WO2022156150A1 - 图像处理方法及装置、电子设备、存储介质及计算机程序 - Google Patents

图像处理方法及装置、电子设备、存储介质及计算机程序 Download PDF

Info

Publication number
WO2022156150A1
WO2022156150A1 PCT/CN2021/103195 CN2021103195W WO2022156150A1 WO 2022156150 A1 WO2022156150 A1 WO 2022156150A1 CN 2021103195 W CN2021103195 W CN 2021103195W WO 2022156150 A1 WO2022156150 A1 WO 2022156150A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
scene
processed
attribute information
information
Prior art date
Application number
PCT/CN2021/103195
Other languages
English (en)
French (fr)
Inventor
鲍虎军
王锐
李佰余
盛崇山
袁亚振
Original Assignee
浙江商汤科技开发有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 浙江商汤科技开发有限公司 filed Critical 浙江商汤科技开发有限公司
Publication of WO2022156150A1 publication Critical patent/WO2022156150A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T15/00—3D [Three Dimensional] image rendering
    • G06T15/005—General purpose rendering architectures

Definitions

  • the present disclosure relates to the field of computer technology, and in particular, to an image processing method and apparatus, an electronic device, a storage medium, and a computer program.
  • Reverse drawing is different from forward drawing. Reverse drawing starts from an image and aims to estimate the basic elements in the image scene. Reverse drawing is widely used in graphics and vision fields such as games and augmented reality (AR, Augmented Reality).
  • AR Augmented Reality
  • the present disclosure proposes a technical solution for image processing.
  • an image processing method including:
  • the scene attribute information is optimized to obtain optimized scene attribute information.
  • the scene attribute information is optimized according to the to-be-processed image and the rendered image to obtain optimized scene attribute information, including:
  • the scene attribute information is optimized, and rendering processing is performed according to the optimized scene attribute information, until the rendered image obtained by performing the rendering processing on the optimized scene attribute information is different from the to-be-processed image.
  • the difference of the images is less than or equal to the difference threshold.
  • the scene attribute information includes at least one item of scene normal information, reflectivity information, and light field information.
  • performing attribute analysis on the image to be processed to obtain scene attribute information of the image to be processed including:
  • the scene attribute estimation network includes a first estimation network, and the first estimation network includes a first feature extraction encoder, a first decoder, and a second decoder, and the The image to be processed is input to the scene attribute estimation network for attribute analysis, and the scene attribute information of the image to be processed is obtained, including:
  • the scene attribute estimation network includes a second estimation network, the second estimation network includes a second feature extraction encoder and a third decoder, and the to-be-processed image is input into the scene
  • the attribute estimation network performs attribute analysis to obtain scene attribute information of the to-be-processed image, including:
  • Attribute analysis is performed on the second image feature by the third decoder to obtain scene normal information of the to-be-processed image.
  • the method further includes:
  • the scene attribute estimation network is trained according to a preset training set, the training set includes a plurality of sample images, and the sample images have pre-labeled scene attribute information, wherein the sample images are based on the pre-labeled scene Attribute information is obtained by rendering operation.
  • the method further includes:
  • an image processing apparatus including:
  • an analysis part configured to perform attribute analysis on the image to be processed to obtain scene attribute information of the image to be processed
  • a first rendering part configured to perform rendering processing according to the scene attribute information to obtain a rendered image
  • the optimization part is configured to optimize the scene attribute information according to the to-be-processed image and the rendered image to obtain optimized scene attribute information.
  • the optimization part is further configured as:
  • the scene attribute information is optimized, and rendering processing is performed according to the optimized scene attribute information, until the rendered image obtained by performing the rendering processing on the optimized scene attribute information is different from the to-be-processed image.
  • the difference of the images is less than or equal to the difference threshold.
  • the scene attribute information includes at least one item of scene normal information, reflectivity information, and light field information.
  • the analysis part is further configured to:
  • the scene attribute estimation network includes a first estimation network, and the first estimation network includes a first feature extraction encoder, a first decoder, and a second decoder, and the analysis part, Also configured as:
  • the scene attribute estimation network includes a second estimation network, and the second estimation network includes a second feature extraction encoder and a third decoder, and the analysis part is further configured to:
  • Attribute analysis is performed on the second image feature by the third decoder to obtain scene normal information of the to-be-processed image.
  • the apparatus further includes:
  • the training part is configured to train the scene attribute estimation network according to a preset training set, the training set includes a plurality of sample images, and the sample images have pre-labeled scene attribute information, wherein the sample images are based on The pre-labeled scene attribute information is obtained by performing a rendering operation.
  • the apparatus further includes:
  • a determining part configured to determine at least one item of first scene attribute information from the optimized scene attribute information
  • the second rendering part is configured to perform rendering processing on the image to be processed according to the second scene attribute information and the first scene attribute information to obtain a scene rendering image.
  • an electronic device comprising: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to call the instructions stored in the memory to Perform the above method.
  • a computer-readable storage medium having computer program instructions stored thereon, the computer program instructions implementing the above method when executed by a processor.
  • a computer program including computer readable code, when the computer readable code is executed in a computer device, the processor in the computer device executes the method for implementing the above method.
  • This embodiment of the present disclosure can perform attribute analysis on the image to be processed, obtain scene attribute information of the image to be processed, and perform rendering processing according to the scene attribute information to obtain a rendered image, and then according to the image to be processed and the rendered image , and optimize the scene attribute information to obtain optimized scene attribute information.
