Magnification processing method and device of image data, electronic equipment and storage medium

Abstract

The application provides an image data magnification processing method and device, electronic equipment and storage medium, relates to the technical field of image data processing, and the method comprises the following steps: acquiring an input image, and performing row buffering on the input image; combining pixel columns of a filter window output by row buffering of the input image into complete filter window data, and generating first original image coordinates of pixels corresponding to the complete filter window data in the input image; the filter window is a filter window of a magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels of the row buffering of the input image; and performing independent magnification filtering processing on the complete filter window data based on the first original image coordinates. The technical scheme provided by the application can solve the row buffering control problem in a multi-channel scene.

Description

Technical Field

[0001] This application relates to the field of image data processing technology, and in particular to a method, apparatus, electronic device, and storage medium for amplifying image data. Background Technology

[0002] Currently, in technical fields such as image rendering that require magnification algorithms, line buffers have become one of the core fundamental components for implementing these algorithms. Since magnification algorithms often face scenarios with multiple output channels in practical applications, and the overall number of line buffers corresponding to multiple channels is relatively large, multi-channel multiplexed line buffers have become a common design solution in the industry to avoid wasting hardware resources due to excessive increases in the number of line buffers. The core objective of multi-channel multiplexed line buffers is to optimize hardware layout through resource reuse, adapt to the implementation requirements of multi-channel magnification algorithms, and support the application of related technologies in various terminal devices and systems.

[0003] In related technologies, the line cache usage scheme for amplification algorithms generally adopts a "multi-channel multiplexing + independent reading" architecture, that is, multiple output channels share the same set of line cache resources, and each channel independently reads the corresponding data from the line cache according to its own data needs. However, this design scheme has significant technical defects: when the number of output channels is large, the line cache needs to interact with multiple channels simultaneously for data interface and timing coordination, which greatly increases the complexity of the line cache control logic; at the same time, each channel needs to be equipped with independent address generation logic, and the increase in the number of channels directly leads to a synchronous increase in the number of address generation logic, further exacerbating the complexity of hardware design; in addition, the number of interaction interfaces between the line cache and each channel increases with the number of channels, which not only consumes more hardware resources, but also makes subsequent expansion of the number of channels face multiple obstacles such as interface compatibility and timing adjustment, resulting in extremely poor flexibility.

[0004] In summary, the line cache usage schemes of amplification algorithms in related technologies will cause a series of problems when the number of channels increases, such as complex line cache control, logical redundancy in addresses, cumbersome interaction interfaces, and difficulty in channel expansion. This not only increases the difficulty of hardware design and implementation, but also limits the scalability and adaptability of the algorithm in multi-channel scenarios. There is an urgent need for an optimized line cache reuse design scheme to solve the line cache control problem in multi-channel scenarios. Summary of the Invention

[0005] The main objective of this application is to provide a method, apparatus, electronic device, and storage medium for image data magnification processing, aiming to solve the problem of line buffer control in multi-channel scenarios.

[0006] To achieve the above objectives, a first aspect of this application provides a method for magnifying image data, the method comprising:

[0007] Acquire the input image and perform row caching on the input image;

[0008] The pixel columns of the filter window output from the row buffer of the input image are combined into complete filter window data, and the first original image coordinates of the pixel corresponding to the complete filter window data in the input image are generated; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image;

[0009] The complete filter window data is subjected to multi-channel independent amplification and filtering processing based on the first original image coordinates.

[0010] In some embodiments, the multi-channel independent amplification and filtering processing of the complete filter window data based on the first original image coordinates includes:

[0011] The first original image coordinates are compared with the second original image coordinates based on the target channel; the target channel is any one of multiple image processing channels, and the second original image coordinates are the original image coordinates corresponding to the pixel to be processed in the target channel;

[0012] When the first original image coordinates are equal to the second original image coordinates, the complete filter window data is independently amplified and filtered based on the target channel.

[0013] In some embodiments, the method further includes:

[0014] The vertical boundary of the input image is mirror-filled based on the row buffer of the input image to obtain the complete pixel column of the filter window;

[0015] The row buffer of the input image outputs the complete pixel column in the current output row order; the pixel column of the filter window output by the row buffer of the input image includes the complete pixel column.

[0016] In some embodiments, the method further includes:

[0017] The system acquires the next row data request signal from each of the multiple image processing channels and synchronizes the next row data request signals. The multiple image processing channels determine whether they need the next row of data based on their own processing requirements and generate the next row data request signal accordingly.

[0018] If all subsequent data request signals after synchronization indicate that the next row of data is needed, the row buffer of the input image is switched to the next row to be output according to the current output row order to output the pixel column of the filter window;

[0019] If the next row data request signal after synchronization does not indicate that the next row of data is needed, the row buffer of the input image continues to output the pixel column corresponding to the current output row according to the current output row order.

[0020] In some embodiments, combining the pixel columns of the filter window output from the row buffer of the input image into complete filter window data includes:

[0021] If the filtering window exceeds the horizontal boundary of the input image, the pixel columns of the filtering window output from the row buffer of the input image are horizontally padded.

[0022] According to the size parameters of the filter window, the horizontally padded pixel columns are stitched together to obtain the complete filter window data.

[0023] In some embodiments, after performing multi-channel independent filtering on the complete filtered window data based on the first original image coordinates, the method further includes:

[0024] Acquire the reverse voltage signals of each of the multiple image processing channels;

[0025] When all the back pressure signals indicate that the channel is in a ready state, the pixel column of the filter window after the row buffer of the input image is shifted to the right is output, and the step of combining the pixel column of the filter window output from the row buffer of the input image into complete filter window data continues to be executed.

[0026] In some embodiments, the row caching of the input image includes:

[0027] The pixels of the input image are written to the corresponding row buffers in order, thus obtaining the row buffer of the input image; the base number of the row buffer of the input image is equal to the vertical number of the filter window plus 1, and the size of each row of pixels in the row buffer of the input image corresponds to the resolution of the input image.

