CN114155464B - Video data storage method and device, storage medium and terminal - Google Patents

Abstract

The invention discloses a video data evidence storing method which is applied to a first client and comprises the following steps: when video data to be transmitted are acquired, acquiring a pre-trained scene type recognition model set for the video data; extracting a plurality of key video frames in the video data, and determining the scene type of each key video frame based on the model; loading driving data of a current vehicle, and constructing a mask picture carrying parameters according to the driving data and the scene type of each key video frame; synthesizing the mask picture carrying the parameters and each key video frame to generate synthesized target video data; and sending the video data and the driving data to be transmitted to a second client, and sending the processed target video data and the driving data to a cloud server. Because this application passes through the scene type of model identification video data to combine driving data to carry out the secondary synthesis to video data, make the video of driver report difficult to falsify, promoted the authenticity of video.

Description

A video data storage method, device, storage medium and terminal

technical field

The invention relates to the technical field of data security, in particular to a video data storage method, device, storage medium and terminal.

Background technique

When the truck driver is in the scene of loading and unloading, traffic jam, accident, refueling, attendance check-in, etc., the truck driver needs to shoot the current operation scene through the watermark video, and report the current operation scene to the owner or the owner of the vehicle for vehicle report or manage the vehicle.

In the prior art, when the truck driver uses the mobile phone to obtain evidence, the realization method is: the truck driver uses the mobile phone sensor to take pictures, video and other evidence collection, stores the corresponding evidence collection documents in the mobile phone, and reports the evidence collection documents to the cargo owner or vehicle owner from the mobile phone. Since the evidence files will be stored in the mobile phone, and then transferred from the mobile phone, the evidence files will be replaced or tampered with, resulting in the evidence files transmitted to the back-end server not being real evidence files, that is, the authenticity of the evidence files cannot be obtained. Warranty, thereby reducing the authenticity of the video.

Contents of the invention

Embodiments of the present application provide a video data certificate storage method, device, storage medium, and terminal. In order to provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is presented below. This summary is not an overview, nor is it intended to identify key/critical elements or delineate the scope of these embodiments. Its sole purpose is to present some concepts in a simplified form as a prelude to the more detailed description that is presented later.

In the first aspect, the embodiment of the present application provides a video data storage method, which is applied to the first client, and the method includes:

When the video data to be transmitted is collected, a pre-trained scene type recognition model set for the video data is obtained;

Extract multiple key video frames in the video data, and determine the scene type of each key video frame based on the pre-trained scene type recognition model;

Load the driving data of the current vehicle, and construct a mask image with parameters according to the driving data and the scene type of each key video frame;

Synthesize the mask image carrying parameters with each key video frame to generate synthesized target video data;

Send the video data and driving data to be transmitted to the second client, and send the target video data to the cloud server together with the driving data after processing.

Optionally, follow the steps below to generate a pre-trained scene type recognition model, including:

Collecting scene images where the vehicle is located to obtain model training samples; where the scene images at least include vehicle unloading scenes, vehicle refueling scenes, vehicle driving scenes, and vehicle accident scenes;

Use the YOLOv5 algorithm to create a scene type recognition model;

Input the model training samples into the scene type recognition model for model training, and output the loss value;

When the loss value reaches the minimum, generate a pre-trained scene type recognition model;

or,

When the loss value does not reach the minimum, the loss value is backpropagated to adjust the model parameters of the scene type recognition model, and continue to input the model training samples into the scene type recognition model for model training.

Optionally, the scene type recognition model includes an input terminal, a reference network, a Neck network and a Head output terminal;

Determine the scene type of each key video frame based on a pre-trained scene type recognition model, including:

The input terminal receives each key video frame, scales each key video frame to a preset size, and performs normalization to obtain a normalized video frame;

The benchmark network performs feature extraction on the normalized video frames to obtain a set of feature maps;

The Neck network fuses each feature map in the feature map set with the preset basic features to obtain the fused feature map;

The head output uses the classification branch to classify the fused feature maps, and uses the regression branch to perform linear regression on the classified types to obtain the scene type of each key video frame.

Optionally, construct a mask image with parameters according to the driving data and the scene type of each key video frame, including:

Get the mask image;

Identify the first parameter identification set on the mask image;

Identify the second parameter identification set corresponding to the driving data and the scene type of each key video frame;

Identify the same parameter identifiers from the first parameter identifier set as the parameter identifiers in the second parameter identifier set to perform data mapping, and generate a mask picture carrying parameters.

