CN111899149B - Image processing method and device based on operator fusion and storage medium - Google Patents
Image processing method and device based on operator fusion and storage medium Download PDFInfo
- Publication number
- CN111899149B CN111899149B CN202010657863.1A CN202010657863A CN111899149B CN 111899149 B CN111899149 B CN 111899149B CN 202010657863 A CN202010657863 A CN 202010657863A CN 111899149 B CN111899149 B CN 111899149B
- Authority
- CN
- China
- Prior art keywords
- target
- atom
- target operation
- atoms
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/20—Processor architectures; Processor configuration, e.g. pipelining
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses an image processing method and device based on operator fusion and a storage medium. Wherein the method comprises the following steps: acquiring an operation atom for executing a target operation on a target image, wherein the operation atom represents a calculation type for executing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms; and combining the operators with the same target operation to obtain a combined target image, namely removing repeated addressing in a storage medium through the operator combining process of the same atomic operation, and loading data and storing the data, so that the performance of the multiple operators when being used simultaneously is improved, the access times of the storage medium and the pressure of the transmission medium are reduced, and the technical problem of lower processing performance caused by excessive access times of the multiple operators in the image processing process in the prior art is solved.
Description
Technical Field
The invention relates to the field of image processing, in particular to an image processing method and device based on operator fusion and a storage medium.
Background
The currently used image processing algorithm needs to call different operator interfaces respectively aiming at different operators, so that the performance of the algorithm does not reach an optimal scheme, and even if the performance of each platform is optimized to the greatest extent aiming at the operators, unnecessary performance and occupation of equipment processing and transmission bandwidth caused by repeated reloading and storing of the same atoms aiming at the same image data can exist, so that the performance cannot reach the optimal standard.
In view of the above problems, no effective solution has been proposed at present.
Disclosure of Invention
The embodiment of the invention provides an image processing method, an image processing device and a storage medium based on operator fusion, which at least solve the technical problem of lower processing performance caused by excessive access times of a plurality of operators in the image processing process in the prior art.
According to an aspect of an embodiment of the present invention, there is provided an image processing method based on operator fusion, including: acquiring an operation atom for executing a target operation on a target image, wherein the operation atom represents a calculation type for executing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing the target operation, and determining target operation operators corresponding to the target operation atoms; and combining the same target operation operators to obtain a combined target image.
According to another aspect of the embodiment of the present invention, there is also provided an image processing apparatus based on operator fusion, including: an acquisition unit configured to acquire an operation atom that performs a target operation on a target image, wherein the operation atom represents a calculation type of performing the target operation; a determining unit, configured to convert the operation atom into a corresponding target operation atom according to attribute information of a target device that performs the target operation, and determine a target operation operator corresponding to the target operation atom; and the merging unit is used for merging the target operators with the same operation operator to obtain a merged target image.
According to a further aspect of embodiments of the present invention, there is also provided a computer readable storage medium having a computer program stored therein, wherein the computer program is configured to perform the above operator fusion based image processing method at run-time.
According to still another aspect of the embodiments of the present invention, there is further provided an electronic device including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor executes the above-mentioned image processing method based on operator fusion through the computer program.
In the embodiment of the invention, the operation atom for executing the target operation on the target image is obtained, wherein the operation atom represents the calculation type for executing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms; the method comprises the steps of carrying out merging processing on the operators with the same target operation to obtain a target image after the merging processing, and carrying out merging processing on the operation atoms with the same operation operators according to attribute information of target equipment for processing the target image and operation atoms of the target image, namely, removing repeated addressing in a storage medium through the operator merging processing of the same operation operator, loading data and storing data, thereby improving the performance of the operators when the operators are used simultaneously, reducing the access times of the storage medium and the pressure of the transmission medium, and further solving the technical problem of lower processing performance caused by excessive access times of the operators in the image processing process in the prior art.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a schematic illustration of an application environment of an alternative operator fusion-based image processing method according to an embodiment of the present invention;
FIG. 2 is a flow chart of an alternative operator fusion-based image processing method according to an embodiment of the invention;
FIG. 3 is a flow chart of an alternative image processing method by operator fusion according to an embodiment of the invention;
FIG. 4 is a schematic diagram of an alternative operator fusion-based image processing apparatus according to an embodiment of the present invention;
Fig. 5 is a schematic structural diagram of an electronic device according to an alternative image processing method based on operator fusion according to an embodiment of the present invention.
