SDI equipment filtering reverse deduction method, device and computer equipment

Abstract

The embodiment of the present application provides a method, device and computer equipment for inverse derivation of SDI device filtering, which relates to the field of computer technology. The method defines a convolution kernel K and randomly initializes the convolution kernel K; decodes the original video file and obtains the decoded Y component image matrix I; performs a convolution operation on the image matrix I through the convolution kernel K to obtain a data matrix C to define the convolution output image; compares the convolution output image with the image of the original video file output by the SDI device through a loss function, and obtains the loss value of the loss function; determines the parameter optimization target, and obtains the minimum loss value based on the parameter optimization target through the gradient descent method to determine the inverse derivation value of the filtering parameter of the SDI device. The method predicts the filtering parameters used by the SDI broadcasting device through the numerical changes of the two images before and after filtering, thereby eliminating errors in picture quality assessment.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a method and apparatus for inverse filtering of an SDI device, and a computer device. Background Art

[0002] Serial Digital Interface (SDI) is a broadcast-grade digital video interface used to transmit uncompressed digital video data and is commonly found in various digital video devices.

[0003] In the broadcasting and television industry, there are often many brands of SDI video playback devices. These different devices will process the original video encoding files to a certain extent. The most common method is to perform filtering operations to eliminate noise.

[0004] While filtering is a means of improving image quality, it can be challenging to assess image quality. Existing evaluation algorithms can easily misinterpret numerical changes caused by filtering as image quality degradation, lowering the score. Summary of the Invention

[0005] In order to solve one of the above technical defects, an embodiment of the present application provides an SDI device filtering back-derivation method, device, computer device, and storage medium.

[0006] According to a first aspect of an embodiment of the present application, a method for inverse filtering of an SDI device is provided, comprising:

[0007] Define a convolution kernel K and randomly initialize the convolution kernel K;

[0008] Decode an original video file and obtain the decoded Y component image matrix I;

[0009] Perform convolution operation on the image matrix I through the convolution kernel K to obtain the data matrix C to define the convolution output image;

[0010] Comparing the convolution output image with the image of the original video file output by an SDI device through a loss function, and obtaining a loss value of the loss function;

[0011] The parameter optimization target is determined, and the minimum loss value based on the parameter optimization target is obtained by the gradient descent method to determine the filter parameter prediction value of the SDI device.

[0012] In an optional embodiment of the present application, the step of decoding the original video file and obtaining the decoded Y component image matrix I in the method further includes:

[0013] The image matrix I is an m*n matrix, and the data matrix C is obtained by the Kronecker product operation of the image matrix I and the convolution kernel K.

[0014] In an optional embodiment of the present application, before comparing the convolution output image with the image of the original video file output by the SDI device using a loss function and obtaining the loss value of the loss function, the method further includes:

[0015] The Y component data Y' of the frame image at the same position as that in the original video file is obtained from the SDI device to define a loss function according to the data matrix C and the Y component data Y'.

[0016] In an optional embodiment of the present application, the loss function in the method uses the MSE mean square error method to compare the convolution output image with the image of the original video file output by the SDI device.

[0017] In an optional embodiment of the present application, the step of obtaining the minimum loss value based on the parameter optimization objective by the gradient descent method in the method further includes:

[0018] Establish the gradient matrix of the loss function and perform operations to obtain the convolution kernel K target in the parameter optimization direction. The formula is:

[0019]

[0020] Where, is the gradient matrix of the loss function.

[0021] In an optional embodiment of the present application, the step of obtaining the minimum loss value based on the parameter optimization objective by the gradient descent method in the method further includes:

[0022] The minimum loss value based on the parameter optimization objective is obtained by iterative calculation based on the gradient descent method.

[0023] In an optional embodiment of the present application, the convolution kernel K in the method is a 5×5 matrix.

[0024] According to a second aspect of an embodiment of the present application, there is provided an SDI device filtering inverse derivation device, the device comprising an original video decoding module, a convolution operation module, a loss function module, and a parameter prediction module;

[0025] The original video decoding module is used to decode an original video file and obtain the decoded Y component image matrix I;

[0026] A convolution operation module is used to perform a convolution operation on the image matrix I based on a randomly initialized convolution kernel K to obtain a data matrix C to define a convolution output image;

[0027] A loss function module, configured to compare the convolution output image with an image of an original video file output by an SDI device using a loss function, and obtain a loss value of the loss function;

[0028] The parameter prediction module is used to determine the parameter optimization target and obtain the minimum loss value based on the parameter optimization target through the gradient descent method to determine the filter parameter prediction value of the SDI device.