Optionally, the target video data is processed and sent to the cloud server together with the driving data, including:

Obtain digital watermark image;

Intercept square RGB images from the image of the target video data and the digital watermark image respectively to obtain the first image and the second image;

performing color channel separation on the first image to obtain a first color component set, and performing color channel separation on the second image to obtain a second color component set;

A transformation matrix is obtained after performing Arnold transformation on the first color component set;

performing DCT transformation on the second color component set according to the transformation matrix to obtain a DC component;

According to the transformation matrix and the DC component, the digital watermark is embedded into the target video data, and the processed video data is generated;

Send the processed video data and driving data to the cloud server.

In the second aspect, the embodiment of the present application provides a video data storage method, which is applied to a cloud server, and the method includes:

receiving the processed video data and driving data sent by the first client to the cloud server;

Convert the processed video data to binary data;

Perform SHA256 hash operation on the binary data and the driving data to obtain the first hash string;

Save the first hash string to the blockchain.

In the third aspect, the embodiment of the present application provides a video data storage method, which is applied to the second client, and the method includes:

When receiving the video data and driving data to be transmitted sent by the first client for the second client, establish communication with the cloud server and obtain the first hash string stored in the block chain;

Performing SHA256 hash operation on the video data to be transmitted and the driving data to obtain a second hash string;

When the first hash string is identical to the second hash string and the digital watermark in the video data to be transmitted is correct, play the video data to be transmitted;

or,

When the first hash string is the same as the second hash string and the digital watermark in the video data to be transmitted is correct, it is determined that the video data to be transmitted fails to be authenticated or has been tampered with, and the video data to be transmitted is prohibited from being played.

In the fourth aspect, the embodiment of the present application provides a video data certificate storage device, which is applied to the first client, and the device includes:

A model acquisition module, used to obtain a pre-trained scene type recognition model set for video data when the video data to be transmitted is collected;

The scene type identification module is used to extract a plurality of key video frames in the video data, and determine the scene type of each key video frame based on a pre-trained scene type identification model;

Mask image construction module, used to load the driving data of the current vehicle, and construct a mask image with parameters according to the driving data and the scene type of each key video frame;

The video synthesis module is used to synthesize the mask image carrying parameters and each key video frame to generate synthesized target video data;

The video sending module is used to send the video data and driving data to be transmitted to the second client, and send the target video data to the cloud server together with the driving data after processing.

In a fifth aspect, embodiments of the present application provide a computer storage medium, where a plurality of instructions are stored in the computer storage medium, and the instructions are adapted to be loaded by a processor and execute the above method steps.

In a sixth aspect, an embodiment of the present application provides a terminal, which may include: a processor and a memory; wherein, the memory stores a computer program, and the computer program is adapted to be loaded by the processor and execute the above method steps.

The technical solutions provided by the embodiments of the present application may include the following beneficial effects:

In the embodiment of the present application, the video data storage device first obtains the pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected, and then extracts multiple key video frames in the video data, and based on The model determines the scene type of each key video frame, then loads the driving data of the current vehicle, and constructs a mask image with parameters according to the driving data and the scene type of each key video frame, and then combines the mask image with parameters and Each key video frame is synthesized to generate synthesized target video data, and finally the video data and driving data to be transmitted are sent to the second client, and the processed target video data and driving data are sent to the cloud server. Since the application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention.

Description of drawings

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

Fig. 1 is a schematic flow diagram of a video data storage method applied to a first client provided by an embodiment of the present application;

Fig. 2 is a schematic flow diagram of a video data storage method applied to a cloud server provided by an embodiment of the present application;

FIG. 3 is a schematic flow diagram of a video data storage method applied to a second client provided by an embodiment of the present application;

Fig. 4 is a schematic flow chart of a video data storage method provided by an embodiment of the present application;

Fig. 5 is a schematic structural diagram of a video data storage device provided by an embodiment of the present application;

FIG. 6 is a schematic structural diagram of a terminal provided by an embodiment of the present application.

Detailed ways

The following description and drawings illustrate specific embodiments of the invention sufficiently to enable those skilled in the art to practice them.

