CN110278370B - Method and device for automatically generating shooting control mechanism and electronic equipment - Google Patents
Method and device for automatically generating shooting control mechanism and electronic equipment Download PDFInfo
- Publication number
- CN110278370B CN110278370B CN201910542321.7A CN201910542321A CN110278370B CN 110278370 B CN110278370 B CN 110278370B CN 201910542321 A CN201910542321 A CN 201910542321A CN 110278370 B CN110278370 B CN 110278370B
- Authority
- CN
- China
- Prior art keywords
- shooting
- image data
- control mechanism
- requirement
- calculation
- 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
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/617—Upgrading or updating of programs or applications for camera control
Abstract
The embodiment of the invention provides a method, a device and electronic equipment for automatically generating a shooting control mechanism, wherein the method comprises the following steps: providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect; acquiring a shooting control mechanism generated by parameter search and architecture search according to the shooting requirement and the image data; and carrying out shooting enabling by utilizing the shooting control mechanism so as to realize shooting operation meeting the shooting requirements of the user. Therefore, the embodiment of the invention can automatically obtain the shooting control mechanism according to the shooting requirement of the user and the image data which corresponds to the shooting requirement and reflects the shooting effect without specific shooting skill of the user or shooting skill learning, thereby completing the shooting operation aiming at the shooting requirement.
Description
Technical Field
The embodiment of the invention relates to the technical field of internet, in particular to a method, a device and electronic equipment for automatically generating a shooting control mechanism.
Background
Because the portable shooting device can complete scene shooting more timely and conveniently, more and more users adopt the portable shooting device to meet specific shooting requirements, and the specific shooting requirements often need some specific shooting skills to set shooting parameters of the portable shooting device. The shooting skill of the user cannot set the shooting parameters, so that the shooting requirements of the shot images cannot be met, and the realization of specific shooting requirements is limited.
The portable photographing apparatus generally provides only general photographing guidance, and if a user needs to realize a specific photographing requirement, the user needs to fumble about a photographing skill through continuous experimental photographing to solve the specific photographing requirement through a grasped photographing skill. For users who are not good at shooting, the users can only use public communities such as forums and the like to search shooting skills shared by other people so as to realize specific shooting requirements.
Therefore, how to easily realize a specific shooting requirement becomes a technical problem which needs to be solved urgently in the prior art.
Disclosure of Invention
Embodiments of the present invention provide a method, an apparatus, and an electronic device for automatically generating a shooting control mechanism, so as to solve or partially solve the above problems.
According to a first aspect of the embodiments of the present invention, there is provided a method for automatically generating a shooting control mechanism, the method including: providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect; acquiring a shooting control mechanism generated by parameter search and architecture search according to the shooting requirement and the image data; and carrying out shooting enabling by utilizing the shooting control mechanism so as to realize shooting operation meeting the shooting requirements of the user.
According to a second aspect of the embodiments of the present invention, there is provided an apparatus for automatically generating a shooting control mechanism, the apparatus including: the data providing module is used for providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect; the mechanism generation module is used for obtaining a shooting control mechanism which carries out parameter search and architecture search generation according to the shooting requirement and the image data; and the shooting enabling module is used for utilizing the shooting control mechanism to carry out shooting enabling so as to realize shooting operation meeting the shooting requirements of the user.
According to a third aspect of embodiments of the present invention, there is provided an electronic apparatus, including: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus; the memory is used for storing at least one executable instruction, and the executable instruction causes the processor to execute the corresponding operation of the method according to the first aspect.
According to a fourth aspect of embodiments of the present invention, there is provided a computer storage medium having stored thereon a computer program which, when executed by a processor, implements the method according to the first aspect.
According to the embodiment of the invention, at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects the shooting effect are provided, so that a shooting control mechanism which is generated by parameter search and architecture search according to the shooting requirement and the image data is obtained, and the shooting control mechanism is enabled, so that the shooting operation meeting the shooting requirement of the user is realized. Therefore, the embodiment of the invention can automatically obtain the shooting control mechanism according to the shooting requirement of the user and the image data which corresponds to the shooting requirement and reflects the shooting effect without specific shooting skill of the user or shooting skill learning, thereby completing the shooting operation aiming at the shooting requirement.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the embodiments of the present invention, and it is also possible for a person skilled in the art to obtain other drawings based on the drawings.