Detailed Description
In order that those skilled in the art will better understand the present invention, a technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
According to an aspect of the embodiment of the present invention, an image processing method based on operator fusion is provided, optionally, as an optional implementation manner, the image processing method based on operator fusion may be applied, but not limited to, in a hardware environment as shown in fig. 1, where the image processing method may include, but is not limited to, a terminal device 102, a network 110, and a server 112. Wherein the terminal device 102 is provided with a view client for presenting the target image.
The terminal device 102 may include, but is not limited to: a human-machine interaction screen 104, a processor 106 and a memory 108. The man-machine interaction screen 104 is used for acquiring man-machine interaction instructions through a man-machine interaction interface and displaying target images; the processor 106 is configured to display a target image in response to the man-machine interaction instruction. The memory 108 is used to store target image attribute information. Here the server 112 may include, but is not limited to: the processing engine 116 is used for calling the target image stored in the database 114 and acquiring an operation atom for executing the target operation on the target image, wherein the operation atom represents the calculation type for executing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms; the method comprises the steps of carrying out merging processing on the operators with the same target operation to obtain a target image after the merging processing, and carrying out merging processing on the operation atoms with the same operation operators according to attribute information of target equipment for processing the target image and operation atoms of the target image, namely, removing repeated addressing in a storage medium through the operator merging processing of the same operation operator, loading data and storing data, thereby improving the performance of the operators when the operators are used simultaneously, reducing the access times of the storage medium and the pressure of the transmission medium, and further solving the technical problem of lower processing performance caused by excessive access times of the operators in the image processing process in the prior art.
The specific process comprises the following steps: the man-machine interaction screen 104 in the terminal device 102 displays an interaction interface (a game screen in a shooting game is shown in fig. 1) where the game client runs a game task. The target image is acquired and sent to the server 112 via the network 110 as in steps S102-S110. Acquiring an operation atom for performing a target operation on a target image at a server 112, wherein the operation atom represents a calculation type for performing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms; and combining the same target operation operators to obtain a combined target image. And then returns the result of the above determination to the terminal device 102.
Then, as by steps S102 to S110, the terminal device 102 acquires an operation atom for performing a target operation on the target image, wherein the operation atom represents a calculation type for performing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms; and combining the same target operators to obtain a combined target image, and combining the operation atoms with the same operators according to the attribute information of target equipment for processing the target image and the operation atoms of the target image. According to the attribute information of the target equipment for processing the target image and the operation atoms of the target image, the operation atoms with the same operation operator are combined. That is, by merging operators with the same atomic operation, repeated addressing in a storage medium is removed, and the processes of loading data and storing data are carried out, so that the performance of a plurality of operators when being used simultaneously is improved, the access times of the storage medium and the pressure of the transmission medium are reduced, and the technical problem of lower processing performance caused by excessive access times of a plurality of operators in the image processing process in the prior art is solved.
Alternatively, in this embodiment, the above image processing method based on operator fusion may be applied to, but not limited to, the server 112, for assisting processing of the target image in the application client. The application client may be, but not limited to, a terminal device 102, where the terminal device 102 may be, but not limited to, a terminal device supporting running of the application client, such as a mobile phone, a tablet computer, a notebook computer, a PC, etc. The server 112 and the terminal device 102 may implement data interaction through, but are not limited to, a network, which may include, but is not limited to, a wireless network or a wired network. Wherein the wireless network comprises: bluetooth, WIFI, and other networks that enable wireless communications. The wired network may include, but is not limited to: wide area network, metropolitan area network, local area network. The above is merely an example, and is not limited in any way in the present embodiment.
Optionally, as an optional embodiment, as shown in fig. 2, the above image processing method based on operator fusion includes:
step S202, an operation atom for performing a target operation on a target image is acquired, wherein the operation atom represents a calculation type for performing the target operation.
Step S204, converting the operation atoms into corresponding target operation atoms according to the attribute information of the target equipment executing the target operation, and determining target operation operators corresponding to the target operation atoms.
Step S206, combining the operation atoms with the same target operation operator to obtain a combined target image.
Alternatively, in the present embodiment, the above-described target image may include, but is not limited to, any form of image, such as a screenshot of a game process, a photographed landscape image, a photographed character image, or the like.