It should be clear that the described embodiments are only some of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of the present invention.

When the following description refers to the accompanying drawings, the same numerals in different drawings refer to the same or similar elements unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with aspects of the invention as recited in the appended claims.

In the description of the present invention, it should be understood that the terms "first", "second" and so on are used for descriptive purposes only, and cannot be interpreted as indicating or implying relative importance. Those of ordinary skill in the art can understand the specific meanings of the above terms in the present invention in specific situations. In addition, in the description of the present invention, unless otherwise specified, "plurality" means two or more. "And/or" describes the association relationship of associated objects, indicating that there may be three types of relationships, for example, A and/or B may indicate: A exists alone, A and B exist simultaneously, and B exists independently. The character "/" generally indicates that the contextual objects are an "or" relationship.

The present application provides a video data certificate storage method, device, storage medium and terminal to solve the problems existing in the above related technical problems. In the technical solution provided by this application, since this application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved. The following uses an example A specific example will be described in detail.

The video data storage method provided by the embodiment of the present application will be described in detail below in conjunction with accompanying drawings 1 to 4 . The method can be implemented relying on a computer program, and can run on a video data storage device based on the von Neumann system. The computer program can be integrated in the application, or run as an independent utility application.

Please refer to FIG. 1 , which provides a schematic flowchart of a method for video data storage in an embodiment of the present application, which is applied to a first client. As shown in Figure 1, the method of the embodiment of the present application may include the following steps:

S101, when the video data to be transmitted is collected, obtain a pre-trained scene type recognition model set for the video data;

Generally, the scene type identification model is a mathematical model for identifying the scene type, which can output the scene type of the video image, and the model is created using the YOLOv5 algorithm.

In the embodiment of the present application, when generating the pre-trained scene type recognition model, the scene image where the vehicle is located is first collected to obtain a model training sample; wherein, the scene image at least includes a vehicle unloading scene, a vehicle refueling scene, a vehicle driving scene and Vehicle accident scene, then use the YOLOv5 algorithm to create a scene type recognition model, then input the model training samples into the scene type recognition model for model training, output the loss value, and finally generate a pre-trained scene type recognition model when the loss value reaches the minimum.

Further, when the loss value does not reach the minimum, the loss value is backpropagated to adjust the model parameters of the scene type recognition model, and the model training samples are continuously input into the scene type recognition model for model training.

In a possible implementation, when the truck driver is in a scene such as loading and unloading, traffic jam, accident, refueling, attendance check-in, etc., first collect the video of the current scene through the watermark camera on the mobile terminal to obtain the video data to be transmitted, At this point, a pre-trained scene type recognition model set for the video data is acquired, and the model can identify the scene type of the video scene.

S102, extracting a plurality of key video frames in the video data, and determining the scene type of each key video frame based on a pre-trained scene type recognition model;

Wherein, the key video frame is a plurality of high-definition video images selected from the data.

In the embodiment of the present application, when extracting multiple key video frames in the video data, the image parameters of each video image frame in the video data, such as sharpness, brightness, exposure, etc., are first calculated, and then according to each video image frame The image parameter calculates the weight value of each video image frame, and compares the weight value of each video image frame with the preset weight value, and determines the video image frame greater than the preset weight value as the key video frame.

Further, the scene type recognition model includes an input terminal, a reference network, a Neck network and a Head output terminal.

Specifically, when determining the scene type of each key video frame based on the pre-trained scene type recognition model, the input terminal receives each key video frame, and scales each key video frame to a preset size for normalization, The normalized video frame is obtained; the benchmark network performs feature extraction on the normalized video frame to obtain a feature map set; the Neck network fuses each feature map in the feature map set with the preset basic features to obtain the fused The feature map of the head output uses the classification branch to classify the fused feature map, and uses the regression branch to perform linear regression on the classified type to obtain the scene type of each key video frame.

S103, load the driving data of the current vehicle, and construct a mask image carrying parameters according to the driving data and the scene type of each key video frame;

In a possible implementation, when constructing a mask image with parameters, first obtain the mask image, then identify the first parameter identification set on the mask image, and then identify the driving data and the scene type of each key video frame For the corresponding second parameter identification set, finally identify the parameter identifications from the first parameter identification set that are the same as the parameter identifications in the second parameter identification set for data mapping, and generate a mask image carrying parameters.