Fig. 1 is a flowchart illustrating steps of a method for automatically generating a shooting control mechanism according to a first embodiment of the present invention;
fig. 2 is a flowchart illustrating steps of a method for generating a shooting control mechanism according to a first embodiment of the present invention;
fig. 3 is a block diagram of an apparatus for automatically generating a shooting control mechanism according to a second embodiment of the present invention;
fig. 4 is a schematic structural diagram of an electronic device according to a third embodiment of the present invention.
Detailed Description
In order to make those skilled in the art better understand the technical solutions in the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments obtained by a person skilled in the art based on the embodiments of the present invention shall fall within the scope of the protection of the embodiments of the present invention.
The following further describes specific implementation of the embodiments of the present invention with reference to the drawings.
Example one
Referring to fig. 1, a flowchart illustrating steps of a method for automatically generating a shooting control mechanism according to a first embodiment of the present invention is shown.
The method for automatically generating the shooting control mechanism comprises the following steps:
step S102: providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect.
In a specific implementation of the present invention, the shooting requirement includes:
at least one of the conditions of shooting scene, shooting light and object motion.
The object includes: at least one of a human, an animal, and an object.
Specifically, the object is a body occupying an area of the screen 1/4 or more.
The shooting control mechanism meeting the shooting requirements is determined according to the shooting requirements formed by at least one of the shooting scene, the shooting light and the object motion conditions and the image data which corresponds to the shooting requirements and reflects the shooting effect.
Specifically, the image data which corresponds to the shooting requirement and reflects the shooting effect comprises at least one frame of image which clearly reflects the shooting effect.
The shooting device applied by the invention is generally various mobile shooting devices such as a camera, a mobile phone, a tablet personal computer and the like.
Step S104: and acquiring a shooting control mechanism generated by parameter search and architecture search according to the shooting requirement and the image data.
In a specific implementation of the present invention, referring to fig. 2, the generating of the shooting control mechanism includes:
step T102: and establishing an image data set containing image data and a shooting requirement label according to the at least one shooting requirement and the image data.
Specifically, if the first three shooting scenes with the highest occurrence frequency, the first three shooting rays and the first three object motion conditions are arranged and combined, twenty-seven shooting requirements are formed. For twenty-seven shooting requirements, twenty-seven image data sets are constructed, each image data set containing image data and a shooting requirement label.
Step T104: and performing classification calculation on the image data set, and performing image enhancement calculation on the image data in the image data set.
Due to different shooting scenes, shooting light rays and the conditions of different degrees of blurring, distortion and the like of images shot under the condition of object motion. Different image optimization methods are selected according to different shooting requirements.
Specifically, the invention selects a lightweight classification infrastructure network architecture mobilenetV2 to perform classification calculation on the image data set.
In order to meet the severe requirements of time consumption and calculation amount of a mobile terminal, the embodiment of the invention selects mobilenetV2 for classification calculation. The mobilenetV2 adopts the idea of a lightweight convolution neural network, uses a deep separable convolution as a basic convolution module, adopts the idea of expanding and then compressing, and adds a 1 × 1 expansion layer before the deep separable convolution, so that the model is lightened and the loss of information is reduced. Meanwhile, the deep separable convolved activation layer does not use Relu, but uses linear activation to prevent the destruction of features. Firstly, a scene classification data set is established, and a well-classified model is obtained by using ImageNet pre-trained mobilenetV2 transfer learning.
Specifically, the invention selects Deep Retinex composition for Low-Light Enhancement (RetinexNet) image Enhancement algorithm to perform image Enhancement calculation on the image data in the image data set.