[0029] According to a third aspect of an embodiment of the present application, there is provided a computer device, including: a memory;

[0030] A processor; and a computer program; wherein the computer program is stored in a memory and is configured to be executed by the processor to implement the steps of any method according to the first aspect of the embodiment of the present application.

[0031] According to a fourth aspect of the embodiments of the present application, a computer-readable storage medium is provided, on which a computer program is stored; the computer program is executed by a processor to implement the steps of any method of the first aspect of the embodiments of the present application.

[0032] The SDI device filtering inversion method provided in the embodiments of this application predicts the filtering parameters used by the SDI broadcast device by analyzing the numerical changes between the pre- and post-filtering images, thereby eliminating errors in image quality assessment. This facilitates quality assessment of file images, preventing the evaluation algorithm from interpreting numerical changes caused by filtering as image quality degradation, thereby obtaining an objective and accurate score. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0034] Figure 1 Flowchart of the SDI device filtering inverse derivation method provided in an embodiment of the present application;

[0035] Figure 2 A structural diagram of the SDI device filtering and reverse derivation device provided in an embodiment of the present application;

[0036] Figure 3 A schematic diagram of the computer device structure provided for one embodiment of the present application. DETAILED DESCRIPTION

[0037] In order to make the technical solutions and advantages of the embodiments of the present application more clearly understood, the exemplary embodiments of the present application are further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, and are not an exhaustive list of all the embodiments. It should be noted that the embodiments and features in the embodiments of the present application can be combined with each other unless they conflict.

[0038] In broadcasting, a wide variety of video playback devices are often used. These different devices often process the original video files to some degree. The most common approach is to filter the files to eliminate noise. While this improves visual quality, it creates significant challenges when evaluating the file's image quality. Evaluation algorithms can easily interpret changes in values caused by filtering as degraded image quality, thereby lowering the score.

[0039] Gradient descent is a first-order optimization algorithm, often also called the steepest descent method. To find a local minimum of a function using gradient descent, an iterative search must be performed in the direction opposite to the gradient (or approximate gradient) of the function at the current point, at a specified step distance. Conversely, if the search is performed in the positive direction of the gradient, the local maximum of the function will be approached; this process is called gradient ascent.

[0040] In response to the above problems, an embodiment of the present application provides an SDI device filtering inverse derivation method. The present application defines a convolution kernel to perform a convolution operation on the image, defines a loss function COST, compares the image convolved with the kernel with the image output by the SDI device, and gives an evaluation value; finally, by continuously iterating the above process, the value of the loss function becomes smaller and smaller, so that the convolution result becomes closer and closer to the actual SDI device output.

[0041] This application starts with a set of randomly initialized convolution kernel parameters and uses gradient descent to adjust the parameters to find the device filter parameters. Figure 1 Steps shown:

[0042] S1: Define a convolution kernel K and randomly initialize the convolution kernel K.

[0043] In a specific implementation, a convolution kernel K is defined and randomly initialized. Optionally, the value range of the convolution kernel is 0~1. Furthermore, the initial convolution kernel cannot achieve the actual filtering effect of the SDI device. Based on the iterative method in the subsequent steps of this application, the convolution kernel K is iteratively optimized so that the image after convolution filtering is as close as possible to the SDI output image, and the filtering of the SDI device will be smaller than the 5×5 specification. Therefore, the convolution kernel of this application can be 5×5, which can effectively simulate the convolution of a smaller device and improve the computing efficiency.

[0044] S2: Decode an original video file and obtain the decoded Y component image matrix I.

[0045] In a specific implementation, I is an m*n matrix.

[0046] S3: Perform a convolution operation on the image matrix I through the convolution kernel K to obtain the data matrix C to define the convolution output image.

[0047] In a specific implementation, a convolution operation is performed on the file-decoded image I using a convolution kernel K, and the data matrix C after the operation is saved.

[0048] In the embodiment of the present application, the convolution output image, that is, the data matrix C is defined as:

[0049]

[0050] Convolution operation: ;

[0051] Where I is the file input image, that is, the decoded Y component image matrix, and K is the convolution kernel matrix.