Specifically, the driving data includes the real-time latitude and longitude of the Beidou terminal of the truck, reverse geocoding, license plate number, GPS latitude and longitude information of the mobile phone, basic user information, and information manually entered by the user.

S104, synthesizing the mask image carrying the parameters and each key video frame to generate synthesized target video data;

In a possible implementation, after the mask image carrying parameters is generated, the mask image carrying parameters can be synthesized with each key video frame to obtain each key video frame after synthesis, and finally the synthesized The video composed of each key video frame is determined as the target video data for synthesis.

S105. Send the video data to be transmitted and the driving data to the second client, and send the target video data to the cloud server together with the driving data after processing.

In a possible implementation manner, after obtaining the synthesized target video data, the video data and driving data to be transmitted can be sent to the second client, that is, the cargo owner or the vehicle owner.

In a possible implementation, when the target video data is processed and sent to the cloud server together with the driving data, the digital watermark image is first obtained, and then a square RGB image is intercepted from the image of the target video data and the digital watermark image respectively. The first image and the second image are obtained, and then the first image is subjected to color channel separation to obtain the first color component set, and the second image is subjected to color channel separation to obtain the second color component set, and then the first color component set is obtained Arnold transformation is performed to obtain the transformation matrix, followed by DCT transformation of the second color component set according to the transformation matrix to obtain the DC component, and then the digital watermark is embedded in the target video data according to the transformation matrix and the DC component to generate the processed video data, and finally the The processed video data and driving data are sent to the cloud server.

Specifically, a square RGB image is selected from the image of the target video data and the digital watermark image, and the side length of the carrier image is a multiple of 8 and the side length of the carrier image is 8 times the side length of the watermark image. The three color channels of the carrier image are separated to obtain the three color components of IR, IG and IB; the three color channels of the digital watermark image are separated to obtain the three color components of WR, WG and WB.

Perform Arnold transformation on the three color components of WR, WG, and WB (the transformation process can be regarded as the process of stretching, compression, folding, and splicing) to obtain WRA, WGA, and WBA after transformation.

Each component of the image of the target video data is divided into several sub-blocks with a size of 8×8 as a unit. Considering the sub-blocks as a whole, the position coordinates of the upper left sub-block are (1,1), the position coordinates of the adjacent right sub-blocks are (1,2), and so on. Apply DCT transformation to each sub-block separately, and then take out the DC component in the upper left corner of each sub-block after transformation to form a new matrix, and the DC component of the sub-block whose position coordinates are (1,1) is used as the new matrix (1,1 ), the DC component of the sub-block whose position coordinates are (1,2) is used as the element of the new matrix (1,2), and so on. The resulting matrices are called DC component matrices IRD, IBD, and IGD.

Embed the watermark on the DC component matrix, the embedding method is to add/subtract the brightness of k times of the scrambled WRA, WGA, WBA components. The component extraction formula is as follows:

IRDE=IRD+k×WRA; IGDE=IGD×WGA; IBDE=IBD+k×WBA;

Finally, the DC component matrices IRDE, IBDE, and IGDE after embedding the watermark replace the DC components of each sub-block according to the corresponding positions, and then apply the inverse DCT transformation to each sub-block to complete the watermark embedding of each color component, and obtain the processed video data .

Specifically, this application obtains driving data based on the real-time location of the Beidou terminal of the truck, the GPS location of the user's mobile phone, the basic information of the vehicle, and the user information, and automatically identifies the current user's operation scene. Finally, the blockchain technology and video digital watermarking technology can be used. It provides a safe and convenient tamper-proof video reporting method specifically for truck driver reporting.

In the embodiment of the present application, the video data storage device first obtains the pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected, and then extracts multiple key video frames in the video data, and based on The model determines the scene type of each key video frame, then loads the driving data of the current vehicle, and constructs a mask image with parameters according to the driving data and the scene type of each key video frame, and then combines the mask image with parameters and Each key video frame is synthesized to generate synthesized target video data, and finally the video data and driving data to be transmitted are sent to the second client, and the processed target video data and driving data are sent to the cloud server. Since the application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved.