The Retinex theory assumes that the observation image can be decomposed into a reflectance image and an illumination image. Most existing Retinex-based approaches elaborate manual constraints and parameters for this highly ill-posed decomposition, which may be limited by the model capacity when applied to various scenarios. Retinex includes a decomposition network (Decom) for decomposition and an enhancement network (Relight) for lighting adjustment. During the training process of the decomposed network, ground truth values of reflection and illumination are not decomposed. The learning of the network has only key constraints including consistent reflectivity for low/normal light image sharing, and smoothness of illumination. On the basis of decomposition, the subsequent brightness enhancement is carried out on illumination by utilizing an enhancement network, and the denoising operation is carried out on the reflectivity by combining denoising. The retina mesh is trainable end-to-end, and the nature of the learning decomposition facilitates the adjustment of brightness.
Step T106: and taking the image data set of the classification calculation and the image enhancement calculation as a basic network architecture, and carrying out training of neural network architecture search calculation to obtain a shooting control mechanism aiming at the shooting requirement.
Specifically, the invention adopts an image data set which is classified and calculated according to a lightweight classification basic Network Architecture mobilenetV2 and an image data set which is enhanced and calculated by using an image enhancement algorithm RDL to carry out neural Network Architecture Search algorithm Path-Level Network Transformation for Efficient Architecture Search, trains the neural Network Architecture Search and calculation, and obtains a shooting control mechanism aiming at the shooting requirement.
Specifically, the invention adopts a neural Network Architecture Search algorithm for Efficient Architecture Search, and obtains an optimal Network by searching with mobilenetV2 and RetinexNet as basic networks. The embodiment of the invention designs an optimal network evaluation method, which judges whether the searched network learns an image beautifying mechanism suitable in different scenes according to the score value output by the aesthetic scoring network. The method comprises the following specific steps: and after the searched network, cascading a pre-trained image aesthetic scoring network, and using a scene classification data set as a training set. The network continuously learns the optimization, and finally, the aesthetic score of the image reaches a higher level. Through the series of training, the obtained network has the performance of classifying scenes and performing specific enhancement optimization on specific scenes.
The step T106 further includes:
and searching the network architecture cascade obtained by the neural network architecture searching, calculating and training and the trained image aesthetic scoring network to obtain an optimized shooting control mechanism aiming at the shooting requirement.
Specifically, twenty-seven data sets are respectively used as training data to search for the network architecture cascade obtained by searching, calculating and training the neural network architecture and the trained image aesthetic scoring network, so that twenty-seven optimized shooting control mechanisms are obtained. The twenty-seven optimized shooting control mechanisms have targeted optimization for twenty-seven shooting requirements.
The shooting control mechanism can be generated by parameter search and architecture search of the electronic equipment applied to the embodiment of the invention, and can also be sent to the electronic equipment applied to the embodiment of the invention after being generated by parameter search and architecture search of the server side.
Step S106: and carrying out shooting enabling by utilizing the shooting control mechanism so as to realize shooting operation meeting the shooting requirements of the user.
For example, if a user needs to take a picture of a soccer game, the user needs to provide user requirements, such as: grass, weak light, fast motion, and provide captured image data. The invention carries out parameter search and architecture search to generate a shooting control mechanism according to the shooting requirement and the image data, enables the shooting control mechanism aiming at grassland, weak light and rapid movement to a shooting device used by a user, and calls the shooting control mechanism aiming at grassland, weak light and rapid movement to carry out image shooting.
According to the embodiment of the invention, at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects the shooting effect are provided, so that a shooting control mechanism which is generated by parameter search and architecture search according to the shooting requirement and the image data is obtained, and the shooting control mechanism is enabled, so that the shooting operation meeting the shooting requirement of the user is realized. Therefore, the embodiment of the invention can automatically obtain the shooting control mechanism according to the shooting requirement of the user and the image data which corresponds to the shooting requirement and reflects the shooting effect without specific shooting skill of the user or shooting skill learning, thereby completing the shooting operation aiming at the shooting requirement.
The method of automatically generating a shooting control mechanism of the present embodiment may be performed by any suitable electronic device having data processing capabilities, including but not limited to: a server, a mobile terminal (such as a tablet computer, a mobile phone, etc.), a PC, etc., may also be an image processing unit installed on an electronic device, such as an image processing unit installed on a drone controller, an image processing unit installed on a camera, etc., and the image processing device may also be a pan/tilt apparatus.
Example two
Referring to fig. 3, a block diagram of an apparatus for automatically generating a shooting control mechanism according to a second embodiment of the present invention is shown.
The apparatus for automatically generating a shooting control mechanism of the present embodiment includes: a data providing module 301, a mechanism generating module 302, and a shooting enabling module 303.
The data providing module 301 is configured to provide at least one shooting requirement of a user and image data corresponding to the shooting requirement and reflecting a shooting effect.
The mechanism generating module 302 is configured to obtain a shooting control mechanism that performs parameter search and architecture search according to the shooting requirement and the image data.
The shooting enabling module 303 is configured to perform shooting enabling by using the shooting control mechanism to achieve a shooting operation that meets the shooting requirements of the user.
In a specific implementation of the present invention, the shooting requirement includes:
at least one of the conditions of shooting scene, shooting light and object motion.
The object includes: at least one of a human, an animal, and an object.
Specifically, the object is a body occupying an area of the screen 1/4 or more.
The shooting control mechanism meeting the shooting requirements is determined according to the shooting requirements formed by at least one of the shooting scene, the shooting light and the object motion conditions and the image data which corresponds to the shooting requirements and reflects the shooting effect.
Specifically, the image data which corresponds to the shooting requirement and reflects the shooting effect comprises at least one frame of image which clearly reflects the shooting effect.
The shooting device applied by the invention is generally various mobile shooting devices such as a camera, a mobile phone, a tablet personal computer and the like.
In a specific implementation of the present invention, referring to fig. 2, the generating of the shooting control mechanism includes:
step T102: and establishing an image data set containing image data and a shooting requirement label according to the at least one shooting requirement and the image data.
Specifically, if the first three shooting scenes with the highest occurrence frequency, the first three shooting rays and the first three object motion conditions are arranged and combined, twenty-seven shooting requirements are formed. For twenty-seven shooting requirements, twenty-seven image data sets are constructed, each image data set containing image data and a shooting requirement label.
Step T104: and performing classification calculation on the image data set, and performing image enhancement calculation on the image data in the image data set.
Due to different shooting scenes, shooting light rays and the conditions of different degrees of blurring, distortion and the like of images shot under the condition of object motion. Different image optimization methods are selected according to different shooting requirements.
Specifically, the invention selects a lightweight classification infrastructure network architecture mobilenetV2 to perform classification calculation on the image data set.
In order to meet the strict requirements of time consumption and calculation amount of a mobile terminal, the embodiment of the invention selects mobilenetV 2. The mobilenetV2 adopts the idea of a lightweight convolution neural network, uses a deep separable convolution as a basic convolution module, adopts the idea of expanding and then compressing, and adds a 1 × 1 expansion layer before the deep separable convolution, so that the model is lightweight and the loss of information is reduced. Meanwhile, the deep separable convolved activation layer does not use Relu, but uses linear activation to prevent the destruction of features. Firstly, a scene classification data set is established, and a well-classified model is obtained by using ImageNet pre-trained mobilenetV2 transfer learning.
Specifically, the invention selects Deep Retinex composition for Low-Light Enhancement (RetinexNet) image Enhancement algorithm to perform image Enhancement calculation on the image data in the image data set.
The Retinex theory assumes that the observation image can be decomposed into a reflectance image and an illumination image. Most existing Retinex-based approaches elaborate manual constraints and parameters for this highly ill-posed decomposition, which may be limited by the model capacity when applied to various scenarios. Retinex includes a decomposition network (Decom) for decomposition and an enhancement network (Relight) for lighting adjustment. During the training process of the decomposed network, ground truth values of reflection and illumination are not decomposed. The learning of the network has only key constraints including consistent reflectivity for low/normal light image sharing, and smoothness of illumination. On the basis of decomposition, the subsequent brightness enhancement is carried out on illumination by utilizing an enhancement network, and the denoising operation is carried out on the reflectivity by combining denoising. The retina mesh is trainable end-to-end, and the nature of the learning decomposition facilitates the adjustment of brightness.
Step T106: and taking the image data set of the classification calculation and the image enhancement calculation as a basic network architecture, and carrying out training of neural network architecture search calculation to obtain a shooting control mechanism aiming at the shooting requirement.
Specifically, the invention adopts an image data set which is classified and calculated according to a lightweight classification basic Network Architecture mobilenetV2 and an image data set which is enhanced and calculated by using an image enhancement algorithm RDL to carry out neural Network Architecture Search algorithm Path-Level Network Transformation for Efficient Architecture Search, trains the neural Network Architecture Search and calculation, and obtains a shooting control mechanism aiming at the shooting requirement.
Specifically, the invention adopts a neural Network Architecture Search algorithm for Efficient Architecture Search, and obtains an optimal Network by searching with mobilenetV2 and RetinexNet as basic networks. A method for judging an optimal network is designed, and whether the network obtained through searching learns an image beautifying mechanism suitable in different scenes or not is judged according to a score value output by an aesthetic scoring network. The method comprises the following specific steps: and after the searched network, cascading a pre-trained image aesthetic scoring network, and using a scene classification data set as a training set. The network continuously learns the optimization, and finally, the aesthetic score of the image reaches a higher level. Through the series of training, the obtained network has the performance of classifying scenes and performing specific enhancement optimization on specific scenes.
The step T106 further includes:
and searching the network architecture cascade obtained by the neural network architecture searching, calculating and training and the trained image aesthetic scoring network to obtain an optimized shooting control mechanism aiming at the shooting requirement.
Specifically, twenty-seven data sets are respectively used as training data to search for the network architecture cascade obtained by searching, calculating and training the neural network architecture and the trained image aesthetic scoring network, so that twenty-seven optimized shooting control mechanisms are obtained. The twenty-seven optimized shooting control mechanisms have targeted optimization for twenty-seven shooting requirements.
For example, if a user needs to take a picture of a soccer game, the user needs to provide user requirements, such as: grass, weak light, fast motion, and provide captured image data. The invention carries out parameter search and architecture search to generate a shooting control mechanism according to the shooting requirement and the image data, enables the shooting control mechanism aiming at grassland, weak light and rapid movement to a shooting device used by a user, and calls the shooting control mechanism aiming at grassland, weak light and rapid movement to carry out image shooting.
The shooting control mechanism can be generated by parameter search and architecture search of the electronic equipment applied to the embodiment of the invention, and can also be sent to the electronic equipment applied to the embodiment of the invention after being generated by parameter search and architecture search of the server side.
According to the embodiment of the invention, at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects the shooting effect are provided, so that a shooting control mechanism which is generated by parameter search and architecture search according to the shooting requirement and the image data is obtained, and the shooting control mechanism is enabled, so that the shooting operation meeting the shooting requirement of the user is realized. Therefore, the embodiment of the invention can automatically obtain the shooting control mechanism according to the shooting requirement of the user and the image data which corresponds to the shooting requirement and reflects the shooting effect without specific shooting skill of the user or shooting skill learning, thereby completing the shooting operation aiming at the shooting requirement.
The apparatus for automatically generating a shooting control mechanism of the present embodiment may be executed by any suitable electronic device with data processing capability, including but not limited to: a server, a mobile terminal (such as a tablet computer, a mobile phone, etc.), a PC, etc., may also be an image processing unit installed on an electronic device, such as an image processing unit installed on a drone controller, an image processing unit installed on a camera, etc., and the image processing device may also be a pan/tilt apparatus.
EXAMPLE III
Referring to fig. 4, a schematic structural diagram of an electronic device according to a third embodiment of the present invention is shown, and the specific embodiment of the present invention does not limit the specific implementation of the electronic device.
As shown in fig. 4, the electronic device may include: a processor (processor)402, a Communications Interface 404, a memory 406, and a Communications bus 408.
Wherein:
the processor 402, communication interface 404, and memory 406 communicate with each other via a communication bus 408.
A communication interface 404 for communicating with other electronic devices such as a terminal device or a server.
The processor 402 is configured to execute the program 410, and may specifically execute the relevant steps in the above-described method embodiment of downloading the application.
In particular, program 410 may include program code comprising computer operating instructions.
The processor 402 may be a central processing unit CPU or an application Specific Integrated circuit asic or one or more Integrated circuits configured to implement embodiments of the present invention. The electronic device comprises one or more processors, which can be the same type of processor, such as one or more CPUs; or may be different types of processors such as one or more CPUs and one or more ASICs.
And a memory 406 for storing a program 410. Memory 406 may comprise high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
The program 410 may specifically be configured to cause the processor 402 to perform the following operations: providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect; acquiring a shooting control mechanism generated by parameter search and architecture search according to the shooting requirement and the image data; and carrying out shooting enabling by utilizing the shooting control mechanism so as to realize shooting operation meeting the shooting requirements of the user.
In an optional implementation, the shooting requirement includes: at least one of the conditions of shooting scene, shooting light and object motion.
In an alternative embodiment, the object comprises: at least one of a human, an animal, and an object.
In an optional embodiment, the generating of the shooting control mechanism includes: establishing an image data set containing image data and a shooting requirement label according to the at least one shooting requirement and the image data; performing classification calculation on the image data set, and performing image enhancement calculation on image data in the image data set; and taking the image data set of the classification calculation and the image enhancement calculation as a basic network architecture, and carrying out training of neural network architecture search calculation to obtain a shooting control mechanism aiming at the shooting requirement.
In an optional implementation, the program 410 may be specifically further configured to cause the processor 402 to perform the following operations: and searching the network architecture cascade obtained by the neural network architecture searching, calculating and training and the trained image aesthetic scoring network to obtain an optimized shooting control mechanism aiming at the shooting requirement.
According to the embodiment of the invention, at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects the shooting effect are provided, so that a shooting control mechanism which is generated by parameter search and architecture search according to the shooting requirement and the image data is obtained, and the shooting control mechanism is enabled, so that the shooting operation meeting the shooting requirement of the user is realized. Therefore, the embodiment of the invention can automatically obtain the shooting control mechanism according to the shooting requirement of the user and the image data which corresponds to the shooting requirement and reflects the shooting effect without specific shooting skill of the user or shooting skill learning, thereby completing the shooting operation aiming at the shooting requirement.
It should be noted that, according to the implementation requirement, each component/step described in the embodiment of the present invention may be divided into more components/steps, and two or more components/steps or partial operations of the components/steps may also be combined into a new component/step to achieve the purpose of the embodiment of the present invention.
The above-described method according to an embodiment of the present invention may be implemented in hardware, firmware, or as software or computer code storable in a recording medium such as a CD ROM, a RAM, a floppy disk, a hard disk, or a magneto-optical disk, or as computer code originally stored in a remote recording medium or a non-transitory machine-readable medium downloaded through a network and to be stored in a local recording medium, so that the method described herein may be stored in such software processing on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware such as an ASIC or FPGA. It will be appreciated that the computer, processor, microprocessor controller or programmable hardware includes memory components (e.g., RAM, ROM, flash memory, etc.) that can store or receive software or computer code that, when accessed and executed by the computer, processor or hardware, implements the method of downloading applications described herein. Further, when a general-purpose computer accesses code for implementing the methods illustrated herein, execution of the code transforms the general-purpose computer into a special-purpose computer for performing the methods illustrated herein.
Those of ordinary skill in the art will appreciate that the various illustrative elements and method steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present embodiments.
The above embodiments are only for illustrating the embodiments of the present invention and not for limiting the embodiments of the present invention, and those skilled in the art can make various changes and modifications without departing from the spirit and scope of the embodiments of the present invention, so that all equivalent technical solutions also belong to the scope of the embodiments of the present invention, and the scope of patent protection of the embodiments of the present invention should be defined by the claims.
Claims (8)
1. A method of automatically generating a shooting control mechanism, the method comprising:
providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect;
acquiring a shooting control mechanism generated by parameter search and architecture search according to the shooting requirement and the image data;
shooting enabling is carried out by utilizing the shooting control mechanism so as to realize shooting operation meeting the shooting requirements of the user;
wherein the generation of the shooting control mechanism comprises:
establishing an image data set containing image data and a shooting requirement label according to the at least one shooting requirement and the image data;
performing classification calculation on the image data set, and performing image enhancement calculation on image data in the image data set; wherein, a lightweight classification basic network architecture mobilenetV2 is selected to perform classification calculation on the image data set; selecting a DeepRetinexDecomposition for Low-light enhancement image enhancement algorithm, and performing image enhancement calculation on image data in the image data set;
taking the image data set of the classification calculation and the image enhancement calculation as a basic network architecture, and carrying out training of neural network architecture search calculation to obtain a shooting control mechanism aiming at the shooting requirement; adopting an image data set which is classified and calculated according to the lightweight classification basic network architecture mobilenetV2 and an image data set which is subjected to enhancement calculation by using the deepretinexDecomposition for low-light enhancement image enhancement algorithm, and training the neural network architecture search calculation for the neural network architecture search algorithm Path-LevelNetworkTransformationfor efficiency architecture search, so as to obtain a shooting control mechanism aiming at the shooting requirement;
and searching the network architecture cascade obtained by the neural network architecture searching, calculating and training and the trained image aesthetic scoring network to obtain an optimized shooting control mechanism aiming at the shooting requirement.
2. The method of claim 1, wherein the photographic requirements comprise:
at least one of the conditions of shooting scene, shooting light and object motion.
3. The method of claim 2, wherein the object comprises: at least one of a human, an animal, and an object.
4. An apparatus for automatically generating a shooting control mechanism, the apparatus comprising:
the data providing module is used for providing at least one shooting requirement of a user and image data which corresponds to the shooting requirement and reflects a shooting effect; the image data which corresponds to the shooting requirement and reflects the shooting effect comprises at least one frame of image which clearly reflects the shooting effect;
the mechanism generation module is used for obtaining a shooting control mechanism which carries out parameter search and architecture search generation according to the shooting requirement and the image data;
the shooting enabling module is used for utilizing the shooting control mechanism to carry out shooting enabling so as to realize shooting operation meeting the shooting requirements of the user;
wherein the generation of the shooting control mechanism comprises:
establishing an image data set containing image data and a shooting requirement label according to the at least one shooting requirement and the image data;
performing classification calculation on the image data set, and performing image enhancement calculation on image data in the image data set; wherein, a lightweight classification basic network architecture mobilenetV2 is selected to perform classification calculation on the image data set; selecting a DeepRetinexDecomposition for Low-light enhancement image enhancement algorithm, and performing image enhancement calculation on image data in the image data set;
taking the image data set of the classification calculation and the image enhancement calculation as a basic network architecture, and carrying out training of neural network architecture search calculation to obtain a shooting control mechanism aiming at the shooting requirement; adopting an image data set which is classified and calculated according to the lightweight classification basic network architecture mobilenetV2 and an image data set which is subjected to enhancement calculation by using the deepretinexDecomposition for low-light enhancement image enhancement algorithm, and training the neural network architecture search calculation for the neural network architecture search algorithm Path-LevelNetworkTransformationfor efficiency architecture search, so as to obtain a shooting control mechanism aiming at the shooting requirement;
and searching the network architecture cascade obtained by the neural network architecture searching, calculating and training and the trained image aesthetic scoring network to obtain an optimized shooting control mechanism aiming at the shooting requirement.
5. The apparatus of claim 4, wherein the photographic requirements comprise:
at least one of the conditions of shooting scene, shooting light and object motion.
6. The apparatus of claim 5, wherein the object comprises: at least one of a human, an animal, and an object.
7. An electronic device, comprising: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus;
the memory is used for storing at least one executable instruction which causes the processor to execute the corresponding operation of the method according to any one of claims 1-3.
8. A computer storage medium having stored thereon a computer program which, when executed by a processor, carries out the method of any one of claims 1 to 3.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910542321.7A CN110278370B (en) | 2019-06-21 | 2019-06-21 | Method and device for automatically generating shooting control mechanism and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910542321.7A CN110278370B (en) | 2019-06-21 | 2019-06-21 | Method and device for automatically generating shooting control mechanism and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110278370A CN110278370A (en) | 2019-09-24 |
| CN110278370B true CN110278370B (en) | 2020-12-18 |
Family
ID=67961437
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910542321.7A Active CN110278370B (en) | 2019-06-21 | 2019-06-21 | Method and device for automatically generating shooting control mechanism and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110278370B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111161155A (en) * | 2019-11-28 | 2020-05-15 | 东南大学 | Deep learning-based visibility enhancement method for underwater pier disease image |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107820020A (en) * | 2017-12-06 | 2018-03-20 | 广东欧珀移动通信有限公司 | Method of adjustment, device, storage medium and the mobile terminal of acquisition parameters |
| CN108174096A (en) * | 2017-12-29 | 2018-06-15 | 广东欧珀移动通信有限公司 | Method, apparatus, terminal and the storage medium of acquisition parameters setting |
| CN109376844A (en) * | 2018-10-30 | 2019-02-22 | 银河水滴科技（北京）有限公司 | The automatic training method of neural network and device recommended based on cloud platform and model |
| CN109615073B (en) * | 2018-12-03 | 2021-06-04 | 郑州云海信息技术有限公司 | Neural network model construction method, device and storage medium |
-
2019
- 2019-06-21 CN CN201910542321.7A patent/CN110278370B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110278370A (en) | 2019-09-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898579B (en) | Image definition recognition method and device and storage medium | |
| US10521705B2 (en) | Automatically selecting images using multicontext aware ratings | |
| US7623686B2 (en) | Techniques and apparatus for increasing accuracy of iris authentication by utilizing a plurality of iris images | |
| US8726324B2 (en) | Method for identifying image capture opportunities using a selected expert photo agent | |
| CN109951595A (en) | Intelligence adjusts method, apparatus, storage medium and the mobile terminal of screen intensity | |
| CN113538273B (en) | Image processing method and image processing apparatus | |
| CN107623819B (en) | A kind of method taken pictures and mobile terminal and related media production | |
| CN111147751B (en) | Photographing mode generation method and device and computer readable storage medium | |
| CN109660714A (en) | Image processing method, device, equipment and storage medium based on AR | |
| US11574477B2 (en) | Highlight video generated with adaptable multimodal customization | |
| KR20200110379A (en) | Method and system for reducing deep neural network structure | |
| CN110278370B (en) | Method and device for automatically generating shooting control mechanism and electronic equipment | |
| CN116612015A (en) | Model training method, image mole pattern removing method and device and electronic equipment | |
| CN106815803A (en) | The processing method and processing device of picture | |
| CN110570383B (en) | Image processing method and device, electronic equipment and storage medium | |
| WO2019128564A1 (en) | Focusing method, apparatus, storage medium, and electronic device | |
| KR20230078755A (en) | Object Re-Illumination Using Neural Networks | |
| CN113194256A (en) | Shooting method, shooting device, electronic equipment and storage medium | |
| CN116805282A (en) | Image super-resolution reconstruction method, model training method, device and electronic equipment | |
| CN112561822B (en) | Beautifying method and device, electronic equipment and storage medium | |
| CN115623313A (en) | Image processing method, image processing apparatus, electronic device, and storage medium | |
| CN114005156A (en) | Face replacement method, face replacement system, terminal equipment and computer storage medium | |
| CN113313790A (en) | Video generation method, device, equipment and storage medium | |
| CN113611318A (en) | Audio data enhancement method and related equipment | |
| CN112508801A (en) | Image processing method and computing device |
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 | ||
| GR01 | Patent grant |