The operation atoms may include, but are not limited to, data used for performing image processing, for example, performing image processing according to pixel points, performing image processing through a matrix, and the like.
Alternatively, in the present embodiment, the target operator may include, but is not limited to, adding the operation atoms, phase processing, and the like.
It should be noted that, acquiring an operation atom for performing a target operation on a target image may include:
and acquiring an operation atom of the target image for executing the target operation according to the attribute information of the target device.
In this embodiment, converting an operation atom into a corresponding target operation atom according to attribute information of a target device executing a target operation, and determining a target operation operator corresponding to the target operation atom may include:
converting the operation atoms into target operation atoms according to preset rules;
and carrying out merging processing according to the target operation atoms based on the same target operation operators to obtain a target image after merging processing.
Wherein the atoms can be manipulated to be converted into the same number of bytes for processing.
It should be noted that, in the case that the operation atom is of a matrix type, converting the operation atom into the target operation atom according to the preset rule may include:
The matrix is converted into columns by a column-row conversion mode to serve as the minimum target operation atom.
It should be further noted that, in the case that the operation atom is of the pixel point type, converting the operation atom into the target operation atom according to the preset rule may include:
and obtaining the minimum vector of the pixel point, and taking the minimum vector as a target operation atom.
Optionally, in this embodiment, converting the operation atom into a corresponding target operation atom according to attribute information of a target device performing the target operation, and determining a target operation operator corresponding to the target operation atom may include:
Acquiring parameters of target equipment for processing a target image;
and determining attribute information of the target equipment according to the parameters.
According to the embodiment of the application, the operation atom for executing the target operation on the target image is obtained, wherein the operation atom represents the calculation type for executing the target operation; converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms; the method comprises the steps of carrying out merging processing on the operators with the same target operation to obtain a target image after the merging processing, and carrying out merging processing on the operation atoms with the same operation operators according to attribute information of target equipment for processing the target image and operation atoms of the target image, namely, removing repeated addressing in a storage medium through the operator merging processing of the same operation operator, loading data and storing data, thereby improving the performance of the operators when the operators are used simultaneously, reducing the access times of the storage medium and the pressure of the transmission medium, and further solving the technical problem of lower processing performance caused by excessive access times of the operators in the image processing process in the prior art.
As an alternative embodiment, the application also provides an alternative image processing method through operator fusion. As shown in fig. 3, a flowchart of an image processing method by operator fusion.
Step S301, selecting an operation atom;
The operation atoms may include, but are not limited to, pixel points, matrices, and the like.
According to the optimal operation atom principle instead of the minimum operation atom principle, different operation atoms are selected, the pixel points can calculate according to the parallel processing capacity of the device, the minimum vector capable of being operated is used as a basic operation atom, or atoms with proper operation are selected according to the size of a cache such as a cache, so that the calculation and the storage unit of the device can be fully utilized to improve the performance, and the atoms such as a matrix can be converted into columns serving as the minimum operation atoms in a row-column conversion mode.
Step S302, selecting a platform;
The platform may include, but is not limited to, a computing unit, a storage medium, and the like.
The method is further screened according to the parallel processing capacity and the cache size of different platforms, the selection can reasonably allocate the performance and the storage medium resource to achieve the required balance through parameter configuration, and the processing can also be performed in a performance optimal mode.
Step S303, selecting an operation data type;
the operation data types may include, but are not limited to, u8, int32, fioat, etc.
The optimization modes used for different lengths of different data types are different, and the needed optimization mode is selected according to the actual operation data types.
Step S304, selecting the same atomic operator;
The operation operators comprise add, sub, ADD WEIGHT and the like.
And selecting the operator combination of the same atomic operation required, judging whether operator merging operation can be performed according to the input combination, and multiplexing a source storage medium and a target storage medium operated by the operators according to an operator cascading mode.
Step S305, selecting an optimization mode;
Different optimization methods are adopted due to different factors such as operation atoms, data types, data sizes, platform caches and the like, and specific optimization methods can be used for specific operator combinations, such as: the multiply and add computations are combined into an accumulate computation unit.
Step S306, carrying out the loop internal combination calculation processing according to the selected calculation unit.
In this embodiment, the scenario processed by the image processing scheme is not data processing on a single atom, and is often a large number of atoms which are orderly arranged, and the atoms are processed according to a certain order, and the arrangement and the use modes of different operators to be processed are the same.