Please refer to FIG. 2 , which provides a schematic flowchart of a video data storage method for an embodiment of the present application, which is applied to a cloud server. As shown in Figure 2, the method of the embodiment of the present application may include the following steps:

S201, receiving the processed video data and driving data sent by the first client to the cloud server;

S202, converting the processed video data into binary data;

S203, performing SHA256 hash operation on the binary data and the driving data to obtain a first hash string;

S204. Save the first hash string to the blockchain.

In the embodiment of the present application, the video data storage device first obtains the pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected, and then extracts multiple key video frames in the video data, and based on The model determines the scene type of each key video frame, then loads the driving data of the current vehicle, and constructs a mask image with parameters according to the driving data and the scene type of each key video frame, and then combines the mask image with parameters and Each key video frame is synthesized to generate synthesized target video data, and finally the video data and driving data to be transmitted are sent to the second client, and the processed target video data and driving data are sent to the cloud server. Since the application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved.

Please refer to FIG. 3 , which provides a schematic flowchart of a video data storage method for an embodiment of the present application, which is applied to a cloud server. As shown in Figure 3, the method of the embodiment of the present application may include the following steps:

S301, when receiving the video data and driving data to be transmitted sent by the first client to the second client, establish communication with the cloud server and obtain the first hash string stored in the block chain;

S302, performing SHA256 hash operation on the video data to be transmitted and the driving data to obtain a second hash string;

S303, when the first hash character string is the same as the second hash character string and the digital watermark in the video data to be transmitted is correct, play the video data to be transmitted; or, when the first hash character string is the same as the second hash character string When the character strings are the same and the digital watermark in the video data to be transmitted is correct, it is determined that the video data to be transmitted fails to be authenticated or has been tampered with, and the video data to be transmitted is prohibited from being played.

For example, as shown in Figure 4, Figure 4 is a schematic block diagram of a video data storage process provided by the present application. On the driver's first client, the first client installed on it should be automatically downloaded and identified after the program is started. model, and then automatically obtain the longitude and latitude of the mobile phone location and the longitude and latitude of the Beidou terminal location of the truck through the freight big data to obtain the driving data, and then the first client starts the watermark camera to record the video, extracts the key video frames in the video and inputs them into the downloaded recognition model to identify The scene type of each key video frame, according to the scene type and driving data to obtain a mask image with parameters, and then combine the mask image with each key video frame to obtain a composite video, and then process the composite video with the driving data Send it to the cloud server together, and send the video and driving data that have just been recorded to the second client.

After the cloud server receives the processed synthetic video and driving data, it performs SHA256 hash operation on the processed synthetic video and driving data to obtain a hash string, and puts the hash string into the blockchain to prevent tamper.

When the second client receives the newly recorded video and driving data, it performs SHA256 hash operation on the newly recorded video and driving data to obtain the target hash string, and obtains the hash string stored on the blockchain , and finally compare whether the target hash string is consistent with the hash string stored on the blockchain. If they are consistent and the digital watermark is correct, play the video that has just been recorded. Otherwise, the authentication fails or the video has been tampered with, then the end .

In the embodiment of the present application, the video data storage device first obtains the pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected, and then extracts multiple key video frames in the video data, and based on The model determines the scene type of each key video frame, then loads the driving data of the current vehicle, and constructs a mask image with parameters according to the driving data and the scene type of each key video frame, and then combines the mask image with parameters and Each key video frame is synthesized to generate synthesized target video data, and finally the video data and driving data to be transmitted are sent to the second client, and the processed target video data and driving data are sent to the cloud server. Since the application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved.

The following are device embodiments of the present invention, which can be used to implement the method embodiments of the present invention. For the details not disclosed in the device embodiment of the present invention, please refer to the method embodiment of the present invention.

Please refer to FIG. 5 , which shows a schematic structural diagram of a video data storage device provided by an exemplary embodiment of the present invention. The video data storage device can be implemented as all or a part of the terminal through software, hardware or a combination of the two. The device 1 includes a model acquisition module 10 , a scene type identification module 20 , a mask picture construction module 30 , a video synthesis module 40 , and a video sending module 50 .

Model acquiring module 10, for when collecting the video data to be transmitted, obtain the pre-trained scene type recognition model set for video data;

The scene type identification module 20 is used to extract a plurality of key video frames in the video data, and determines the scene type of each key video frame based on the pre-trained scene type identification model;

Mask picture construction module 30, is used to load the driving data of current vehicle, and constructs the mask picture carrying parameter according to the scene type of driving data and each key video frame;

Video synthesis module 40 is used to synthesize the mask picture carrying parameters and each key video frame to generate synthesized target video data;

The video sending module 50 is configured to send the video data and driving data to be transmitted to the second client, and send the target video data to the cloud server together with the driving data after processing.