[0052] In a specific implementation, the calculation process of each element in the C matrix is:

[0053]

[0054] S4: Compare the convolution output image with the image of the original video file output by an SDI device through a loss function, and obtain a loss value of the loss function.

[0055] In a specific implementation, the Y component data Y' of the frame image at the same position as in the original video file is first obtained from the SDI device to define a loss function based on the data matrix C and the Y component data Y'; wherein Y' is an m*n matrix.

[0056] In the implementation, a loss function (COST) is defined that compares the image convolved with the K kernel with the image output by the SDI device and provides an evaluation value. The loss function uses the mean squared error (MSE) to define the difference between the convolved image and the SDI device image.

[0057] Loss function definition:

[0058]

[0059] S5: Determine the parameter optimization target and obtain the minimum loss value based on the parameter optimization target through the gradient descent method to determine the filter parameter prediction value of the SDI device.

[0060] In the specific implementation, to determine the parameter optimization target, it is necessary to find the optimal filtering parameter K' so that the cost value is minimized, that is:

[0061]

[0062] Furthermore, the gradient matrix of the loss function is established and calculated to obtain the convolution kernel K target in the parameter optimization direction. The formula is:

[0063]

[0064] Where, is the gradient matrix of the loss function.

[0065] Furthermore, the optimized K-core is obtained:

[0066]

[0067]

[0068] In the specific implementation, the above steps are continuously iterated to make the value of the loss function smaller and smaller, so that the convolution result is closer and closer to the actual SDI device output.

[0069] Based on the method of the embodiment of the present application, in practice, after W filtering and then differencing with the SDI image, the deviation can be significantly eliminated, and the mean square error (MSE) value is only 1.1.

[0070] It should be understood that, although the various steps in the flowchart are shown in sequence as indicated by the arrows, these steps are not necessarily performed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and these steps may be performed in other orders. Moreover, at least a portion of the steps in the figure may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily performed at the same time, but may be performed at different times. The execution order of these sub-steps or stages is not necessarily to be performed in sequence, but may be performed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0071] See Figure 2, an embodiment of the present application provides an SDI device filter inverse derivation device, including an original video decoding module 10, a convolution operation module 20, a loss function module 30 and a parameter prediction module 40;

[0072] The original video decoding module 10 is used to decode an original video file and obtain the decoded Y component image matrix I.

[0073] In a specific implementation, I is an m*n matrix.

[0074] The convolution operation module 20 is used to perform a convolution operation on the image matrix I based on a randomly initialized convolution kernel K to obtain a data matrix C to define a convolution output image.

[0075] In a specific implementation, a convolution kernel K is defined and randomly initialized. Optionally, the value range of the convolution kernel is 0~1. Furthermore, the initial convolution kernel cannot achieve the actual filtering effect of the SDI device. Based on the iterative method in the subsequent steps of this application, the convolution kernel K is iteratively optimized so that the image after convolution filtering is as close as possible to the SDI output image, and the filtering of the SDI device will be smaller than the 5×5 specification. Therefore, the convolution kernel of this application can be 5×5, which can effectively simulate the convolution of a smaller device and improve the computing efficiency.

[0076] In a specific implementation, a convolution operation is performed on the file-decoded image I using a convolution kernel K, and the data matrix C after the operation is saved.

[0077] In the embodiment of the present application, the convolution output image is defined, and the convolution operation is: ;

[0078] Where I is the file input image, that is, the decoded Y component image matrix, and K is the convolution kernel matrix.

[0079] The loss function module 30 is used to compare the convolution output image with the image of the original video file output by an SDI device through a loss function, and obtain the loss value of the loss function.

[0080] In a specific implementation, the Y component data Y' of the frame image at the same position as in the original video file is first obtained from the SDI device to define a loss function based on the data matrix C and the Y component data Y'; wherein Y' is an m*n matrix.

[0081] In the implementation, a loss function (COST) is defined that compares the image convolved with the K kernel with the image output by the SDI device and provides an evaluation value. The loss function uses the mean squared error (MSE) to define the difference between the convolved image and the SDI device image.

[0082] The parameter prediction module 40 is used to determine the parameter optimization target and obtain the minimum loss value based on the parameter optimization target by the gradient descent method to determine the filter parameter prediction value of the SDI device.

[0083] In a specific implementation, to determine the parameter optimization target, it is necessary to find the optimal filtering parameter K' so that the cost value is minimized.