[0028] To achieve the above objectives, a second aspect of this application provides an image data magnification processing apparatus, the apparatus comprising:

[0029] A line caching module is used to acquire the input image and cache the input image in lines;

[0030] A filter window generation module is used to combine the pixel columns of the filter window output from the row buffer of the input image into complete filter window data, and generate the first original image coordinates of the pixels corresponding to the complete filter window data in the input image; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image;

[0031] A multi-channel processing module is used to perform multi-channel independent amplification and filtering processing on the complete filter window data based on the first original image coordinates.

[0032] To achieve the above objectives, a third aspect of this application provides an electronic device, which includes a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the image data magnification processing method described in the first aspect.

[0033] To achieve the above objectives, a fourth aspect of the present application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the image data magnification processing method described in the first aspect.

[0034] To achieve the above objectives, a fifth aspect of the present application provides a computer program product, which includes a computer program that, when executed by a processor, implements the image data magnification processing method described in the first aspect above.

[0035] The image data magnification processing method, apparatus, electronic device, computer-readable storage medium, and computer program product proposed in this application acquire an input image and perform row buffering on the input image; combine the pixel columns of the filter window output from the row buffer of the input image into complete filter window data, and generate the first original image coordinates of the pixels corresponding to the complete filter window data in the input image; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image; and perform multi-channel independent magnification filtering processing on the complete filter window data based on the first original image coordinates.

[0036] Thus, this embodiment combines the pixel columns of the filter window output by the row buffer into complete filter window data, and simultaneously generates the first original image coordinates of the corresponding pixels in the input image. Then, based on these first original image coordinates, multi-channel independent amplification and filtering processing is performed on the complete filter window data. In this way, when the number of image processing channels for amplifying the input image increases, multiple image processing channels can reuse address generation logic, having minimal impact on the row buffer module. This reduces the amount of address generation logic required for the row buffer and lowers the complexity of row buffer control. In other words, this embodiment effectively solves the row buffer control problem in multi-channel scenarios, avoiding a chain reaction of problems such as complex row buffer control, redundant address generation logic, cumbersome interaction interfaces, and difficulties in channel expansion during image amplification processing. It reduces the difficulty of hardware design and implementation and improves the scalability and adaptability of the amplification algorithm in multi-channel scenarios. Attached Figure Description

[0037] Figure 1 A schematic diagram of a structure for image magnification processing by independently reading data from a multi-channel multiplexed line buffer;

[0038] Figure 2 A flowchart illustrating the steps of the image data magnification processing method provided in some embodiments of this application;

[0039] Figure 3 for Figure 2 A detailed flowchart of step S201;

[0040] Figure 4 A flowchart illustrating the steps of the image data magnification processing method provided in the embodiments of this application in other embodiments;

[0041] Figure 5 A flowchart illustrating the steps of the image data magnification processing method provided in some embodiments of this application;

[0042] Figure 6 for Figure 2 A detailed flowchart of step S202;

[0043] Figure 7 for Figure 2 A detailed flowchart of step S203;

[0044] Figure 8 A flowchart illustrating the steps of the image data magnification processing method provided in some embodiments of this application;

[0045] Figure 9A schematic diagram of the multi-channel amplification module structure involved in a complete embodiment of the image data amplification processing method provided in this application;

[0046] Figure 10 This is a schematic diagram of the structure of the image data magnification processing device provided in the embodiments of this application;

[0047] Figure 11 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0049] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0050] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0051] First, the overall concept of the image data magnification processing method provided in the embodiments of this application will be explained.

[0052] In technical fields such as image rendering that require magnification algorithms, a certain number of line buffers are typically used when implementing these algorithms. When the number of output channels is large, the overall number of lines is also large. Therefore, multiple channels often reuse the line buffer, with each channel independently reading the required data from the line buffer. For example... Figure 1 As shown, after the original image to be magnified is input pixel by pixel row by row, the data is first stored in the row buffer. The size of the row buffer is related to the magnification algorithm. Then, each channel generates a different read address to obtain the data corresponding to each channel. Then, the horizontal magnification logic and vertical magnification logic of each channel are used. Finally, the magnified image is output through the output interface of each channel.

[0053] Thus, since the row cache needs to interact with multiple channels in terms of data interface and timing, when the number of channels is large, the control of the row cache will be more complex, the number of channel address generation logics will also increase, the interaction interface between the row cache and the channels will also increase, and it will be more troublesome to expand the number of channels.

[0054] To address this, embodiments of this application provide an image data magnification processing method, apparatus, electronic device, computer-readable storage medium, and computer program product. The method involves acquiring an input image and performing row buffering on the input image; combining the pixel columns of the filter window output from the row buffer of the input image into complete filter window data, and generating the first original image coordinates of the pixels corresponding to the complete filter window data in the input image; the filter window is the filter window of a magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image; and performing multi-channel independent magnification filtering processing on the complete filter window data based on the first original image coordinates.

[0055] Thus, this embodiment combines the pixel columns of the filter window output by the row buffer into complete filter window data, and simultaneously generates the first original image coordinates of the corresponding pixels in the input image. Then, based on these first original image coordinates, multi-channel independent amplification and filtering processing is performed on the complete filter window data. In this way, when the number of image processing channels for amplifying the input image increases, multiple image processing channels can reuse address generation logic, having minimal impact on the row buffer module. This reduces the amount of address generation logic required for the row buffer and lowers the complexity of row buffer control. In other words, this embodiment effectively solves the row buffer control problem in multi-channel scenarios, avoiding a chain reaction of problems such as complex row buffer control, redundant address generation logic, cumbersome interaction interfaces, and difficulties in channel expansion during image amplification processing. It reduces the difficulty of hardware design and implementation and improves the scalability and adaptability of the amplification algorithm in multi-channel scenarios.