It should be noted that, when the video data storage device provided by the above-mentioned embodiments executes the video data storage method, it only uses the division of the above-mentioned functional modules for illustration. In practical applications, the above-mentioned functions can be assigned by different The functional modules are completed, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the video data certificate storage device provided in the above embodiments and the video data certificate storage method embodiment belong to the same idea, and the implementation process thereof is detailed in the method embodiment, and will not be repeated here.

The serial numbers of the above embodiments of the present application are for description only, and do not represent the advantages and disadvantages of the embodiments.

In the embodiment of the present application, the video data storage device first obtains the pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected, and then extracts multiple key video frames in the video data, and based on The model determines the scene type of each key video frame, then loads the driving data of the current vehicle, and constructs a mask image with parameters according to the driving data and the scene type of each key video frame, and then combines the mask image with parameters and Each key video frame is synthesized to generate synthesized target video data, and finally the video data and driving data to be transmitted are sent to the second client, and the processed target video data and driving data are sent to the cloud server. Since the application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved.

The present invention also provides a computer-readable medium on which program instructions are stored. When the program instructions are executed by a processor, the video data storage methods provided by the above method embodiments are implemented.

The present invention also provides a computer program product containing instructions, which, when running on a computer, causes the computer to execute the video data storage method of each method embodiment above.

Referring to FIG. 6 , it provides a schematic structural diagram of a terminal according to an embodiment of the present application. As shown in FIG. 6 , a terminal 1000 may include: at least one processor 1001 , at least one network interface 1004 , a user interface 1003 , a memory 1005 , and at least one communication bus 1002 .

Wherein, the communication bus 1002 is used to realize connection and communication between these components.

Wherein, the user interface 1003 may include a display screen (Display) and a camera (Camera), and the optional user interface 1003 may also include a standard wired interface and a wireless interface.

Wherein, the network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

Wherein, the processor 1001 may include one or more processing cores. The processor 1001 uses various interfaces and lines to connect various parts of the entire electronic device 1000, and by running or executing instructions, programs, code sets or instruction sets stored in the memory 1005, and calling data stored in the memory 1005, execute Various functions of the electronic device 1000 and processing data. Optionally, the processor 1001 may use at least one of Digital Signal Processing (Digital Signal Processing, DSP), Field-Programmable Gate Array (Field-Programmable Gate Array, FPGA), and Programmable Logic Array (Programmable Logic Array, PLA). implemented in the form of hardware. The processor 1001 may integrate one or a combination of a central processing unit (Central Processing Unit, CPU), an image processor (Graphics Processing Unit, GPU), a modem, and the like. Among them, the CPU mainly handles the operating system, user interface and application programs, etc.; the GPU is used to render and draw the content that needs to be displayed on the display screen; the modem is used to handle wireless communication. It can be understood that the above modem may also not be integrated into the processor 1001, but implemented by a single chip.

Wherein, the memory 1005 may include a random access memory (Random Access Memory, RAM), or may include a read-only memory (Read-Only Memory). Optionally, the storage 1005 includes a non-transitory computer-readable storage medium (non-transitory computer-readable storage medium). The memory 1005 may be used to store instructions, programs, codes, sets of codes or sets of instructions. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playback function, an image playback function, etc.), Instructions and the like for implementing the above method embodiments; the storage data area can store the data and the like involved in the above method embodiments. Optionally, the memory 1005 may also be at least one storage device located away from the aforementioned processor 1001 . As shown in FIG. 6 , the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and a video data storage application program.

In the terminal 1000 shown in FIG. 6 , the user interface 1003 is mainly used to provide the user with an input interface to obtain the data input by the user; and the processor 1001 can be used to call the video data storage application program stored in the memory 1005, and Specifically perform the following operations:

When the video data to be transmitted is collected, a pre-trained scene type recognition model set for the video data is obtained;

Extract multiple key video frames in the video data, and determine the scene type of each key video frame based on the pre-trained scene type recognition model;

Load the driving data of the current vehicle, and construct a mask image with parameters according to the driving data and the scene type of each key video frame;

Synthesize the mask image carrying parameters with each key video frame to generate synthesized target video data;

Send the video data and driving data to be transmitted to the second client, and send the target video data to the cloud server together with the driving data after processing.