In this embodiment, redundancy overhead of the storage medium and the transmission medium is reduced, redundancy overhead of the repetition cycle is reduced, and data loading and storing are regularly and conveniently optimized
In this embodiment, in the case of lifting the mathematical layer, redundant operations of loading and storing the storage medium are considered to be removed, so that not only is the pressure of the storage medium and the transmission medium reduced, but also the cyclic overhead of repetition is reduced by combining the cyclic structures. After the merging calculation, the cost of loading and storing is saved, and the cost of loop jump and pipeline updating is saved.
Performance improvement: taking the improvement of the performance of the pure c code and the arm cpu as an example, the combination of arithmetic operators can be improved by about one time.
It should be further noted that, by eliminating repeated processes of addressing in the storage medium, loading data and storing data through the operator merging process of the same atomic operation, performance when multiple operators are used simultaneously is improved, and access times of the storage medium and pressure on the transmission medium are reduced.
In this implementation, the use condition is limited to the condition that the same atomic operations are processed under the condition of a certain cascade relationship, the atomic loading and storing sequences of the operations are different after the operators are fused, and the reloading or storing is required after the fusion, so that the performance cannot be improved and the pressure on the storage and transmission medium is reduced. If the stepless connection relation is too complex for input and output judging operation, the performance is finally affected, and the effect of performance optimization cannot be achieved if the calculation pressure is improved.
According to the embodiment provided by the application, the operator fusion method is applied to the general image processing or data processing process, so that other atomic calculations which are operated by the same atoms except the neural network operator and have specific relations can be fused, the processing efficiency is improved, and the patent proposal also solves the problem that the repeated loading and storing processes of different operators on data are reduced under the condition of loading and storing the data from a storage medium aiming at the operator application, and the performance is improved.
It should be noted that, for simplicity of description, the foregoing method embodiments are all described as a series of acts, but it should be understood by those skilled in the art that the present invention is not limited by the order of acts described, as some steps may be performed in other orders or concurrently in accordance with the present invention. Further, those skilled in the art will also appreciate that the embodiments described in the specification are all preferred embodiments, and that the acts and modules referred to are not necessarily required for the present invention.
According to another aspect of the embodiment of the present invention, there is also provided an operator fusion-based image processing apparatus for implementing the above operator fusion-based image processing method. As shown in fig. 4, the apparatus includes: an acquisition unit 41, a determination unit 43, and a merging unit 45.
An acquisition unit 41 for acquiring an operation atom for performing a target operation on a target image, wherein the operation atom represents a calculation type for performing the target operation.
A determining unit 43, configured to convert the operation atom into a corresponding target operation atom according to attribute information of a target device performing the target operation, and determine a target operator corresponding to the target operation atom.
And the merging unit 45 is configured to perform merging processing on the operation atoms with the same target operation operator, so as to obtain a target image after the merging processing.
Alternatively, in the present embodiment, the acquiring unit 41 may include:
and the first acquisition module is used for acquiring an operation atom of the target image for executing the target operation according to the attribute information of the target device.
Alternatively, in the present embodiment, the determining unit 43 may include:
The conversion module is used for converting the operation atoms into target operation atoms according to preset rules;
and the merging module is used for carrying out merging processing based on the same target operation operator according to the target operation atoms to obtain a target image after the merging processing.
The conversion module is further configured to convert, when the operation atom is of a matrix type, the matrix into a column as a minimum target operation atom by a column-row conversion method.
The conversion module is further configured to obtain a minimum vector of the pixel point and use the minimum vector as a target operation atom when the operation atom is of a pixel point type.
Alternatively, in the present embodiment, the determining unit 43 may include:
The second acquisition module is used for acquiring parameters of the target equipment for processing the target image;
And the determining module is used for determining the attribute information of the target equipment according to the parameters.
With the embodiment provided by the present application, the acquisition unit 41 acquires an operation atom that performs a target operation on a target image, wherein the operation atom represents a calculation type of performing the target operation; the determining unit 43 converts the operation atom into a corresponding target operation atom according to attribute information of a target device performing the target operation, and determines a target operator corresponding to the target operation atom; the merging unit 45 performs merging processing on the operation atoms having the same target operation operator, and obtains a target image after the merging processing. According to the attribute information of the target equipment for processing the target image and the operation atoms of the target image, the operation atoms with the same operation operator are combined, namely, repeated addressing in a storage medium is removed through the operator combining process of the same operation, the process of loading data and storing the data is carried out, so that the performance of a plurality of operators in simultaneous use is improved, the access times of the storage medium and the pressure of the transmission medium are reduced, and the technical problem of lower processing performance caused by excessive access times of the plurality of operators in the image processing process in the prior art is solved.