[0056] Next, the image data magnification processing method, apparatus, electronic device, computer-readable storage medium, and computer program product provided in the embodiments of this application will be specifically described through the following examples, and the image data magnification processing method provided in the embodiments of this application will be described in detail first.

[0057] It should be noted that in all specific embodiments of this application, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of this application require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to confirmation pages. Only after obtaining the user's separate permission or consent is the necessary user-related data required for the proper functioning of these embodiments acquired.

[0058] It should be noted that the image data magnification processing method provided in this application embodiment can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be an electronic device such as a mobile phone, tablet computer, laptop computer, or desktop computer. The server can be a server terminal device, configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms. The software can be an application implementing the image data magnification processing method, a computer program, and a storage medium carrying the computer program. It should be understood that, based on different design needs of actual applications, in different feasible embodiments, the terminal, server, and software applying the image data magnification processing method provided in this application embodiment can, of course, be other forms not listed here, and the image data magnification processing method provided in this application embodiment does not specifically limit these.

[0059] Furthermore, the image data magnification processing method provided in this application embodiment can also be used in numerous general-purpose or special-purpose computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer computer devices, personal computers (PCs), minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In a distributed computing environment, program modules can reside in local and remote computer storage media, including storage devices.

[0060] For ease of understanding and explanation, the following text will use the image data magnification processing method provided in the embodiments of this application for electronic devices as examples to describe in detail the various specific embodiments of this application. The implementation of the image data magnification processing method provided in the embodiments of this application for any of the above-described subject matter can refer to the specific embodiments of the image data magnification processing method for electronic devices described below.

[0061] Please refer to Figure 2 , Figure 2 The illustrated flowchart shows the steps of the image data magnification processing method provided in some embodiments of this application. It should be understood that, although... Figure 2 The flowcharts illustrating subsequent steps show the execution order of some method steps. However, based on different design needs in practical applications, the image data magnification processing method provided in this application embodiment can, of course, employ a different execution order of method steps than shown in the figures. That is, Figure 2 The order of the method steps shown does not constitute a limitation on the execution logic order of the image data magnification processing method provided in the embodiments of this application. Any other order based on... Figure 2 Reasonable changes to the sequence of steps shown should be included within the protection scope of the image data magnification processing method provided in the embodiments of this application.

[0062] like Figure 2 As shown, in some embodiments, the image data magnification processing method provided in this application may include steps S201 to S203 as shown below.

[0063] Step S201: Obtain the input image and perform row caching on the input image.

[0064] In the process of performing image magnification tasks based on magnification algorithms, electronic devices first acquire the original image to be magnified as the input image, and then input the input image row by row, pixel by pixel, into the system running the magnification algorithm to perform row buffering of the input image.

[0065] It should be noted that magnification algorithms can include ESRGAN, R-ESRGAN, and LDSR. For example, electronic devices can use ESRGAN or R-ESRGAN when magnifying portrait or landscape images. As another example, R-ESRGAN anime6B can be used when magnifying anime images. Furthermore, if an electronic device needs to complete image magnification within a tight timeframe, it can choose a faster magnification algorithm such as BSRGAN, while if better image quality is required, an LDSR magnification algorithm can be used.

[0066] In some embodiments, the electronic device can select different magnification algorithms to perform small-scale magnification processing on the input image, and determine the magnification algorithm that best meets the requirements by comparing the final output effects of different algorithms.

[0067] Step S202: Combine the pixel columns of the filter window output from the row buffer of the input image into complete filter window data, and generate the first original image coordinates of the pixel corresponding to the complete filter window data in the input image; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image.

[0068] After the electronic device writes the input image into the row buffer, it outputs the pixel columns of the filter window corresponding to each row of pixels in sequence (from left to right in the horizontal direction and from top to bottom in the vertical direction). In this way, the electronic device combines the pixel columns of the filter window output by the row buffer of the input image into complete filter window data, and at the same time generates the first original image coordinates of the pixel corresponding to the current filter window data in the input image.

[0069] In some embodiments, the line buffer of the input image can be the line buffer module in the system that runs the magnification algorithm described above. The system may also include a filter window generation module. The electronic device can use the filter window generation module to combine the pixel columns of the filter window input by the line buffer into complete filter window data, and at the same time generate the first original image coordinates of the pixels corresponding to the filter window data in the input image.

[0070] Step S203: Perform multi-channel independent amplification and filtering processing on the complete filter window data based on the first original image coordinates.

[0071] After obtaining the complete filter window data, the electronic device further distributes the complete filter window data to the processing modules of all image processing channels simultaneously. These processing modules then perform coordinate comparison based on the first original image coordinates, and determine the data required for each channel based on the comparison results for amplification and filtering processing. This achieves multi-channel independent amplification and filtering processing of the complete filter window data.

[0072] In some embodiments, when the electronic device performs multi-channel independent amplification and filtering processing on complete filter window data through multiple image processing channels, the processing module of each channel that receives the complete filter window data compares the second original image coordinates generated by the coordinate transformation module (the second original image coordinates generated by the coordinate transformation module are different for different channels) with the first original image coordinates input from the filter window generation module. Only when the two coordinates are equal will the received data be subjected to the corresponding window filtering processing.

[0073] In this embodiment, the original image to be magnified is acquired as the input image via an electronic device. Then, the input image is input pixel-by-pixel, row by row, to the system running the magnification algorithm for row buffering. Subsequently, the row buffers of the input image are output sequentially (from left to right horizontally and from top to bottom vertically) as pixel columns of the filtering window corresponding to each row of pixels. The electronic device combines these pixel columns of the filtering window output from the row buffers of the input image into complete filtering window data, and simultaneously generates the first original image coordinates of the pixel corresponding to the current filtering window data in the input image. Finally, the electronic device simultaneously distributes the complete filtering window data to all image processing channels, allowing these channels to perform coordinate comparisons based on the first original image coordinates and determine the data they need for magnification and filtering based on the comparison results. This achieves multi-channel independent magnification and filtering processing of the complete filtering window data.