In one embodiment, when the processor 1001 generates the pre-trained scene type recognition model, it specifically performs the following operations:

Collecting scene images where the vehicle is located to obtain model training samples; where the scene images at least include vehicle unloading scenes, vehicle refueling scenes, vehicle driving scenes, and vehicle accident scenes;

Use the YOLOv5 algorithm to create a scene type recognition model;

Input the model training samples into the scene type recognition model for model training, and output the loss value;

When the loss value reaches the minimum, generate a pre-trained scene type recognition model;

or,

When the loss value does not reach the minimum, the loss value is backpropagated to adjust the model parameters of the scene type recognition model, and continue to input the model training samples into the scene type recognition model for model training.

In one embodiment, when the processor 1001 determines the scene type of each key video frame based on the pre-trained scene type identification model, it specifically performs the following operations:

The input terminal receives each key video frame, scales each key video frame to a preset size, and performs normalization to obtain a normalized video frame;

The benchmark network performs feature extraction on the normalized video frames to obtain a set of feature maps;

The Neck network fuses each feature map in the feature map set with the preset basic features to obtain the fused feature map;

The head output uses the classification branch to classify the fused feature maps, and uses the regression branch to perform linear regression on the classified types to obtain the scene type of each key video frame.

In one embodiment, when the processor 1001 constructs a mask picture carrying parameters according to the driving data and the scene type of each key video frame, it specifically performs the following operations:

Get the mask image;

Identify the first parameter identification set on the mask image;

Identify the second parameter identification set corresponding to the driving data and the scene type of each key video frame;

Identify the same parameter identifiers from the first parameter identifier set as the parameter identifiers in the second parameter identifier set to perform data mapping, and generate a mask picture carrying parameters.

In one embodiment, when the processor 1001 executes processing the target video data and sending it to the cloud server together with the driving data, it specifically performs the following operations:

Obtain digital watermark image;

Intercept square RGB images from the image of the target video data and the digital watermark image respectively to obtain the first image and the second image;

performing color channel separation on the first image to obtain a first color component set, and performing color channel separation on the second image to obtain a second color component set;

A transformation matrix is obtained after performing Arnold transformation on the first color component set;

performing DCT transformation on the second color component set according to the transformation matrix to obtain a DC component;

According to the transformation matrix and the DC component, the digital watermark is embedded into the target video data, and the processed video data is generated;

Send the processed video data and driving data to the cloud server.

In the embodiment of the present application, the video data storage device first obtains the pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected, and then extracts multiple key video frames in the video data, and based on The model determines the scene type of each key video frame, then loads the driving data of the current vehicle, and constructs a mask image with parameters according to the driving data and the scene type of each key video frame, and then combines the mask image with parameters and Each key video frame is synthesized to generate synthesized target video data, and finally the video data and driving data to be transmitted are sent to the second client, and the processed target video data and driving data are sent to the cloud server. Since the application uses the model to identify the scene type of the video data, and combines the driving data to perform secondary synthesis on the video data, the video reported by the driver is not easy to be tampered with, and the authenticity of the video is improved.

Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be implemented through computer programs to instruct related hardware. The program for video data storage can be stored in a computer-readable storage medium. When the program is executed, it may include the processes of the embodiments of the above-mentioned methods. Wherein, the storage medium may be a magnetic disk, an optical disk, a read-only memory or a random access memory, and the like.

The above disclosures are only preferred embodiments of the present application, which certainly cannot limit the scope of the present application. Therefore, equivalent changes made according to the claims of the present application still fall within the scope of the present application.