According to a further aspect of the embodiments of the present invention there is also provided an electronic device for implementing the above operator fusion based image processing method, as shown in fig. 5, the electronic device comprising a memory 502 and a processor 504, the memory 502 having stored therein a computer program, the processor 504 being arranged to perform the steps of any of the method embodiments described above by means of the computer program.
Alternatively, in this embodiment, the electronic apparatus may be located in at least one network device of a plurality of network devices of the computer network.
Alternatively, in the present embodiment, the above-described processor may be configured to execute the following steps by a computer program:
s1, acquiring an operation atom for executing a target operation on a target image, wherein the operation atom represents a calculation type for executing the target operation;
S2, converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms;
S3, combining the operation atoms with the same target operation operator to obtain a combined target image.
Alternatively, it will be understood by those skilled in the art that the structure shown in fig. 5 is only schematic, and the electronic device may also be a terminal device such as a smart phone (e.g. an Android phone, an iOS phone, etc.), a tablet computer, a palm computer, and a Mobile internet device (Mobile INTERNET DEVICES, MID), a PAD, etc. Fig. 5 is not limited to the structure of the electronic device. For example, the electronic device may also include more or fewer components (e.g., network interfaces, etc.) than shown in FIG. 5, or have a different configuration than shown in FIG. 5.
The memory 502 may be used to store software programs and modules, such as program instructions/modules corresponding to the image processing method and apparatus based on operator fusion in the embodiment of the present invention, and the processor 504 executes the software programs and modules stored in the memory 502, thereby performing various functional applications and data processing, that is, implementing the image processing method based on operator fusion. Memory 502 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 502 may further include memory located remotely from processor 504, which may be connected to the terminal 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. The memory 502 may be, but is not limited to, information for the target image and an operation atom corresponding to the target image. As an example, as shown in fig. 5, the above memory 502 may include, but is not limited to, the acquisition unit 41, the determination unit 43, and the merging unit 45 in the above operator fusion-based image processing apparatus. In addition, other module units in the image processing apparatus based on operator fusion may be included, but are not limited to, and are not described in detail in this example.
Optionally, the transmission device 506 is configured to receive or transmit data via a network. Specific examples of the network described above may include wired networks and wireless networks. In one example, the transmission device 506 includes a network adapter (Network Interface Controller, NIC) that may be connected to other network devices and routers via a network cable to communicate with the internet or a local area network. In one example, the transmission device 506 is a Radio Frequency (RF) module, which is used to communicate with the internet wirelessly.
In addition, the electronic device further includes: a display 508 for displaying the combined target image information; and a connection bus 510 for connecting the respective module parts in the above-described electronic device.
According to a further aspect of embodiments of the present invention, there is also provided a computer readable storage medium having a computer program stored therein, wherein the computer program is arranged to perform the steps of any of the method embodiments described above when run.
Alternatively, in the present embodiment, the above-described computer-readable storage medium may be configured to store a computer program for executing the steps of:
s1, acquiring an operation atom for executing a target operation on a target image, wherein the operation atom represents a calculation type for executing the target operation;
S2, converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing target operation, and determining target operation operators corresponding to the target operation atoms;
S3, combining the operation atoms with the same target operation operator to obtain a combined target image.
Alternatively, in this embodiment, it will be understood by those skilled in the art that all or part of the steps in the methods of the above embodiments may be performed by a program for instructing a terminal device to execute the steps, where the program may be stored in a computer readable storage medium, and the storage medium may include: flash disk, read-Only Memory (ROM), random-access Memory (Random Access Memory, RAM), magnetic disk or optical disk, etc.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
The integrated units in the above embodiments may be stored in the above-described computer-readable storage medium if implemented in the form of software functional units and sold or used as separate products. Based on such understanding, the technical solution of the present invention may be embodied in essence or a part contributing to the prior art or all or part of the technical solution in the form of a software product stored in a storage medium, comprising several instructions for causing one or more computer devices (which may be personal computers, servers or network devices, etc.) to perform all or part of the steps of the method described in the embodiments of the present invention.