[0074] Thus, this embodiment combines the pixel columns of the filter window output by the row buffer into complete filter window data, and simultaneously generates the first original image coordinates of the corresponding pixels in the input image. Then, based on these first original image coordinates, multi-channel independent amplification and filtering processing is performed on the complete filter window data. In this way, when the number of image processing channels for amplifying the input image increases, multiple image processing channels can reuse address generation logic, having minimal impact on the row buffer module. This reduces the amount of address generation logic required for the row buffer and lowers the complexity of row buffer control. In other words, this embodiment effectively solves the row buffer control problem in multi-channel scenarios, avoiding a chain reaction of problems such as complex row buffer control, redundant address generation logic, cumbersome interaction interfaces, and difficulties in channel expansion during image amplification processing. It reduces the difficulty of hardware design and implementation and improves the scalability and adaptability of the amplification algorithm in multi-channel scenarios.

[0075] Please refer to Figure 3 , Figure 3 for Figure 2 A detailed flowchart of step S201.

[0076] like Figure 3 As shown, in some embodiments, step S201 described above may include step S301 as shown below.

[0077] Step S301: Write the pixels of the input image into the corresponding row buffer row by row to obtain the row buffer of the input image; the base number of the row buffer of the input image is equal to the vertical number of the filter window plus 1, and the size of each row of pixels in the row buffer of the input image corresponds to the resolution of the input image.

[0078] It should be noted that the filter window can be the filter window of the magnification algorithm selected by the electronic device for performing the current image magnification task.

[0079] When an electronic device performs line buffering on an input image, it sequentially inputs the pixels of the input image into the corresponding line buffer of the system, row by row, to obtain the line buffer of the input image. The luminance and chroma channels of the input image can be processed independently. Furthermore, the line buffer configuration can be as follows: the base number of rows equals the number of vertical rows in the filtering window plus one, and the size of each row of pixels in the input image's line buffer corresponds to the resolution of the input image. That is: the base number of rows in the luminance line buffer = the number of vertical rows in the luminance direction in the filtering window of the amplification algorithm plus one row for performance optimization; for example, after performance optimization, the total number of rows in the luminance line buffer is fixed at 6 + 1 = 7 rows in the vertical luminance direction of the filtering window; the base number of rows in the chroma line buffer = the number of vertical rows in the chroma direction of the filtering window plus one row for performance optimization; for example, after performance optimization, the total number of rows in the chroma line buffer is fixed at 4 + 1 = 5 rows in the vertical chroma direction of the filtering window.

[0080] Furthermore, the size of each row of pixels in the line buffer of the input image is fixedly equal to the horizontal resolution of the original input image before magnification, regardless of the number of channels or the resolution of the magnified image.

[0081] In some embodiments, when writing an input image to a line buffer, the electronic device can write the pixels of the input image to the corresponding line buffer row by row (i.e., write luminance to the luminance buffer and chrominance to the chrominance buffer) until the line buffer is full of initial data.

[0082] Please refer to Figure 4 , Figure 4 The following are schematic flowcharts of the steps in some other embodiments of the image data magnification processing method provided in the embodiments of this application.

[0083] like Figure 4 As shown, in some embodiments, the image data magnification processing method provided in this application may further include steps S401 and S402 as shown below.

[0084] Step S401: Based on the row buffer of the input image, perform mirror filling processing on the vertical boundary of the input image to obtain the complete pixel column of the filter window;

[0085] Step S402: Output the complete pixel column based on the row buffer of the input image in the current output row order; the pixel column of the filter window output by the row buffer of the input image includes the complete pixel column.

[0086] After the electronic device writes the input image to the line buffer, when outputting the pixel column of the filter window corresponding to each row of pixels, it also performs mirroring processing on the vertical boundary of the input image to obtain the complete pixel column of the filter window. Then, it outputs this complete pixel column according to the current output row order (i.e., horizontally from left to right and vertically from top to bottom). At this point, the complete pixel column of the filter window is the pixel column of the filter window output by the aforementioned line buffer.

[0087] In some embodiments, when an electronic device performs mirroring on the vertical boundaries of an input image using a row buffer, it can use the row buffer to perform mirroring on the top / bottom boundary regions of the input image when the pixels required by the filtering window exceed the vertical range of the input image. This ensures that subsequent output pixel columns can support the entire filtering window.

[0088] Please refer to Figure 5 , Figure 5 A flowchart illustrating the steps of the image data magnification processing method provided in some embodiments of this application.

[0089] like Figure 5 As shown, in some embodiments, the image data magnification processing method provided in this application may further include steps S501 to S503 as shown below.

[0090] Step S501: Obtain the next row data request signal of each of the multiple image processing channels and synchronize the next row data request signal; the multiple image processing channels determine whether they need the next row data based on their own processing needs to generate the next row data request signal;

[0091] Step S502: When all the next row data request signals after synchronization indicate that the next row of data is needed, the row buffer of the input image is switched to the next row to be output according to the current output row order to output the pixel column of the filter window;

[0092] Step S503: If the next row data request signal after synchronization does not indicate that the next row of data is needed, the pixel column corresponding to the current output row continues to be output based on the row buffer of the input image in the order of the current output rows.

[0093] When an electronic device outputs the pixel column of a filter window corresponding to each row of pixels from the row buffer of the input image, in order to adapt to differences in vertical magnification (i.e., different channels have different vertical magnification, and the same row of pixels is used different times), the electronic device can control whether the row buffer switches to the next row through the next row data request signal (next_row signal) of each of the multiple image processing channels. That is, the electronic device first obtains the next_row signal output by each image processing channel before and after the start and end of the current row of pixels in the row buffer. Here, each channel can output a high-level next_row signal to indicate that this channel needs the next row of data. Then, the electronic device synchronizes the next_row signals of each channel. For example, when the filter window generation module generates the filter window for the first pixel of the current row, it performs a logical AND operation on the next_row signals of all channels, and the synchronized next_row signals are then sent to the row buffer. Here, a synchronized next_row signal of all 1s indicates that all channels need the next row of data, while a synchronized next_row signal of not all 1s indicates that some channels do not yet need the next row of data. Based on this, the electronic device can use row buffering to determine whether the synchronized next_row is all 1 after outputting the rightmost pixel column of the current row. If so, it can switch to the next row to be output according to the current output row order to output the pixel column of the filter window corresponding to the next row of pixels. If not (not all are 1), it can continue to output the pixel column of the filter window corresponding to the current output row of pixels according to the current output row order.

[0094] Please refer to Figure 6 , Figure 6 for Figure 2 A detailed flowchart of step S202.

[0095] like Figure 6 As shown, in some embodiments, the step of “combining the pixel columns of the filter window output by the row buffer of the input image into complete filter window data” in step S202 above may include steps S601 and S602 as shown below.

[0096] Step S601: When the filtering window exceeds the horizontal boundary of the input image, perform horizontal padding on the pixel columns of the filtering window output from the row buffer of the input image;

[0097] Step S602: According to the size parameters of the filter window, the horizontally padded pixel columns are stitched together to obtain complete filter window data.

[0098] When combining the pixel columns of the filter window output from the line buffer into complete filter window data, the electronic device needs to perform horizontal mirroring at the left and right boundaries of the image. That is, for the left / right boundary regions of the input image, when the pixels required for the combined filter window exceed the horizontal range of the original input image, the pixel columns output from the line buffer are mirrored based on the excess portion to perform horizontal completion processing, ensuring the horizontal pixel integrity of the combined filter window. Then, the electronic device further combines the horizontally completed pixel column data output from the line buffer into a complete filter window according to the filter window size parameters required by the magnification algorithm (e.g., 6 vertical rows for luminance and 4 vertical rows for chroma) to obtain the complete filter window data.

[0099] Please refer to Figure 7 , Figure 7 for Figure 2 A detailed flowchart of step S203.

[0100] like Figure 7 As shown, in some embodiments, step S203 above: performing multi-channel independent amplification and filtering processing on the complete filter window data based on the first original image coordinates may include steps S701 and S702 as shown below.

[0101] Step S701: Compare the first original image coordinates with the second original image coordinates based on the target channel; the target channel is any one of multiple image processing channels, and the second original image coordinates are the original image coordinates corresponding to the pixel to be processed in the target channel;

[0102] Step S702: When the first original image coordinates are equal to the second original image coordinates, perform independent amplification and filtering processing on the complete filter window data based on the target channel.

[0103] When an electronic device performs multi-channel independent amplification and filtering on an input image through multiple image processing channels, each of the multiple image processing channels generates a second original image coordinate corresponding to the pixel to be processed in that channel through a coordinate transformation module. Thus, after receiving complete filtering window data, any target channel among the multiple image processing channels can first compare the second original image coordinate (the original image coordinate corresponding to the pixel to be processed in that channel, generated by the coordinate transformation module) with the first original image coordinate input by the filtering window generation module (the original image coordinate of the pixel corresponding to the complete filtering window data generated when combining the pixel columns input from the row buffer to form the complete filtering window data). Then, if the comparison finds that the two coordinates are completely equal, the target channel performs independent amplification and filtering on the complete filtering window data.

[0104] Please refer to Figure 8 , Figure 8 A flowchart illustrating the steps of the image data magnification processing method provided in some embodiments of this application.

[0105] like Figure 8 As shown, in some embodiments, after step S203 above: performing multi-channel independent amplification filtering processing on the complete filter window data based on the first original image coordinates, the image data amplification processing method provided in this application embodiment may further include steps S801 and S802 as shown below.

[0106] Step S801: Obtain the reverse voltage signals of each of the multiple image processing channels;

[0107] Step S802: When all the back pressure signals indicate that the channel is in a ready state, output the right-shifted pixel column of the filter window based on the row buffer of the input image, and return to the step of combining the pixel column of the filter window output from the row buffer of the input image into complete filter window data.

[0108] When an electronic device performs multi-channel independent magnification and filtering on an input image through multiple image processing channels, it can control whether to shift the filtering window to the right using a back pressure signal to accommodate differences in horizontal magnification (because different channels have different horizontal magnification, the number of times each channel uses the horizontal magnification for each pixel may differ). Specifically, each image processing channel sends a back pressure signal `chn_ready` to the filtering window generation module. Only when all channels' `chn_ready` values ​​indicate that the channel is ready (e.g., all are 1) will the filtering window shift to the right to output the pixel column of the right-shifted filtering window based on the row buffer of the input image (otherwise, the filtering window pauses, waiting for all channels to be ready). Then, the electronic device can return to step S102 to continue multi-channel independent magnification and filtering of the input image. This process is repeated until all channels have completed their respective magnification and filtering of the input image.

[0109] The following is a complete embodiment of the image data magnification processing method provided in this application.

[0110] Please refer to Figure 9 , Figure 9 This is a schematic diagram of the multi-channel amplification module structure involved in a complete embodiment of the image data amplification processing method provided in this application.

[0111] In this embodiment, the electronic device can... Figure 9The multi-channel amplification module structure shown performs multi-channel amplification processing on the original image. Therefore, when the number of channels increases, multiple channels reuse address generation logic, which has virtually no impact on the line buffer module, reducing the number of line buffer address generation logics and lowering the complexity of line buffer control. Furthermore, this embodiment allows for parameterization of the number of output channels. The only impact of increasing the number of channels on the line buffer is an increase in the number of next_row signals, and line buffer control based on these next_row signals is very simple.