Claims (10)

1. A method for depositing video data, characterized in that it is applied to the first client, and the method comprises: When the video data to be transmitted is collected, a pre-trained scene type recognition model set for the video data is obtained; Extracting a plurality of key video frames in the video data, and determining the scene type of each key video frame based on the pre-trained scene type recognition model; Load the driving data of the current vehicle, and construct a mask picture carrying parameters according to the driving data and the scene type of each key video frame; Synthesizing the mask image carrying parameters and each key video frame to generate synthesized target video data; Send the video data to be transmitted and the driving data to the second client, and send the target video data to the cloud server together with the driving data after processing. 2. The method according to claim 1, characterized in that generating a pre-trained scene type recognition model according to the following steps, comprising: Collecting scene images where the vehicle is located to obtain model training samples; wherein, the scene images at least include a vehicle unloading scene, a vehicle refueling scene, a vehicle driving scene, and a vehicle accident scene; Use the YOLOv5 algorithm to create a scene type recognition model; Inputting the model training sample into the scene type recognition model for model training, and outputting a loss value; When the loss value reaches the minimum, generate a pre-trained scene type recognition model; or, When the loss value does not reach the minimum, the loss value is backpropagated to adjust the model parameters of the scene type recognition model, and continue to input the model training samples into the scene type recognition model for model training . 3. The method according to claim 1, wherein the scene type recognition model comprises an input terminal, a reference network, a Neck network and a Head output terminal; The scene type identification model based on the pre-training described determines the scene type of each key video frame, including: The input end receives each key video frame, and normalizes each key video frame after scaling to a preset size to obtain a normalized video frame; The benchmark network performs feature extraction on the normalized video frames to obtain a set of feature maps; The Neck network performs feature fusion on each feature map in the feature map set and preset basic features to obtain a fused feature map; The head output uses the classification branch to classify the fused feature maps, and uses the regression branch to perform linear regression on the classified types to obtain the scene type of each key video frame. 4. The method according to claim 1, wherein said building a mask picture carrying parameters according to the driving data and the scene type of each key video frame includes: Get the mask image; Identifying the first parameter identification set on the mask picture; Identifying the second parameter identification set corresponding to the driving data and the scene type of each key video frame; Identifying parameter identifiers from the first parameter identifier set that are the same as the parameter identifiers in the second parameter identifier set to perform data mapping to generate a mask picture carrying parameters. 5. The method according to claim 1, wherein said processing said target video data and sending said driving data to a cloud server includes: Obtain digital watermark image; Respectively intercepting square RGB images from the image of the target video data and the digital watermark image to obtain a first image and a second image; performing color channel separation on the first image to obtain a first color component set, and performing color channel separation on the second image to obtain a second color component set; A transformation matrix is obtained after performing Arnold transformation on the first color component set; performing DCT transformation on the second color component set according to the transformation matrix to obtain a DC component; Embedding a digital watermark into the target video data according to the transformation matrix and the DC component to generate processed video data; Sending the processed video data and the driving data to a cloud server. 6. The method according to claim 1, wherein it is applied to a cloud server, and the method comprises: receiving the processed video data and driving data sent by the first client for the cloud server; Convert the processed video data to binary data; performing SHA256 hash operation on the binary data and the driving data to obtain a first hash string; Save the first hash string to the block chain. 7. The method according to claim 6, wherein it is applied to the second client, and the method comprises: When receiving the video data and driving data to be transmitted sent by the first client for the second client, establish communication with the cloud server and obtain the first hash string stored in the block chain; Performing SHA256 hash operation on the video data to be transmitted and the driving data to obtain a second hash string; When the first hash character string is identical to the second hash character string and the digital watermark in the video data to be transmitted is correct, play the video data to be transmitted; Otherwise, it is determined that the video data to be transmitted fails to be authenticated or has been tampered with, and the video data to be transmitted is prohibited from being played. 8. A video data storage device, characterized in that it is applied to the first client, and the device includes: A model acquisition module, used to obtain a pre-trained scene type recognition model set for the video data when the video data to be transmitted is collected; A scene type identification module, configured to extract a plurality of key video frames in the video data, and determine the scene type of each key video frame based on the pre-trained scene type identification model; A mask picture construction module, used to load the driving data of the current vehicle, and construct a mask picture carrying parameters according to the driving data and the scene type of each key video frame; A video synthesis module, configured to synthesize the mask image carrying parameters and each key video frame to generate synthesized target video data; The video sending module is used to send the video data to be transmitted and the driving data to the second client, and send the target video data to the cloud server together with the driving data after processing. 9. A computer storage medium, characterized in that the computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor and executing the method steps according to any one of claims 1-7. 10. A terminal, characterized in that it comprises: a processor and a memory; wherein, the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing any one of claims 1-7 method steps.

Images