In the foregoing embodiments of the present invention, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In several embodiments provided by the present application, it should be understood that the disclosed client may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, such as the division of the units, is merely a logical function division, and may be implemented in another manner, for example, multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The foregoing is merely a preferred embodiment of the present invention and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present invention, which are intended to be comprehended within the scope of the present invention.
Claims (9)
1. An image processing method based on operator fusion, which is characterized by comprising the following steps:
Acquiring an operation atom for executing a target operation on a target image, wherein the operation atom represents a calculation type for executing the target operation;
Converting the operation atoms into corresponding target operation atoms according to attribute information of target equipment for executing the target operation, and determining target operation operators corresponding to the target operation atoms; combining the operation atoms with the same target operation operator to obtain a combined target image;
the acquiring an operation atom for executing a target operation on a target image includes: and acquiring the operation atom of the target image for executing the target operation according to the attribute information of the target device, wherein the attribute information comprises a cache size.
2. The method of claim 1, converting the operation atom into a corresponding target operation atom according to attribute information of a target device performing the target operation, and determining a target operation operator of the corresponding target operation atom, comprising:
Converting the operation atoms into the target operation atoms according to preset rules;
And carrying out merging processing according to the target operation atoms based on the same target operation operators to obtain the target image after merging processing.
3. The method according to claim 1, in the case that the operation atom is of a matrix type, converting the operation atom into the target operation atom according to a preset rule, comprising:
and converting the matrix into columns serving as the minimum target operation atoms in a column-row conversion mode.
4. The method according to claim 1, in the case that the operation atom is of a pixel point type, converting the operation atom into the target operation atom according to a preset rule, including:
and obtaining the minimum vector of the pixel point, and taking the minimum vector as the target operation atom.
5. The method of claim 1, converting the operation atom into a corresponding target operation atom according to attribute information of a target device performing the target operation, and determining a target operation operator of the corresponding target operation atom, comprising:
Acquiring parameters of the target equipment for processing a target image;
And determining attribute information of the target equipment according to the parameters.
6. An image processing apparatus based on operator fusion, comprising:
An acquisition unit configured to acquire an operation atom that performs a target operation on a target image, wherein the operation atom represents a calculation type of performing the target operation;
A determining unit, configured to convert the operation atom into a corresponding target operation atom according to attribute information of a target device that performs the target operation, and determine a target operation operator corresponding to the target operation atom;
The merging unit is used for merging the operation atoms with the same target operation operator to obtain a merged target image;
The acquisition unit includes: the first acquisition module is used for acquiring the operation atom of the target image for executing the target operation according to the attribute information of the target device, wherein the attribute information comprises a cache size.
7. The apparatus of claim 6, the determining unit comprising:
The conversion module is used for converting the operation atoms into the target operation atoms according to preset rules;
and the merging module is used for carrying out merging processing based on the same target operation operator according to the target operation atoms to obtain the target image after the merging processing.
8. The apparatus of claim 6, the determining unit comprising:
The second acquisition module is used for acquiring parameters of the target equipment for processing the target image;
and the determining module is used for determining the attribute information of the target equipment according to the parameters.
9. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored program, wherein the program when run performs the method of any of the preceding claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010657863.1A CN111899149B (en) | 2020-07-09 | 2020-07-09 | Image processing method and device based on operator fusion and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010657863.1A CN111899149B (en) | 2020-07-09 | 2020-07-09 | Image processing method and device based on operator fusion and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111899149A CN111899149A (en) | 2020-11-06 |
| CN111899149B true CN111899149B (en) | 2024-07-02 |
Family
ID=73192134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010657863.1A Active CN111899149B (en) | 2020-07-09 | 2020-07-09 | Image processing method and device based on operator fusion and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111899149B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114254563A (en) * | 2021-12-20 | 2022-03-29 | Oppo广东移动通信有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN114492737B (en) * | 2021-12-31 | 2022-12-09 | 北京百度网讯科技有限公司 | Data processing method, data processing device, electronic equipment, storage medium and program product |
| CN114615519B (en) * | 2022-01-27 | 2024-06-18 | 百果园技术(新加坡)有限公司 | Video processing method, device, equipment and storage medium |
| CN114881214B (en) * | 2022-05-17 | 2025-09-16 | 北京灵汐科技有限公司 | Processing method and processing device for neural network calculation graph |
Family Cites Families (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8478946B2 (en) * | 2009-09-08 | 2013-07-02 | Advanced Micro Devices, Inc. | Method and system for local data sharing |
| WO2016041126A1 (en) * | 2014-09-15 | 2016-03-24 | 华为技术有限公司 | Method and device for processing data stream based on gpu |
| CN108960012B (en) * | 2017-05-22 | 2022-04-15 | 中科创达软件股份有限公司 | Feature point detection method and device and electronic equipment |
| CN107798714A (en) * | 2017-09-11 | 2018-03-13 | 深圳创维数字技术有限公司 | A kind of image data display method and relevant apparatus and computer-readable storage medium |
| CN110007748B (en) * | 2018-01-05 | 2021-02-19 | Oppo广东移动通信有限公司 | Terminal control method, processing device, storage medium and terminal |
| CN108805901B (en) * | 2018-05-04 | 2022-02-22 | 北京航空航天大学 | Visual target rapid detection tracking parallel computing and fusion method based on multi-core DSP |
| CN109033323B (en) * | 2018-07-18 | 2020-10-23 | 中国人民解放军91776部队 | Tree structure basic data change recording method based on operator |
| CN110780982A (en) * | 2018-07-27 | 2020-02-11 | 深圳百迈技术有限公司 | An image processing method, device and equipment |
| CN109345497B (en) * | 2018-09-14 | 2021-12-03 | 淮阴师范学院 | Image fusion processing method and system based on fuzzy operator and computer program |
| US20190392296A1 (en) * | 2019-06-28 | 2019-12-26 | John Brady | Hardware agnostic deep neural network compiler |
| CN110780921B (en) * | 2019-08-30 | 2023-09-26 | 腾讯科技（深圳）有限公司 | Data processing method and device, storage medium and electronic device |
| CN110533742B (en) * | 2019-09-03 | 2021-05-11 | 广州视源电子科技股份有限公司 | Image color filling method, device, equipment and storage medium |
| CN111309479B (en) * | 2020-02-14 | 2023-06-06 | 北京百度网讯科技有限公司 | Method, device, equipment and medium for realizing task parallel processing |
-
2020
- 2020-07-09 CN CN202010657863.1A patent/CN111899149B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 一种适用于GPU图像处理算法的合并存储结构;左宪禹等;计算机工程与科学;第42卷(第02期);正文第1-4节，附图1-7 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111899149A (en) | 2020-11-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111899149B (en) | Image processing method and device based on operator fusion and storage medium | |
| US20190324809A1 (en) | Method, apparatus, and computer program product for processing computing task | |
| KR20210005035A (en) | Renewable smart contract | |
| EP3678030B1 (en) | Distributed system for executing machine learning, and method therefor | |
| CN109992406B (en) | Picture request method, picture request response method and client | |
| CN111723933A (en) | Training method of neural network model and related product | |
| CN110334074B (en) | Data processing method, device, server and storage medium | |
| CN115880132B (en) | Graphics processor, matrix multiplication task processing method, device and storage medium | |
| CN110489126A (en) | Execution method and apparatus, storage medium and the electronic device of compiler task | |
| CN112632566A (en) | Vulnerability scanning method and device, storage medium and electronic equipment | |
| CN116991560B (en) | Parallel scheduling method, device, equipment and storage medium for language model | |
| CN113703407A (en) | Method, system and equipment for constructing robot production line operating system based on digital twin | |
| CN112953994A (en) | Data acquisition method, acquisition device, terminal equipment and readable storage medium | |
| CN112965809A (en) | Deep learning task processing system and method | |
| CN113254215B (en) | Data processing method and device, storage medium and electronic equipment | |
| EP3479235A1 (en) | Real-time application behavior changes | |
| CN116228544B (en) | Image processing method and device and computer equipment | |
| CN113553489B (en) | Method, device, equipment, medium and program product for content capture | |
| CN118536615A (en) | Data processing method, device, equipment and storage medium based on model call | |
| CN112416470B (en) | Command execution method and device, storage medium and electronic device | |
| CN103617052B (en) | Method and device for processing cache of application program | |
| CN114816575A (en) | Component loading method, hosting method, system, electronic device and storage medium | |
| CN110737780A (en) | Method and device for sending data and method and device for receiving data | |
| CN115729518A (en) | Information processing method and apparatus, storage medium, and electronic device | |
| CN114675872A (en) | Data processing method, device and equipment for application program and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |