CN113034396B

CN113034396B - Image filtering method, device, electronic equipment and storage medium

Info

Publication number: CN113034396B
Application number: CN202110331415.7A
Authority: CN
Inventors: 陈昊; 殷时友
Original assignee: Longxin Zhongke Hefei Technology Co ltd
Current assignee: Longxin Zhongke Hefei Technology Co ltd
Priority date: 2021-03-26
Filing date: 2021-03-26
Publication date: 2023-05-12
Anticipated expiration: 2041-03-26
Also published as: CN113034396A

Abstract

The embodiment of the invention provides an image filtering method, an image filtering device, electronic equipment and a storage medium, and relates to the technical field of computers. Wherein the method comprises the following steps: determining the initial window size of the filter and the number of pixels stored in a register corresponding to the source image; determining a data alignment mode of filtering processing according to the initial window size and the pixel number; according to the data alignment mode, the filter is adjusted; and filtering the source image through the SIMD instruction by adopting the adjusted filter. In the invention, the number of pixels stored in one register corresponding to the source image, namely the number of pixels which can be processed at one time in SIMD operation, the data alignment mode of filtering processing can be set according to the initial window size of the filter and the number of pixels which can be processed at one time in SIMD operation, so that the filter can be adapted to the size of the source image, and meanwhile, the low-efficiency SIMD operation is reduced, the filtering operation efficiency is improved, and the power consumption is saved.

Description

Image filtering method, device, electronic equipment and storage medium

Technical Field

The present invention relates to the field of computer technology, and in particular, to an image filtering method, an image filtering apparatus, an electronic device, and a storage medium.

Background

In the algorithm implementation of image filtering, the window size of the filter can determine the filtering performance, a filter with a proper window size can greatly improve the filtering performance, and the filtering algorithm can be optimized through a SIMD (Single Instruction Multiple Data, single instruction multiple data stream) instruction. SIMD instructions may process a set of pixel data in a register within one instruction cycle. For example, for a 128-bit register, 4 32-bit, 8-bit, 16-bit, or 16-bit 8-bit pixel data may be processed simultaneously.

The pixel components of the common image are all composed of 8-bit unsigned numbers, so that in order to prevent the situation that a computer has out of range in the process of processing weighting, the values of pixels of the image are expanded to 16 bits when the SIMD instruction is used for optimization, and therefore, the number of pixels which can be processed at one time by the SIMD instruction through registers with different bits is different, for example, a 128-bit register can be used for processing 8 pixels at one time by the SIMD instruction. For a certain register, n pixels can be processed at one time through a SIMD instruction, if the size of a filter window is not a multiple of n, the size of the filter window is required to be expanded to be a multiple of n in a 0 supplementing mode, namely n is aligned, so that the advantage of SIMD instruction vector optimization can be fully exerted, and the filtering operation efficiency is improved.

However, with the above-described method, even if the filter window size exceeds n by a factor of x, the filter window size is extended to n by a factor of (x+1), for example, the filter window size is extended from n+1 to 2n, and thus, an inefficient SIMD operation is performed more than once, resulting in a reduction in the filtering operation efficiency and a waste of power consumption.

Disclosure of Invention

In view of the foregoing, embodiments of the present invention have been made to provide an image filtering method that overcomes or at least partially solves the foregoing problems, so as to solve the problems of low filtering operation efficiency and waste of power consumption of an image.

Correspondingly, the embodiment of the invention also provides an image filtering device, electronic equipment and a storage medium, which are used for ensuring the realization and application of the method.

In order to solve the above problems, an embodiment of the present invention discloses an image filtering method, including:

determining the initial window size of the filter and the number of pixels stored in a pixel data register corresponding to the source image;

determining a data alignment mode of filtering processing according to the initial window size and the pixel number;

according to the data alignment mode, the filter is adjusted;

And filtering the source image through a single-instruction multi-data-stream instruction by adopting the adjusted filter.

