-frame time sequence association.

Optionally, redundant time sequence segments such as non-targeted static background are time compressed, only segment head and tail frames are reserved, high information density segments such as multi-person interactions are time expanded, and key frame extraction frequency is encrypted.

In one embodiment of extracting key frames, in a massive monitoring video without labels, the model learns key frame characteristics autonomously through positive and negative sample comparison, wherein positive samples represent representative frames in the same event, such as frames with initial fire and enlarged fire, and negative samples represent redundant frames, such as repeated frames of static background. The model automatically grasps the law of key frames=frames with remarkable information increment through comparison learning. The model is integrated into various scenes such as markets, factories, cells and high speeds in training, and can accurately extract frames in new scenes through the field self-adaptive technology, so that the problem that parameters need to be readjusted in scene changing in the traditional scheme is solved.

In the technical scheme of the embodiment, the frame extraction is carried out on the monitoring video in the disk buffer area according to the play rate before the monitoring video is migrated to the disk, so that not only is the key picture required by event backtracking reserved, but also the redundant frame is greatly deleted, thereby compressing the volume of cold data, directly reducing the long-term archiving capacity and bandwidth occupation of the tape end, and simultaneously, the monitoring video after frame extraction is written into the disk, so that the second-level quick retrieval can be continuously satisfied.

In the fourth embodiment of the present application, the same or similar contents as those of the above-described embodiments can be referred to the above description, and the description thereof will be omitted. On this basis, referring to fig. 7, the method includes:

Step S60, when the monitoring video is stored in a magnetic tape, acquiring video features corresponding to the monitoring video, wherein the video features comprise at least one of basic features, content features and historical features, the basic features comprise camera marks and/or labels, the content features comprise content importance scores and/or key event types, and the historical features comprise historical access frequencies of the monitoring video and/or average retrieval time of similar videos;

step S70, inputting the video features into a prediction model, and obtaining the predicted migration time of the monitoring video output by the prediction model, wherein the prediction model is obtained by training based on video training features and historical migration records corresponding to video training data;

and step S80, reading the monitoring video from the magnetic tape at the predicted migration time, and migrating the monitoring video to a magnetic disk.

In the present embodiment, the predicted migration time precedes the time at which the playback request is received, or the predicted migration time precedes the time at which the video of the playback request is played back. Alternatively, the predicted migration time may be a point in time or a period of time.

The basic features include camera identification and/or tags, wherein the camera identification may be a person tag, an event tag, an object tag, etc. in order to distinguish cameras of different areas or angles.

The content features include content importance scoring and/or key event types, content analysis of surveillance videos stored in the tape, identification of whether key events are included, and determination of key event types, wherein the key events include abnormal events such as personnel conflicts, vehicle accidents, etc., and the key events include events of high access probability scenes such as entrances and exits, dynamic pictures of accident high-incidence areas, etc. The content importance score quantifies the importance of the content, illustratively, contains an anomaly event, importance=0.9, no key object and still picture, importance=0.1, as the content importance score. Key targets such as designated personnel, suspicious items.

The historical characteristics include historical access frequency of the surveillance video and/or average retrieval time of the same type of video. The similar video is the same type of video as the monitoring video, and can also be a plurality of videos corresponding to the same event.

Optionally, the predictive model is trained based on video training features of video training data consisting of historical surveillance video and migration records from tape migration to disk. In this embodiment, the prediction model may be an LSTM model. By utilizing the prediction model trained based on the historical migration record, the predicted migration time of the most likely to be called up of the monitoring video can be calculated in advance during the storage period of the magnetic tape, and the monitoring video is prefetched from the magnetic tape to the magnetic disk before the real arrival of the playback request, so that the long-term storage advantage of the magnetic tape with low cost can be continuously enjoyed, the mechanical loading delay of the original minute level or even the hour level is compressed to be almost zero, and the instant call experience in the emergency is remarkably improved.

Optionally, determining the monitoring video to be migrated and the predicted migration time based on the prediction model, reading the monitoring video to be migrated from the magnetic tape at the predicted migration time before receiving the playback request, and migrating the monitoring video to be migrated to the magnetic disk. The most probably replayed monitoring video and the time point thereof are calculated in advance through the prediction model, before a user really initiates a replay request, the corresponding monitoring video is prefetched from the magnetic tape to the magnetic disk, when the user clicks the replay, the data are ready, the zero waiting experience close to the magnetic disk is realized, meanwhile, the waste of the magnetic disk space caused by blind preloading is avoided, and the response speed and the storage resource utilization rate are both considered.

In the technical scheme of the embodiment, the predicted migration time of the early migration of the monitoring video is calculated in advance through the prediction model, the corresponding monitoring video is prefetched from the magnetic tape to the magnetic disk before the user really initiates the playback request, and when the user clicks the playback, the data are ready, so that zero waiting experience close to the magnetic disk is realized, the waste of magnetic disk space caused by blind preloading is avoided, and the response speed and the storage resource utilization rate are both considered.