[0112] During operation, the improved original image is input pixel-by-pixel, row by row, and the input original image is buffered. The number of rows in the line buffer is the same as the number of rows in the vertical direction of the filtering window of the upscaling algorithm; that is, there are 6 rows for luminance and 4 rows for chroma. To improve performance, one more row is added for both luminance and chroma, resulting in 7 rows for luminance and 5 rows for chroma. The size of each row in the line buffer corresponds to the resolution of the original image before upscaling. Regardless of the number of channels or the horizontal resolution of the upscaled image, the number of rows and the size of each row in the line buffer remain fixed.

[0113] In the horizontal direction, the row buffer outputs the pixel columns of the filter window corresponding to each row of pixels in left-to-right order (vertical mirroring is required at the top and bottom of the image). The filter window generation module mainly combines the pixel column data of the window input from the row buffer into a filter window (horizontal mirroring is required at the left and right boundaries of the image), and generates the coordinates of the pixel corresponding to the current window in the original image. Because the horizontal magnification factor is different for different channels, the number of times the horizontal direction is used for each pixel may be different for different channels. In this case, each channel needs to provide a backpressure signal chn_ready to the filter window generation module. The window will only move to the right when all channels' chn_ready is 1.

[0114] In the vertical direction, the row buffer outputs the pixel column of the filter window corresponding to each row of pixels in a top-left-to-bottom order. Because the vertical magnification factor is different for different channels, the number of times the vertical direction is used for each pixel may be different for different channels. In this case, each channel needs to give a next_row signal to the row buffer module before the row buffer finishes outputting the pixel column of the current row, indicating that the data of the next row of this channel needs the data of the next row. In this way, when the row buffer finishes outputting the rightmost pixel column of the current row, if the next_row signals of all channels are high, the row buffer will output the pixel column of the next row; otherwise, it will continue to output the pixel column of the current row. In order to ensure that the next_row signal received by the row buffer is at the beginning and end of the current row output, the next_row signal of each channel needs to be synchronized in the filter window generation module. That is, when the filter window generation module has generated the filter window for the first pixel of the current row, the next_row signal of each channel is logically ANDed before being sent to the row buffer. In some scenarios, different channels may need to crop the output image. To reduce power consumption, data outside the cropping area in the vertical direction is not processed. In this case, both the next_row signal and the chn_ready signal can be set to high level.

[0115] The filtered window data is sent to the processing modules of all channels simultaneously. Each channel's processing module compares the original image coordinates generated by the coordinate transformation module with the coordinates input from the filtered window. Only when the two coordinates are equal, the corresponding window filtering process is performed.

[0116] It should be noted that the entire image magnification process described above revolves around three core modules: the line buffer module, the filter window generation module, and the multi-channel processing module. It combines rules such as vertical / horizontal boundary mirror processing, channel control signal synchronization, and cropping power consumption optimization to adapt to the magnification differences of different channels.

[0117] In some embodiments, the specific steps for an electronic device to amplify and filter the original image based on a multi-channel amplification module structure may include four stages:

[0118] Phase 1: Raw Image Input and Line Buffer Initialization

[0119] Input rules: The original image is input into the system in the order of "row → pixel", and the data of the luminance channel and chrominance channel are processed independently.

[0120] Line cache configuration:

[0121] Luminance line buffer: Base number of lines = number of vertical lines of the amplification algorithm filter window (6 lines), after adding 1 line for performance optimization, the total number of lines is fixed at 7 lines;

[0122] Chroma line buffer: Base number of lines = number of vertical lines of the filter window (4 lines), add 1 line for performance optimization and the total number of lines is fixed at 5 lines;

[0123] The pixel size per row is fixed at the "horizontal resolution of the original image before magnification" and is independent of the number of channels and the resolution of the magnified image.

[0124] Data writing: The input raw image pixels are written to the corresponding row buffers in sequence (luminance → luminance buffer, chrominance → chrominance buffer) until the row buffers are full of initial data.

[0125] Phase 2: Output of row buffered pixel columns (vertical processing)

[0126] The output unit of the row buffer is "the column of pixels in the filter window corresponding to each row of pixels". The core processing involves the vertical boundary and the difference in channel magnification.

[0127] Vertical boundary mirroring: For the top / bottom boundary regions of the image, when the number of pixels required by the filtering window exceeds the vertical range of the original image, the row buffer fills the excess portion with mirrors to ensure that the output pixel column can support the complete filtering window.

[0128] Vertical output order: Output the corresponding pixel columns of each row in the order of "top → bottom".

[0129] Channel next_row signal control (adapting to differences in vertical magnification): Different channels have different vertical magnifications, and the number of times the same row of pixels is used varies. The next_row signal is used to control whether the row buffer switches to the next row.

[0130] Signal generation: Each channel needs to provide a next_row signal (high level = this channel needs the next row of data) after the start of the current row of pixels and before the end of the row buffer output.

[0131] Signal synchronization: When the filter window generation module generates the filter window for the first pixel of the current row, it performs a logical AND operation on the next_row signal of all channels (all 1s = all channels need the next row), and sends it to the row buffer after synchronization;

[0132] Row switching rules: After the row buffer finishes outputting the rightmost pixel column of the current row, check the synchronized next_row: if all are 1, output the pixel column of the next row; if not all are 1, continue outputting the pixel column of the current row (reusing the current row to adapt to the needs of low magnification channels).

[0133] Vertical cropping power consumption optimization: If a channel needs to perform vertical cropping on the output image, the data outside the cropping area does not need to be processed. In this case, the next_row signal of that channel can be directly set to high level to reduce invalid processing and lower power consumption.

[0134] Phase 3: Filtering Window Generation (Horizontal Processing + Coordinate Generation)

[0135] The filter window generation module receives the pixel columns output from the row buffer, combines them into a complete filter window, and simultaneously handles horizontal boundaries and differences in channel horizontal magnification.

[0136] Horizontal boundary mirroring: For the left / right boundary regions of the image, when the pixels required by the combined filtering window exceed the horizontal range of the original image, the excess part is mirrored to ensure the horizontal pixel integrity of the filtering window.

[0137] Filter window combination: The pixel column data input from the row buffer is combined into a complete filter window according to the filter window size required by the magnification algorithm (e.g., 6 vertical rows for luminance and 4 vertical rows for chrominance). The window data is simultaneously distributed to the processing modules of all channels.

[0138] Original coordinate generation: Synchronously generate "the coordinates of the pixel corresponding to the current filtering window in the original image" (denoted as window coordinates), which will serve as the basis for comparison in subsequent channel processing.

[0139] Channel chn_ready signal control (adapting to differences in horizontal magnification): Different channels have different horizontal magnifications, and the number of times the same pixel is used horizontally varies. The chn_ready signal is used to control whether the window shifts to the right.

[0140] Signal meaning: Each channel outputs the chn_ready signal (high level = this channel is ready and can receive the next filter window);

[0141] Window right-shifting rule: The filter window only moves to the right (outputs the filter window corresponding to the next column) when all channels' chn_ready is 1; otherwise, the window pauses and waits for all channels to be ready.

[0142] Horizontal cropping power consumption optimization: If a channel needs to perform horizontal cropping on the output image, the data outside the cropping area does not need to be processed. In this case, the chn_ready signal of that channel can be directly set to high level to reduce unnecessary power consumption.

[0143] Phase 4: Multi-channel filtering processing (coordinate comparison + window filtering)

[0144] Each channel processes the filtered window data independently, and amplification filtering is only performed when coordinates match:

[0145] The coordinate transformation module outputs: "the original image coordinates corresponding to the current output pixel to be processed" (denoted as the transformed coordinates).

[0146] Coordinate comparison: The processing module of each channel compares the window coordinates (output of the filter window generation module) and the transformed coordinates (output of the coordinate transformation module).

[0147] Filtering execution condition: The channel performs the corresponding amplification filtering process on the current filtering window only when the two coordinates are completely equal (ensuring that each output pixel only matches the filtering window of the corresponding coordinate in the original image).

[0148] In this embodiment, by employing fixed caching rules (the number of rows and the size of each row in the luminance / chrominance line buffer are fixed, and are only related to the resolution before magnification and the number of vertical rows in the filtering window), superimposed boundary mirroring to ensure the integrity of the filtering window (the vertical and horizontal boundaries are filled with mirrors respectively), channel control synchronization (the vertical next_row signal is synchronized in the filtering window module to control the switching of the line buffer rows, and the filtering window only moves to the right when the horizontal chn_ready signal is all 1), cropping optimization (the next_row / chn_ready outside the cropping area is set to a high level) to reduce power consumption, and coordinate matching triggering filtering (the channel performs filtering processing only when the window coordinates match the transformation coordinates), the entire image magnification process adapts to the magnification differences of different channels through caching, boundary processing, channel signal synchronization, and coordinate verification, while taking into account good performance and power consumption.

[0149] Please see Figure 10 This application also provides an image data magnification processing apparatus, which can be applied to electronic devices to implement the above-described image data magnification processing method.

[0150] like Figure 10 As shown, the image data magnification processing device provided in this embodiment includes a line buffer module 1001, a filter window generation module 1002, and a multi-channel processing module 1003. Among them,

[0151] The line caching module 1001 is used to acquire the input image and perform line caching on the input image;

[0152] The filter window generation module 1002 is used to combine the pixel columns of the filter window output from the row buffer of the input image into complete filter window data, and generate the first original image coordinates of the pixel corresponding to the complete filter window data in the input image; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image;

[0153] The multi-channel processing module 1003 is used to perform multi-channel independent amplification and filtering processing on the complete filter window data based on the first original image coordinates.

[0154] In some embodiments, the multi-channel processing module 1003 is further configured to compare the first original image coordinates with the second original image coordinates based on the target channel; the target channel is any one of a plurality of image processing channels, and the second original image coordinates are the original image coordinates corresponding to the pixel to be processed in the target channel; when the first original image coordinates and the second original image coordinates are equal, the complete filter window data is independently amplified and filtered based on the target channel.

[0155] In some embodiments, the row cache module 1001 is further configured to perform mirror filling processing on the vertical boundary of the input image based on the row cache of the input image to obtain a complete pixel column of the filter window; output the complete pixel column based on the row cache of the input image in the current output row order; the pixel column of the filter window output by the row cache of the input image includes the complete pixel column.

[0156] In some embodiments, the row buffer module 1001 is further configured to acquire the next row data request signal of each of the multiple image processing channels and synchronize the next row data request signal; the multiple image processing channels determine whether they need the next row data based on their own processing needs to generate the next row data request signal; when all synchronized next row data request signals indicate that the next row data is needed, the row buffer of the input image is switched to the next output row according to the current output row order to output the pixel column of the filter window; when none of the synchronized next row data request signals indicate that the next row data is needed, the row buffer of the input image continues to output the pixel column corresponding to the current output row according to the current output row order.

[0157] In some embodiments, the filter window generation module 1002 is further configured to perform horizontal padding on the pixel columns of the filter window output from the row buffer of the input image when the filter window exceeds the horizontal boundary of the input image; and to stitch the horizontally padded pixel columns together according to the size parameters of the filter window to obtain complete filter window data.

[0158] In some embodiments, the filter window generation module 1002 is further configured to acquire the back pressure signals of each of the multiple image processing channels; when the back pressure signals all indicate that the channels are in a ready state, the pixel column of the filter window after right shift is output based on the row buffer of the input image, and the step of combining the pixel column of the filter window output from the row buffer of the input image into complete filter window data is returned to continue execution.

[0159] In some embodiments, the row cache module 1001 is further configured to write the pixels of the input image into the corresponding row cache in sequence, thereby obtaining the row cache of the input image; the base number of the row cache of the input image is equal to the vertical number of the filter window plus 1, and the size of each row of pixels in the row cache of the input image corresponds to the resolution of the input image.

[0160] It should be noted that the specific implementation of the image data magnification processing device provided in this application embodiment is basically the same as the specific implementation of the image data magnification processing method described above, and will not be repeated here.

[0161] Please see Figure 11 This application also provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the above-described method for evaluating the visual language model.

[0162] like Figure 11 As shown, in some embodiments, the electronic device provided in this application includes:

[0163] The processor 1101 can be implemented using a general-purpose central processing unit (CPU), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this application.

[0164] The memory 1102 can be implemented as a read-only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1102 can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory 1102 and is called and executed by the processor 1101 to execute the image data magnification processing method of the embodiments of this application.

[0165] Input / output interface 1103 is used to implement information input and output;

[0166] The communication interface 1104 is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0167] Bus 1105 transmits information between various components of the device (e.g., processor 1101, memory 1102, input / output interface 1103, and communication interface 1104);

[0168] The processor 1101, memory 1102, input / output interface 1103 and communication interface 1104 are connected to each other within the device via bus 1105.

[0169] This application also provides an electronic device, which includes a processor and a memory. The memory stores a computer program, and the processor executes the computer program to implement the above-described image data magnification processing method.

[0170] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-described image data magnification processing method.

[0171] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0172] This application also provides a computer program product, including a computer program. The steps implemented by the computer program when executed by a processor are basically the same as those in the specific embodiments of the image data magnification processing method described above, and will not be repeated here.

[0173] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0174] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0175] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0176] Those skilled in the art will understand that all or some of the steps, apparatuses, or functional modules / units in the methods disclosed above can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0177] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0178] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0179] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of the units described above is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0180] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0181] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0182] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes multiple instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0183] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

Claims

1. A method for amplifying image data, characterized in that, The method is applied to an electronic device, and the method includes: Acquire the input image and perform row caching on the input image; The pixel columns of the filter window output from the row buffer of the input image are combined into complete filter window data, and the first original image coordinates of the pixel corresponding to the complete filter window data in the input image are generated; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image; Based on the first original image coordinates, the complete filter window data is subjected to multi-channel independent amplification and filtering processing; wherein, the complete filter window data is simultaneously sent to the processing modules of multiple image processing channels, and each of the multiple image processing channels generates the second original image coordinates corresponding to the current pixel to be processed in this channel through the coordinate transformation module. After receiving the complete filter window data, any target channel among the multiple image processing channels first compares the second original image coordinates with the first original image coordinates. Only when the two coordinates are completely equal does the target channel perform independent amplification and filtering processing on the complete filter window data.

2. The method according to claim 1, characterized in that, The method further includes: The vertical boundary of the input image is mirror-filled based on the row buffer of the input image to obtain the complete pixel column of the filter window; The row buffer of the input image outputs the complete pixel column in the current output row order; the pixel column of the filter window output by the row buffer of the input image includes the complete pixel column.

3. The method according to claim 1, characterized in that, The method further includes: The system acquires the next row data request signal from each of the multiple image processing channels and synchronizes the next row data request signals. The multiple image processing channels determine whether they need the next row of data based on their own processing requirements and generate the next row data request signal accordingly. If all subsequent data request signals after synchronization indicate that the next row of data is needed, the row buffer of the input image is switched to the next row to be output according to the current output row order to output the pixel column of the filter window; If the next row data request signal after synchronization does not indicate that the next row of data is needed, the row buffer of the input image continues to output the pixel column corresponding to the current output row according to the current output row order.

4. The method according to claim 1, characterized in that, The step of combining the pixel columns of the filter window output from the row buffer of the input image into complete filter window data includes: If the filtering window exceeds the horizontal boundary of the input image, the pixel columns of the filtering window output from the row buffer of the input image are horizontally padded. According to the size parameters of the filter window, the horizontally padded pixel columns are stitched together to obtain the complete filter window data.

5. The method according to claim 1, characterized in that, After performing multi-channel independent filtering on the complete filter window data based on the first original image coordinates, the method further includes: Acquire the reverse voltage signals of each of the multiple image processing channels; When all the back pressure signals indicate that the channel is in a ready state, the pixel column of the filter window after the row buffer of the input image is shifted to the right is output, and the step of combining the pixel column of the filter window output from the row buffer of the input image into complete filter window data continues to be executed.

6. The method according to any one of claims 1 to 5, characterized in that, The process of caching the input image by rows includes: The pixels of the input image are written to the corresponding row buffers in order, thus obtaining the row buffer of the input image; the base number of the row buffer of the input image is equal to the vertical number of the filter window plus 1, and the size of each row of pixels in the row buffer of the input image corresponds to the resolution of the input image.

7. An image data magnification processing device, characterized in that, Applied to electronic devices, the device includes: A line caching module is used to acquire the input image and cache the input image in lines; A filter window generation module is used to combine the pixel columns of the filter window output from the row buffer of the input image into complete filter window data, and generate the first original image coordinates of the pixels corresponding to the complete filter window data in the input image; the filter window is the filter window of the magnification algorithm, and the pixel columns of the filter window correspond to each row of pixels in the row buffer of the input image; A multi-channel processing module is used to perform multi-channel independent amplification and filtering processing on the complete filtering window data based on the first original image coordinates; wherein, the complete filtering window data is simultaneously sent to the processing modules of multiple image processing channels, and the processing module of any target channel among the multiple image processing channels compares the first original image coordinates with the second original image coordinates corresponding to the pixel to be processed in this channel generated by the coordinate transformation module of this channel, and performs independent amplification and filtering processing on the complete filtering window data when the first original image coordinates and the second original image coordinates are equal.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing a computer program, and the processor executing the computer program implements the image data magnification processing method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the image data magnification processing method as described in any one of claims 1 to 6.

Images