Optionally, the determining the data alignment mode of the filtering process according to the initial window size and the pixel number includes:

setting the data alignment mode of the filtering process as n alignment, and executing the following steps in sequence;

modifying the data alignment mode into m alignment under the condition that the initial window size meets a first preset condition;

modifying the data alignment mode to be 1 alignment under the condition that the initial window size meets a second preset condition;

the first preset condition and the second preset condition are related to the initial window size and the pixel number, n is the pixel number, and m is a multiple of 4.

setting the data alignment mode of the filtering process to be 1 alignment under the condition that the initial window size meets a second preset condition;

setting the data alignment mode as m alignment when the initial window size does not meet the second preset condition and meets the first preset condition;

Setting the data alignment mode as n alignment when the initial window size does not meet the second preset condition and does not meet the first preset condition;

Optionally, the first preset condition includes that the initial window size does not exceed a first preset value, and the first preset value does not exceed half of the number of pixels.

Optionally, the second preset condition includes that a remainder of dividing the initial window size by the number of pixels is smaller than a second preset value, and the second preset value is not more than half of the number of pixels.

The embodiment of the invention also discloses an image filtering device, which comprises:

a first determining module, configured to determine an initial window size of the filter and a number of pixels stored in a pixel data register corresponding to the source image;

the second determining module is used for determining a data alignment mode of filtering processing according to the initial window size and the pixel number;

the adjusting module is used for adjusting the filter according to the data alignment mode;

And the processing module is used for filtering the source image through a single-instruction multi-data-stream instruction by adopting the adjusted filter.

Optionally, the second determining module includes:

the first setting submodule is used for setting the data alignment mode of the filtering process to be n alignment and calling the first modification module and the second modification module in sequence;

the first modifying module is configured to modify the data alignment mode into m alignment when the initial window size meets a first preset condition;

the second modifying module is configured to modify the data alignment mode to 1 alignment when the initial window size meets a second preset condition;

Optionally, the second determining module includes:

the second setting module is used for setting the data alignment mode of the filtering processing to be 1 alignment under the condition that the initial window size meets a second preset condition;

a third setting module, configured to set the data alignment mode to be m alignment when the initial window size does not meet the second preset condition and meets the first preset condition;

A fourth setting module, configured to set the data alignment mode to n alignment when the initial window size does not meet the second preset condition and does not meet the first preset condition;

The embodiment of the invention also discloses an electronic device which comprises a memory and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by one or more processors, and the one or more programs comprise instructions for:

according to the data alignment mode, the filter is adjusted;

The embodiment of the invention also discloses a readable storage medium, which enables the electronic equipment to execute one or more of the image filtering methods in the embodiment of the invention when the instructions in the storage medium are executed by the processor of the electronic equipment.

The embodiment of the invention has the following advantages:

in the embodiment of the invention, the initial window size of the filter and the pixel number stored in one pixel data register corresponding to the source image, namely the pixel number which can be processed at one time by SIMD operation, can be determined firstly, then for different initial window sizes and pixel numbers, different alignment modes of filtering processing data can be determined, and the filter can be adjusted according to different alignment modes to obtain filters with different sizes, and then the filter with proper size after adjustment is adopted, and the filtering processing is carried out on the source image through SIMD instructions. In the embodiment of the invention, the data alignment mode of the filtering process can be set according to the initial window size of the filter and the number of the pixels which can be processed at one time in the SIMD operation, so that the filter can be adapted to the size of the source image, and meanwhile, the low-efficiency SIMD operation is reduced, the filtering operation efficiency is improved, and the power consumption is saved.

Drawings

FIG. 1 is a flow chart of steps of an embodiment of an image filtering method of the present invention;

FIG. 2 is a flow chart of steps of another embodiment of an image filtering method of the present invention;

FIG. 3 is a flow chart of steps of yet another embodiment of an image filtering method of the present invention;

fig. 4 is a block diagram showing the structure of an embodiment of an image filtering apparatus of the present invention;

fig. 5 is a block diagram illustrating a configuration of an electronic device for image filtering, according to an exemplary embodiment.

Detailed Description

In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.

Before describing embodiments of the present invention in detail, some concepts related to the present invention will be described first.

Filter window size: the filter window, which may be simply referred to as a filter, refers to the number of pixels contained in the filter window. For example, for a 5×5 filter window, the size of the filter window is 25, and for a 2×4 filter window, the size of the filter window is 8.

Alignment of x: the data alignment of the filtering process, x alignment meaning the expansion of the filter window size to a multiple of x, may be set before filtering the image. In the current implementation, the data alignment of the filtering process is generally set to n-alignment, where n refers to the number of pixels stored in a pixel data register corresponding to the source image, that is, the number of pixels that can be processed at one time by the SIMD instruction.