In the fifth embodiment of the present application, the same or similar contents as those of the above-described embodiments can be referred to the above description, and the description thereof will be omitted. On this basis, referring to fig. 8, the method further includes:

step S90, scanning the monitoring video and the corresponding label stored in the magnetic disk;

Step S100, determining the invoking heat of the monitoring video according to the type of the tag of the monitoring video;

step S110, determining a storage duration corresponding to the monitoring video according to the invoking heat, and migrating the monitoring video from the magnetic disk to the magnetic tape according to the storage duration.

In this embodiment, the higher the retrieval heat, the longer the storage duration of the monitoring video in the disk, the lower the retrieval heat, and the shorter the storage duration of the monitoring video in the disk.

Optionally, the scanning period is shortened for the video of the high-heat label, so that the heat change can be captured in time, such as the heat rise after burst calling and temporary non-migration, and the scanning period is prolonged for the video of the low-heat label, so that the occupation of the ineffective scanning on the disk performance is reduced. For example, when the number of times a certain video is called in the last 1 hour is less than or equal to 1 time, and the total call in 7 historical days is less than or equal to 3 times, the low-heat early warning is automatically triggered, the preset period is not required to wait, and the migration evaluation flow is directly entered. For example, a warehouse monitoring tag is a goods storage area, but no call is made for 15 consecutive days, and after the system detects in real time, the system starts to migrate to a magnetic tape in advance to release the disk space.

Optionally, the tags of the monitoring video include people, animals, objects, events and the like, different tag types correspond to different heat adjustment degrees, and different tags also correspond to different heat adjustment degrees in different time periods.

In an embodiment, the invoking heat of the monitoring video is determined according to the type of the tag of the monitoring video and the tag corresponding to the high-frequency event in the current time period.

Optionally, after determining the heat of calling of the monitoring video according to the type of the tag of the monitoring video, if the monitoring video contains an abnormal event and a key target, increasing the heat of calling, and if the monitoring video is a pure static picture or nonsensical repeated content, reducing the heat of calling.

Optionally, the invoking heat of the monitoring video is determined according to at least two of the type, the invoking duration, the invoking subject and the associated invoking of the tag of the monitoring video. The method comprises the steps of determining that the calling heat is high if single check is larger than preset time, only fast browsing, determining that the calling heat is low if single check is larger than preset time, determining that the calling heat is high if security personnel call for many times, determining that the calling heat is high if common personnel check accidentally, determining that the calling heat is low if related calling, and determining that the calling heat is high if a certain video is called as incident related evidence for many times.

In an embodiment, the storage duration corresponding to the monitoring video is determined according to the invoking heat, and the storage duration=max (duration determined by heat, minimum compliance duration).

In the technical scheme of the embodiment, the strategy of dynamically determining the retention time of the data in the disk according to the call heat is adopted, so that the high-value video with frequent access is remained in the high-speed disk as much as possible, and the low-frequency video is migrated to the low-cost tape as early as possible, thereby obviously reducing the online storage capacity and the total possession cost while ensuring the quick response of the hot spot data, and realizing the optimal balance between the performance and the cost of monitoring video storage.

It should be noted that the foregoing examples are only for understanding the present application, and do not limit the processing method of the surveillance video of the present application, and more forms of simple transformation based on the technical concept are all within the scope of the present application.

The application provides a processing device of a monitoring video, which comprises at least one processor and a memory in communication connection with the at least one processor, wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the processing method of the monitoring video in the first embodiment.

Referring now to fig. 9, a schematic diagram of a processing device suitable for implementing surveillance video in accordance with an embodiment of the present application is shown. The processing device for monitoring video in the embodiment of the present application may include, but is not limited to, mobile terminals such as notebook computers, tablet computers (PAD, portable Application Description), portable multimedia players (PMP, portable MEDIA PLAYER), vehicle terminals (e.g., vehicle navigation terminals), and the like, and fixed terminals such as digital TVs, desktop computers, and the like. The processing device for monitoring video shown in fig. 9 is only an example, and should not impose any limitation on the functions and application scope of the embodiments of the present application.

As shown in fig. 9, the processing apparatus of the surveillance video may include a processing device 1001 (e.g., a central processing unit, a graphic processor, etc.), which may perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM) 1002 or a program loaded from a storage device 1003 into a random access Memory (RAM, random Access Memory) 1004. In the RAM1004, various programs and data necessary for the operation of the processing apparatus for monitoring video are also stored. The processing device 1001, the ROM1002, and the RAM1004 are connected to each other by a bus 1005. An input/output (I/O) interface 1006 is also connected to the bus. In general, a system including an input device 1007 such as a touch screen, a touch pad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc., an output device 1008 including a Liquid Crystal Display (LCD) CRYSTAL DISPLAY, a speaker, a vibrator, etc., a storage device 1003 including a magnetic tape, a hard disk, etc., and a communication device 1009 may be connected to the I/O interface 1006. The communication means 1009 may allow the processing device of the surveillance video to communicate wirelessly or by wire with other devices to exchange data. Although the figures illustrate a processing device for surveillance video with various systems, it should be understood that not all illustrated systems are required to be implemented or provided. More or fewer systems may alternatively be implemented or provided.

In particular, according to embodiments of the present disclosure, the processes described above with reference to flowcharts may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method shown in the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through a communication device, or installed from the storage device 1003, or installed from the ROM 1002. The above-described functions defined in the method of the disclosed embodiment of the application are performed when the computer program is executed by the processing device 1001.

The processing equipment of the monitoring video provided by the application adopts the processing method of the monitoring video in the embodiment, and can solve the technical problem of low video data retrieval efficiency. Compared with the prior art, the beneficial effects of the processing device for the monitoring video provided by the application are the same as those of the processing method for the monitoring video provided by the embodiment, and other technical features of the processing device for the monitoring video are the same as those disclosed by the method of the previous embodiment, so that the description is omitted.

It is to be understood that portions of the present disclosure may be implemented in hardware, software, firmware, or a combination thereof. In the description of the above embodiments, particular features, structures, materials, or characteristics may be combined in any suitable manner in any one or more embodiments or examples.

The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon for performing the processing method of surveillance video in the above-described embodiments.

The computer readable storage medium provided by the present application may be, for example, a USB flash disk, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system or device, or a combination of any of the foregoing. More specific examples of a computer-readable storage medium may include, but are not limited to, an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access Memory (RAM, random Access Memory), a Read-Only Memory (ROM), an erasable programmable Read-Only Memory (EPROM, erasable Programmable Read Only Memory, or flash Memory), an optical fiber, a portable compact disc Read-Only Memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In this embodiment, a computer-readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, or device. Program code embodied on a computer readable storage medium may be transmitted using any appropriate medium, including but not limited to electrical wiring, fiber optic cable, radio Frequency (RF), the like, or any suitable combination of the foregoing.

The computer readable storage medium may be included in the processing device of the surveillance video or may exist alone without being incorporated in the processing device of the surveillance video.

The computer readable storage medium is loaded with one or more programs, and when the one or more programs are executed by the processing device of the monitoring video, the processing device of the monitoring video firstly judges whether the monitoring video is stored on the magnetic tape or not when receiving a playback request, and if the monitoring video is confirmed to be stored on the magnetic tape, the monitoring video is automatically migrated to the magnetic disk and then read and played from the magnetic disk. When the monitoring video is played back, cold data in the magnetic tape is migrated to the magnetic disk, so that the advantages of low cost and large capacity of long-term archiving of the magnetic tape are reserved, the data positioning and loading time is obviously shortened through magnetic disk reading, quick playback response is realized, and user experience and system usability are improved.

Computer program code for carrying out operations of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, smalltalk, C ++ and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a local area network (LAN, local Area Network) or a wide area network (WAN, wide Area Network), or may be connected to an external computer (e.g., through the internet using an internet service provider).

The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The modules involved in the embodiments of the present application may be implemented in software or in hardware. Wherein the name of the module does not constitute a limitation of the unit itself in some cases.

The readable storage medium provided by the application is a computer readable storage medium, and the computer readable storage medium stores computer readable program instructions (namely computer program) for executing the processing method of the monitoring video, so that the technical problem of low video data retrieval efficiency can be solved. Compared with the prior art, the beneficial effects of the computer readable storage medium provided by the application are the same as those of the method for processing the monitoring video provided by the above embodiment, and are not described in detail herein.

The application also provides a computer program product comprising a computer program which, when executed by a processor, implements the steps of the method for processing surveillance video as described above.

The computer program product provided by the application can solve the technical problem of low efficiency of video data retrieval. Compared with the prior art, the beneficial effects of the computer program product provided by the application are the same as those of the method for processing the monitoring video provided by the above embodiment, and are not described herein.

The foregoing description is only a partial embodiment of the present application, and is not intended to limit the scope of the present application, and all the equivalent structural changes made by the description and the accompanying drawings under the technical concept of the present application, or the direct/indirect application in other related technical fields are included in the scope of the present application.

Claims

1. The processing method of the monitoring video is characterized by comprising the following steps of:

2. The method for processing a surveillance video according to claim 1, characterized in that the method comprises:

3. The method for processing a surveillance video according to claim 2, wherein the step of reading the surveillance video corresponding to the playback request from the magnetic tape and migrating the surveillance video to a magnetic disk comprises:

4. The method for processing surveillance video of claim 3, wherein the step of determining a bandwidth employed for data migration based on the packets corresponding to the playback request comprises:

5. The method for processing a surveillance video according to claim 1, wherein the step of reading the surveillance video corresponding to the playback request from the magnetic tape and migrating the surveillance video to a magnetic disk comprises:

6. The method for processing a surveillance video according to claim 1, characterized in that the method comprises:

7. The method for processing a surveillance video according to claim 1, further comprising, before the step of determining a storage medium in which the surveillance video is located when a playback request of the surveillance video is received:

8. The method for processing a surveillance video of claim 1, further comprising:

9. A processing device for surveillance video, characterized in that it comprises a memory, a processor and a computer program stored on the memory and executable on the processor, the computer program being configured to implement the steps of the processing method for surveillance video according to any of claims 1 to 8.

10. A storage medium, characterized in that the storage medium is a computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, realizes the steps of the method for processing surveillance video according to any one of claims 1 to 8.