[0084] Furthermore, the gradient matrix of the loss function is established and calculated to obtain the convolution kernel K target in the parameter optimization direction.

[0085] Furthermore, the optimized K-core is obtained.

[0086] In the specific implementation, the above steps are continuously iterated to make the value of the loss function smaller and smaller, so that the convolution result is closer and closer to the actual SDI device output.

[0087] For the specific limitations of the above-mentioned SDI device filter inversion derivation device, please refer to the limitations of the SDI device filter inversion derivation method above, and will not be repeated here. The various modules in the above-mentioned SDI device filter inversion derivation device can be implemented in whole or in part through software, hardware, or a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the corresponding operations of the above-mentioned modules.

[0088] In one embodiment, a computer device is provided, wherein the internal structure diagram of the computer device can be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data. The network interface of the computer device is used to communicate with an external terminal via a network connection. When the computer program is executed by the processor, it implements the above-mentioned SDI device filtering inverse derivation method. It includes: a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements any step in the above-mentioned video signal processing method.

[0089] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, any step in the above-mentioned SDI device filtering inverse derivation method can be implemented.

[0090] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application may be implemented in various computer languages, such as C, VHDL, Verilog, object-oriented programming language Java, and interpreted scripting language JavaScript.

[0091] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0092] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0093] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0094] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0095] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A method for inverse derivation of SDI equipment filtering, characterized in that: include: Define a convolution kernel K and randomly initialize the convolution kernel K; Decode an original video file and obtain the decoded Y component image matrix I; Performing a convolution operation on the image matrix I through the convolution kernel K to obtain a data matrix C to define a convolution output image; Comparing the convolution output image with an image of the original video file output by an SDI device through a loss function, and obtaining a loss value of the loss function; A parameter optimization target is determined, and a minimum loss value based on the parameter optimization target is obtained by a gradient descent method to determine a reverse-derived value of a filter parameter of the SDI device.

2. The SDI device filtering inverse derivation method according to claim 1, characterized in that: The step of decoding the original video file and obtaining the decoded Y component image matrix I further includes: The image matrix I is an m*n matrix, and the data matrix C is obtained by performing a Kronecker product operation on the image matrix I and the convolution kernel K.

3. The SDI device filtering inverse derivation method according to claim 2, characterized in that: Before comparing the convolution output image with the image of the original video file output by the SDI device using the loss function and obtaining the loss value of the loss function, the method further includes: The Y component data Y' of the frame image at the same position as that in the original video file is obtained from the SDI device to define the loss function according to the data matrix C and the Y component data Y'.

4. The SDI device filtering inverse derivation method according to claim 3, characterized in that: The loss function uses an MSE mean square error method to compare the convolution output image with the image of the original video file output by the SDI device.

5. The SDI device filtering inverse derivation method according to claim 1 or 4, characterized in that: The step of obtaining the minimum loss value based on the parameter optimization target by the gradient descent method further includes: The gradient matrix of the loss function is established and calculated to obtain the convolution kernel K target in the parameter optimization direction. The formula is: Where, is the gradient matrix of the loss function.

6. The SDI device filtering inverse derivation method according to claim 5, characterized in that: The step of obtaining the minimum loss value based on the parameter optimization target by the gradient descent method further includes: The minimum loss value based on the parameter optimization target is obtained by iterative calculation based on the gradient descent method.

7. The SDI device filtering inverse derivation method according to claim 1, characterized in that: The convolution kernel K is a 5×5 matrix.

8. An SDI equipment filtering reverse derivation device, characterized in that: It includes original video decoding module, convolution operation module, loss function module and parameter prediction module; The original video decoding module is used to decode an original video file and obtain the decoded Y component image matrix I; a convolution operation module, configured to perform a convolution operation on the image matrix I based on a randomly initialized convolution kernel K to obtain a data matrix C to define a convolution output image; A loss function module, configured to compare the convolution output image with an image of the original video file output by an SDI device using a loss function, and obtain a loss value of the loss function; The parameter prediction module is used to determine the parameter optimization target and obtain the minimum loss value based on the parameter optimization target through the gradient descent method to determine the filter parameter prediction value of the SDI device.

9. A computer device, characterized in that: include: Memory; processor; as well as computer programs; The computer program is stored in the memory and configured to be executed by the processor to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that A computer program is stored thereon; the computer program is executed by a processor to implement the method according to any one of claims 1 to 7.

Images