For example, when a 128-bit register is used for storing pixel data in a source image, and when the pixel data of a pixel is expanded from 8 bits to 16 bits through optimization of a SIMD instruction, a 128-bit register can store pixel values of 8 pixels, that is, the number of pixels stored in a pixel data register corresponding to the source image is 8. Therefore, the data alignment manner of the filtering process is set to 8 alignment, and further, in the case where the filter window size is not a multiple of 8, the filter window size can be expanded to a multiple of 8. For example, when the filter window size is 5, the filter window size may be extended to 8, and when the filter window size is 12, the filter window size may be extended to 16, and the extended portion may be supplemented with 0.

One of the core concepts of the embodiments of the present invention is that before filtering a source image by a SIMD instruction, different data alignment modes may be set according to an initial window size of a filter and the number of pixels stored in a pixel data register corresponding to the source image, where the number of pixels stored in a register corresponding to the source image is the number of pixels that can be processed at one time by SIMD operation, and the data alignment mode of the filtering process may be set according to the initial window size of the filter, so that the filter is ensured to be adapted to the source image size, and the data alignment mode may also be set according to the number of SIMD pixels that can be processed at one time by SIMD operation, thereby reducing low-efficiency SIMD operation, improving filtering operation efficiency, and saving power consumption.

Referring to fig. 1, a flowchart illustrating steps of an embodiment of an image filtering method according to the present invention may specifically include the following steps:

step 101: the initial window size of the filter is determined, as well as the number of pixels stored in a pixel data register corresponding to the source image.

In the embodiment of the invention, the electronic device can determine the initial window size of the filter according to the size of the source image and the size of the target image, and the related technology can be referred to specifically. The electronic device may further determine the number of pixels n stored in one pixel data register corresponding to the source image according to the number of bits of the pixel data register corresponding to the source image. For example, where the value of one pixel is 16 bits, then 8 pixels may be stored for a 128 bit register, i.e., 8 pixels may be processed at a time by a SIMD instruction. For a 256 bit register, 16 pixels may be stored, i.e., 16 pixels may be processed at a time by a SIMD instruction.

The number n of pixels stored in one pixel data register corresponding to the source image is determined based on the value of each pixel expanded to 16 bits by SIMD expansion.

Step 102: and determining the data alignment mode of the filtering process according to the initial window size and the pixel number.

In this step, different preset conditions can be combined according to the number of pixels, each preset condition can be divided into a satisfied condition and an unsatisfied condition, and when the satisfied conditions of the preset conditions are different, different data alignment modes can be set.

In practical applications, when the remainder of dividing the initial window size L by the number of pixels n is small, the data alignment may be determined as 1 alignment.

When the initial window size L is small compared to the number of pixels n, the data alignment may be determined as an alignment of a small value, which may not exceed half of n.

When the initial window size L is larger than the number of pixels n and the remainder of dividing the initial window size L by the number of pixels n is larger, the data alignment manner may be determined as n-aligned.

Step 103: and adjusting the filter according to the data alignment mode.

In this step, when the data alignment mode is x alignment, the filter may be extended to a multiple of x, and the extended bit number may be supplemented with 0.

Specifically, in the case where the data alignment mode is 1 alignment, since the initial window size is necessarily a multiple of 1, when the data alignment mode is 1 alignment, the initial window size may not be adjusted, and a filter with the initial window size may be directly adopted.

For the case that the data alignment mode is smaller value alignment, the size of the filter window can be finely adjusted so that the size of the filter window is a multiple of the smaller value.

And (3) for the case that the data alignment mode is n alignment, performing size fine adjustment on the filter window so that the size of the filter window is a multiple of n.

Step 104: and filtering the source image through a single-instruction multi-data-stream instruction by adopting the adjusted filter.

In this step, an appropriately sized filter may be used to filter the source image via SIMD instructions, and then at least x pixels may be filtered at a time.

And for the case that the data alignment mode is 1 alignment, directly adopting a filter with an initial window size to carry out filtering processing. Specifically, L (L is the initial window size) pixels may be a group, SIMD optimization is first used, weighting of multiple groups of pixels is sequentially processed, and when the number of remaining pixels is less than L, SIMD optimization may not be used any more, and weighting processing of pixels may be directly performed. By filtering with such a filter, SIMD optimization can be employed for most pixels, exerting SIMD advantages.

For the case that the data alignment mode is smaller value alignment, the window size of the filter can be adjusted to be a multiple of the smaller value, the adjustment degree is smaller, the number of pixels which can be processed at one time of the SIMD instruction can be properly increased by filtering through the filter, and excessive 0 compensation is avoided, so that excessive invalid SIMD operation is caused.

For the case that the data alignment mode is n alignment, the window size of the filter can be adjusted to n, the adjustment degree in the case is smaller, and 0 supplement is smaller when the filter is used for filtering, so that the advantage of SIMD is exerted, and meanwhile, less invalid SIMD operation is realized.

Referring to fig. 2, a flowchart illustrating steps of another embodiment of an image filtering method according to the present invention may specifically include the following steps:

step 201: the initial window size of the filter is determined, as well as the number of pixels stored in a pixel data register corresponding to the source image.

In the embodiment of the present invention, reference may be made to step 101 for specific implementation of this step.

Step 202: setting the data alignment mode of the filtering process to n alignment, and sequentially executing the following

steps

203 and 204; n is the number of pixels stored in a pixel data register corresponding to the source image.

In this step, the data alignment method of the filtering process may be first defaulted to n-alignment, and then the following

steps

203 and 204 may be sequentially performed.

For example, if the source image stores pixel data in a 128-bit register, and n is 8 for the 128-bit register, the data alignment of the filtering process may be set to 8 alignment first.

Step 203: under the condition that the initial window size meets a first preset condition, modifying the data alignment mode into m alignment; the first preset condition is related to the initial window size and the number of pixels, m being a multiple of 4.

The first preset condition includes that the initial window size does not exceed a first preset value, and the first preset value does not exceed half of the number of pixels. The initial window size is not more than half of the number of pixels, namely, the initial window size is smaller than the number of pixels, and the data alignment mode can be modified into m alignment with smaller values, so that the size of the filter can be finely adjusted subsequently, and the excessive 0 compensation is avoided under the condition that the number of pixels which can be processed at one time by a SIMD instruction is properly increased, and the excessive invalid SIMD operation is avoided.

In an alternative implementation, for a 128-bit register, the SIMD instruction is one-time processable pixel count bit 8, then the first preset value may be set to 4, and the corresponding first preset condition includes the initial window size not exceeding 4, i.e. the initial window size is greater than or equal to 4.

For example, when the initial window size is 3, 3 is not more than 4, and the first preset condition is satisfied, the data alignment mode can be modified from 8 alignment to 4 alignment, so that for a 128-bit register, 2 filter weighting processes can be performed at a time subsequently, and for a 256-bit register, 4 filter weighting processes can be performed at a time subsequently, so that the register can perform better performance. When the initial window size is 10, the initial window size exceeds 4, the first preset condition is not satisfied, the data alignment mode can not be modified, and the data alignment mode is still 8. When the initial window size is 15, 15 exceeds 4, the first preset condition is not satisfied, the data alignment mode can not be modified, and the data alignment is still 8.

In the embodiment of the present invention, in the case that the initial window size does not meet the first preset condition, the following step 204 is directly performed, that is, whether the initial window size meets the first preset condition or not, the following step 204 is continuously performed.

Step 204: under the condition that the initial window size meets a second preset condition, modifying the data alignment mode into 1 alignment; the second preset condition is related to the initial window size and the number of pixels.

The remainder of dividing the initial window size by the number of pixels is smaller than a second preset value, and the second preset value is not more than half of the number of pixels. The remainder obtained by dividing the initial window size by the number of pixels is smaller than half of the number of pixels, namely, the remainder obtained by dividing the initial window size by the number of pixels is smaller, and then the data alignment mode can be modified into 1 alignment, namely, the size of the filter is not adjusted, and the filter with the initial window size is directly adopted. In this way, during the subsequent filtering processing, L (L is the initial window size) pixels can be a group, SIMD optimization is firstly used, weighting of a plurality of groups of pixels is sequentially processed, when the number of the remaining pixels is less than L, SIMD optimization can be not used any more, and weighting processing of one pixel by one pixel is directly performed, so that SIMD optimization can be adopted for most pixels, and the advantage of SIMD is exerted.

In step 203, if the initial window size satisfies the first preset condition, the alignment mode is modified from the default n alignment to the m alignment. However, if the initial window size does not meet the first preset condition, the alignment mode is still the default n alignment, and no modification is performed. In step 204, regardless of whether the alignment is modified in step 203, the alignment is modified to be 1 alignment as long as the initial window size satisfies the second preset condition.

Specifically, if the alignment is not modified in step 203 and still is n-aligned, in step 204, if the initial window size satisfies the second preset condition, the alignment is modified from n-aligned to 1-aligned, and if the initial window size does not satisfy the second preset condition, the alignment is not modified and still is n-aligned.

If the alignment is modified from n alignment to m alignment in step 203, in step 204, if the initial window size satisfies the second preset condition, the alignment is modified from m alignment to 1 alignment, and if the initial window size does not satisfy the second preset condition, the alignment is not modified and still is m alignment.

In an alternative implementation, for a 128-bit register, SIMD instruction one-time-processible pixel count bit 8, then the second preset value may be set to 3, and correspondingly, the second preset condition includes that the remainder of the initial window size to the pixel count is less than 3.

For example, when the initial window size is 3, the remainder of 3 pairs of pixels 8 is 3, and the second preset condition is not satisfied, the data alignment mode may not be modified, and the data alignment is still 4. When the initial window size is 10, the remainder of 10 pairs of pixels 8 is smaller than 3, the second preset condition is met, and the data alignment mode can be modified from 8 alignment to 1 alignment. When the initial window size is 15, the remainder of 15 to 8 pixels is 7, and the second preset condition is not satisfied, the data alignment mode can be not modified, and the data alignment mode is still 8.

In the embodiment of the invention, the important second preset condition can be put at the end for judgment, if the second preset condition is met, even if the data alignment mode is modified before, the data alignment mode can be modified to be more proper at the end, and if the second preset condition is not met, the data alignment mode before can be described to be more proper.

Step 205: and adjusting the filter according to the data alignment mode.

In this step, the window size of the filter may be extended by a multiple of the data alignment, the extended number of bits being complemented by 0. For example, when the initial window size is 3, the data alignment is 4, the window size of the filter can be extended from 3 to 4 (1 times 4), and the last bit of the filter is complemented with 0. When the initial window size is 10, the data alignment is 1 alignment, and the window size of the filter may not be extended. When the initial window size is 15, the data alignment is 8, the window size of the filter can be expanded from 15 to 16 (2 times of 8), and the last bit of the filter is complemented with 0.

Step 206: and filtering the source image through a single-instruction multi-data-stream instruction by adopting the adjusted filter.

In this step, the pixel data in the pixel data register may be weighted with the adjusted filter as a weight.

When the window size of the adjusted filter is 4, the data alignment is 4 alignment, and for 128-bit registers, weighting processing can be performed pixel by pixel. Specifically, the filter vector may be dot multiplied with the pixels, and the dot multiplication results of 4 pixels are added to obtain one piece of processed pixel data.

When the window size of the adjusted filter is 10, the data alignment is 1 alignment, and for the 128-bit register, the weighting process may be performed on 8 pixels first, and then the remaining 2 pixels may be weighted pixel by pixel. Specifically, the filter vector may be dot multiplied by the pixel, the dot multiplied results of 8 pixels are added, the dot multiplied result of 9 th pixel is added, and the dot multiplied result of 10 th pixel is added to obtain a processed pixel data.

When the window size of the adjusted filter is 16, the data alignment mode is 8 alignment, and for 128-bit registers, weighting processing can be performed on every 8 pixels in turn. Specifically, the filter vector may be dot multiplied by the pixels, the dot multiplied results of the 8 pixels of the first group are added, then the filter vector is dot multiplied by the pixels, the dot multiplied results of the 8 pixels of the second group are added, and then the dot multiplied results of the two groups of pixels are added to obtain one processed pixel data.

When the window size of the adjusted filter is 16, the data alignment mode is 8 alignment, and for 256-bit registers, weighting processing can be directly performed on 16 pixels. Specifically, the filter vector may be dot multiplied with the pixels, and the dot multiplication results of 16 pixels are added to obtain one piece of processed pixel data.

When the weighting operation of the SIMD instruction optimization filter is used, the size of the filter window is reasonably designed, so that the alignment mode is more flexible and more reasonable, better performance is achieved, unnecessary register operation is reduced, the optimization performance of the SIMD can be better exerted, and the energy consumption caused by the SIMD is reasonably reduced.

It should be noted that, the embodiment of the present invention only takes a specific filter window size and the number of pixels as an example to assist in describing the technical details of the embodiment of the present invention, and the present invention is not limited thereto.

In the embodiment of the invention, the initial window size of the filter and the pixel number stored in one pixel data register corresponding to the source image, namely the pixel number which can be processed at one time by SIMD operation, are firstly determined, then different alignment modes of filtering processing data can be determined when the first preset condition and the second preset condition are different, wherein the first preset condition and the second preset condition are related to the initial window size and the pixel number, then the filter is adjusted according to different alignment modes to obtain filters with different sizes, and further the filter with proper adjusted size is adopted to carry out filtering processing on the source image through SIMD instructions. In the embodiment of the invention, the data alignment mode of the filtering process can be set according to the initial window size of the filter and the meeting condition of the first preset condition and the second preset condition related to the number of the pixels which can be processed at one time by SIMD operation, so that the filter can be adapted to the size of a source image, and meanwhile, the low-efficiency SIMD operation is reduced, the filtering operation efficiency is improved, and the power consumption is saved.

Referring to fig. 3, a flowchart illustrating steps of yet another embodiment of an image filtering method according to the present invention may specifically include the following steps:

step 301: the initial window size of the filter is determined, as well as the number of pixels stored in a pixel data register corresponding to the source image.

In the embodiment of the present invention, the specific implementation manner of determining the initial window size of the filter and the number of pixels stored in one pixel data register corresponding to the source image may refer to step 101.

Step 302: setting the data alignment mode of the filtering process to be 1 alignment under the condition that the initial window size meets a second preset condition; the second preset condition is related to the initial window size and the number of pixels.

In the embodiment of the invention, whether the initial window size meets the second preset condition can be judged first. The remainder of dividing the initial window size by the number of pixels is smaller than a second preset value, and the second preset value is not more than half of the number of pixels. The remainder obtained by dividing the initial window size by the number of pixels is smaller than half of the number of pixels, namely, the remainder obtained by dividing the initial window size by the number of pixels is smaller, and then the data alignment mode can be directly set to be 1 alignment, namely, the size of the filter is not adjusted, and the filter with the initial window size is directly adopted. In this way, during the subsequent filtering processing, L (L is the initial window size) pixels can be a group, SIMD optimization is firstly used, weighting of a plurality of groups of pixels is sequentially processed, when the number of the remaining pixels is less than L, SIMD optimization can be not used any more, and weighting processing of one pixel by one pixel is directly performed, so that SIMD optimization can be adopted for most pixels, and the advantage of SIMD is exerted.

For example, when the initial window size is 10, the remainder of 10 to the number of pixels 8 is less than 3, the second preset condition is satisfied, and the data alignment manner may be directly set to 1 alignment.

Step 303: setting the data alignment mode as m alignment when the initial window size does not meet the second preset condition and meets the first preset condition; the first preset condition is related to the initial window size and the number of pixels, m being a multiple of 4.

In the embodiment of the invention, under the condition that the initial window size does not meet the second preset condition, whether the initial window size meets the first preset condition can be continuously judged.

For example, when the initial window size is 3, the remainder of 3 pairs of the number 8 of pixels is equal to 3, the second preset condition is not satisfied, but 3 does not exceed 4, the first preset condition is satisfied, and the data alignment manner may be set to 4 alignment.

Step 304: setting the data alignment mode as n alignment when the initial window size does not meet the second preset condition and does not meet the first preset condition; n is the number of pixels stored in a pixel data register corresponding to the source image.

For example, when the initial window size is 15, the remainder of 15 to the number of pixels 8 is 7, the second preset condition is not satisfied, and 15 exceeds 4, the first preset condition is not satisfied, and the data alignment manner may be set to 8 alignment.

In the embodiment of the invention, the important second preset condition can be judged first, then the first preset condition is judged, if the second preset condition is met, the data alignment mode can be directly set to be the proper 1 alignment, if the second preset condition is not met, but the first preset condition is met, the data alignment mode can be directly set to be the proper m alignment, and if the second preset condition is not met, and the first preset condition is not met, the data alignment mode can be directly set to be the proper n alignment. Therefore, multiple modifications to the data alignment mode can be avoided, and the overall efficiency of filtering processing is improved.

Step 305: and adjusting the filter according to the data alignment mode.

For a specific implementation of this step, reference is made to step 205 described above.

Step 306: and filtering the source image through a single-instruction multi-data-stream instruction by adopting the adjusted filter.

For a specific implementation of this step, reference is made to step 206 above.

It should be noted that, for simplicity of description, the method embodiments are shown as a series of acts, but it should be understood by those skilled in the art that the embodiments are not limited by the order of acts, as some steps may occur in other orders or concurrently in accordance with the embodiments. Further, those skilled in the art will appreciate that the embodiments described in the specification are presently preferred embodiments, and that the acts are not necessarily required by the embodiments of the invention.

Referring to fig. 4, which shows a block diagram of an embodiment of an image filtering apparatus according to the present invention, an image filtering apparatus 400 may specifically include the following modules:

a first determining module 401, configured to determine an initial window size of the filter and a number of pixels stored in a pixel data register corresponding to the source image;

a second determining module 402, configured to determine a data alignment mode of the filtering process according to the initial window size and the number of pixels;

an adjustment module 403, configured to adjust the filter according to the data alignment manner;

and the processing module 404 is configured to perform filtering processing on the source image through a single instruction multiple data stream instruction by using the adjusted filter.

Optionally, the second determining module includes:

In the embodiment of the invention, the initial window size of the filter and the pixel number stored in one pixel data register corresponding to the source image, namely the pixel number which can be processed at one time by SIMD operation, are firstly determined through a first determining module, then different alignment modes of filtering processing data can be determined for different initial window sizes and pixel numbers through a second determining module, the filter is adjusted according to different alignment modes through an adjusting module, different-size filters are obtained, and then the source image is filtered through a processing module by adopting the filter with proper adjusted size through a SIMD instruction. In the embodiment of the invention, the data alignment mode of the filtering process can be set according to the initial window size of the filter and the number of the pixels which can be processed at one time in the SIMD operation, so that the filter can be adapted to the size of the source image, and meanwhile, the low-efficiency SIMD operation is reduced, the filtering operation efficiency is improved, and the power consumption is saved.

For the device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference is made to the description of the method embodiments for relevant points.

Fig. 5 is a block diagram illustrating a configuration of an electronic device 500 for image filtering, according to an example embodiment. For example, the electronic device 500 may be a computer, a server, or the like.

Referring to fig. 5, an electronic device 500 may include one or more of the following components: a processing component 502, a memory 504, a power supply component 506, a multimedia component 508, an audio component 510, an input/output (I/O) interface 512, a sensor component 514, and a communication component 516.

The processing component 502 generally controls overall operation of the electronic device 500, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 502 may include one or more processors 520 to execute instructions to perform all or part of the steps of the methods described above. Further, the processing component 502 can include one or more modules that facilitate interactions between the processing component 502 and other components. For example, the processing component 502 may include a multimedia module to facilitate interaction between the multimedia component 508 and the processing component 502.

Memory 504 is configured to store various types of data to support operations at device 500. Examples of such data include instructions for any application or method operating on the electronic device 500, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 504 may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.

The power supply component 506 provides power to the various components of the electronic device 500. The power components 506 may include a power management system, one or more power sources, and other components associated with generating, managing, and distributing power for the electronic device 500.

The multimedia component 508 includes a screen between the electronic device 500 and the user that provides an output interface. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touches, swipes, and gestures on the touch panel. The touch sensor may sense not only the boundary of a touch or slide action, but also the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 508 includes a front-facing camera and/or a rear-facing camera. When the electronic device 500 is in an operational mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each front camera and rear camera may be a fixed optical lens system or have focal length and optical zoom capabilities.

The audio component 510 is configured to output and/or input audio signals. For example, the audio component 510 includes a Microphone (MIC) configured to receive external audio signals when the electronic device 500 is in an operational mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may be further stored in the memory 505 or transmitted via the communication component 516. In some embodiments, the audio component 510 further comprises a speaker for outputting audio signals.

The I/O interface 512 provides an interface between the processing component 502 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: homepage button, volume button, start button, and lock button.

The sensor assembly 514 includes one or more sensors for providing status assessment of various aspects of the electronic device 500. For example, the sensor assembly 514 may detect the on/off state of the device 500, the relative positioning of components, such as a display and keypad of the electronic device 500, the sensor assembly 514 may also detect a change in position of the electronic device 500 or a component of the electronic device 500, the presence or absence of a user's contact with the electronic device 500, the orientation or acceleration/deceleration of the electronic device 500, and a change in temperature of the electronic device 500. The sensor assembly 514 may include a proximity sensor configured to detect the presence of nearby objects without any physical contact. The sensor assembly 514 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 514 may also include an acceleration sensor, a gyroscopic sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 516 is configured to facilitate communication between the electronic device 500 and other devices, either wired or wireless. The electronic device 500 may access a wireless network based on a communication standard, such as WiFi,2G, or 3G, or a combination thereof. In one exemplary embodiment, the communication component 516 receives broadcast signals or broadcast-related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 516 further includes a Near Field Communication (NFC) module to facilitate short range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, ultra Wideband (UWB) technology, bluetooth (BT) technology, and other technologies.

In an exemplary embodiment, the electronic device 500 may be implemented by one or more Application Specific Integrated Circuits (ASICs), digital Signal Processors (DSPs), digital Signal Processing Devices (DSPDs), programmable Logic Devices (PLDs), field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic elements for executing the methods described above.

In an exemplary embodiment, a non-transitory computer readable storage medium is also provided, such as memory 504, including instructions executable by processor 520 of electronic device 500 to perform the above-described method. For example, the non-transitory computer readable storage medium may be ROM, random Access Memory (RAM), CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

A non-transitory computer readable storage medium, which when executed by a processor of a terminal, causes the terminal to perform a method of image filtering, the method comprising:

according to the data alignment mode, the filter is adjusted;

Optionally, the determining the data alignment mode of the filtering process according to the initial window size and the pixel number includes sequentially executing the following steps:

In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described by differences from other embodiments, and identical and similar parts between the embodiments are all enough to be referred to each other.

It will be apparent to those skilled in the art that embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the invention may take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, terminal devices (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the scope of the embodiments of the invention.

Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or terminal. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or terminal device comprising the element.

The foregoing has described in detail a method and apparatus for filtering images, an electronic device and a storage medium, and specific examples have been used herein to illustrate the principles and embodiments of the present invention, and the above examples are only for aiding in the understanding of the method and core ideas of the present invention; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present invention, the present description should not be construed as limiting the present invention in view of the above.

Claims

1. An image filtering method, comprising:

determining a data alignment mode of filtering according to the initial window size and the pixel number, wherein the data alignment mode of filtering comprises the following steps: n-alignment, m-alignment, and 1-alignment, where n is the number of pixels and m is a multiple of 4;

according to the data alignment mode, adjusting the filter comprises the following steps: expanding the window size of the filter to be a multiple of the data alignment mode, and supplementing the expansion bit number by 0;

2. The method of claim 1, wherein determining the data alignment of the filtering process based on the initial window size and the number of pixels comprises:

the first preset condition and the second preset condition are both related to the initial window size and the pixel number, the first preset condition comprises that the initial window size does not exceed a first preset value, the first preset value does not exceed half of the pixel number, the second preset condition comprises that the remainder obtained by dividing the initial window size by the pixel number is smaller than a second preset value, the second preset value does not exceed half of the pixel number, n is the pixel number, and m is a multiple of 4.

3. The method of claim 1, wherein determining the data alignment of the filtering process based on the initial window size and the number of pixels comprises:

4. An image filtering apparatus, comprising:

a second determining module, configured to determine, according to the initial window size and the number of pixels, a data alignment manner of the filtering process, where the data alignment manner of the filtering process includes: n-alignment, m-alignment, and 1-alignment, where n is the number of pixels and m is a multiple of 4;

The adjusting module is configured to adjust the filter according to the data alignment mode, and includes: expanding the window size of the filter to be a multiple of the data alignment mode, and supplementing the expansion bit number by 0;

5. The apparatus of claim 4, wherein the second determining module comprises:

6. The apparatus of claim 4, wherein the second determining module comprises:

7. An electronic device comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by one or more processors, the one or more programs comprising instructions for:

8. A readable storage medium, characterized in that instructions in the storage medium, when executed by a processor of an electronic device, enable the electronic device to perform the image filtering method according to one or more of the method claims 1-3.