  • the scene attribute information of the image to be processed can be optimized to improve the accuracy of the scene attribute information of the image to be processed.
  • FIG. 1A shows a schematic diagram of an image processing method according to an embodiment of the present disclosure
  • FIG. 1B shows a schematic diagram 1 of an application scenario of an electronic device according to an embodiment of the present disclosure
  • FIG. 1C shows a second schematic diagram of an application scenario of an electronic device according to an embodiment of the present disclosure
  • FIG. 2 shows a flowchart of an image processing method according to an embodiment of the present disclosure
  • FIG. 3 shows a schematic diagram of an image processing method according to an embodiment of the present disclosure
  • FIG. 4 shows a schematic diagram of an image processing method according to an embodiment of the present disclosure
  • FIG. 5A shows an image A to be processed according to an embodiment of the present disclosure
  • FIG. 5B shows an image B to be processed according to an embodiment of the present disclosure
  • FIG. 6A shows a reflectance map A corresponding to an image A to be processed according to an embodiment of the present disclosure
  • FIG. 6B shows a reflectance map B corresponding to an image B to be processed according to an embodiment of the present disclosure
  • FIG. 7A shows a scene normal map A corresponding to an image A to be processed according to an embodiment of the present disclosure
  • FIG. 7B shows a scene normal map B corresponding to an image B to be processed according to an embodiment of the present disclosure
  • FIG. 8A shows a light field diagram A corresponding to an image A to be processed according to an embodiment of the present disclosure
  • FIG. 8B shows a light field diagram B corresponding to an image B to be processed according to an embodiment of the present disclosure
  • FIG. 9A shows a scene rendering diagram A corresponding to an image A to be processed according to an embodiment of the present disclosure
  • FIG. 9B shows a scene rendering diagram B corresponding to an image B to be processed according to an embodiment of the present disclosure
  • FIG. 10A shows a relight map corresponding to an image A to be processed according to an embodiment of the present disclosure
  • FIG. 10B shows a texture replacement map corresponding to an image B to be processed according to an embodiment of the present disclosure
  • FIG. 11 shows a block diagram of an image processing apparatus according to an embodiment of the present disclosure
  • FIG. 12 shows a block diagram of an electronic device 800 according to an embodiment of the present disclosure
  • FIG. 13 shows a block diagram of an electronic device 1900 according to an embodiment of the present disclosure.
  • An embodiment of the present disclosure provides an image processing method.
  • the scene attribute information of the to-be-processed image can be obtained after attribute analysis of the to-be-processed image, and rendering processing is performed according to the scene attribute information to obtain the corresponding rendered image. Then, the scene attribute information of the to-be-processed image is optimized according to the difference between the rendered image and the to-be-processed image.
  • the optimized scene attribute information is rendered again, the difference between the obtained rendered image and the to-be-processed image is greater than the difference threshold, then Continue to optimize the optimized scene attribute information until the difference between the rendered image obtained by rendering the optimized scene attribute information and the to-be-processed image is less than or equal to the difference threshold, and it can be considered that the optimized scene attribute obtained at this time
  • the information is most precise for high fidelity and efficient reverse rendering.
  • the image processing method provided by the embodiments of the present disclosure may be executed by an electronic device such as a terminal device or a server, and the terminal device may be a user equipment (User Equipment, UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital processing (Personal Digital Assistant, PDA), handheld device, computing device, vehicle-mounted device, wearable device, etc.
  • the method can be implemented by the processor calling the computer-readable instructions stored in the memory.
  • the method may be performed by a server.
  • the following describes an application scenario in which the execution body of the image processing method is implemented as an electronic device.
  • the electronic device may include a processor 11 and a camera 12 .
  • the electronic device 11 can collect images to be processed through the camera 12 , and the processor 12
  • the to-be-processed image is analyzed and processed to determine the optimized scene attribute information in the to-be-processed image.
  • the electronic device may be implemented as a cell phone.
  • the electronic device 10 can receive images to be processed transmitted by other devices 13 through the network 14 , so that the electronic device 10 can The received image to be processed is analyzed and processed to determine the optimized scene attribute information in the image to be processed.
  • the electronic device can be implemented as a server, and the server can receive images to be processed collected by a smartphone through a network.
  • FIG. 2 shows a flowchart of an image processing method according to an embodiment of the present disclosure.
  • the image processing method may include:
  • step S21 attribute analysis is performed on the image to be processed to obtain scene attribute information of the image to be processed.
  • the embodiments of the present disclosure do not have strict restrictions on the format and content of the image to be processed, and the image to be processed may include common low dynamic image formats such as png and jpg.
  • the to-be-processed image may be an image captured by a common camera, for example, the to-be-processed image may be an image collected in the scene shown in FIG. 1B ; the to-be-processed image may also be a network picture, for example, the to-be-processed image
  • the processed image may be an image captured in the scene shown in Figure 1C.
  • the image to be processed in the embodiment of the present disclosure may be an indoor scene image or an outdoor scene image.
  • Attribute analysis can be performed on the image to be processed to obtain scene attribute information of the image to be processed.
  • a neural network for estimating the scene attribute information of the image to be processed can be pre-trained, and the attribute analysis of the image to be processed can be performed through the neural network to obtain the scene attribute information of the image to be processed.
  • the neural network can be a deep neural network. The present disclosure The embodiment does not limit the neural network.
  • the scene attribute information may include at least one item of scene normal information, reflectivity information, and light field information.
  • the scene normal information includes the normal value of the plane where each pixel point in the image to be processed is located, the reflectivity information includes the reflectance value corresponding to each pixel point in the image to be processed, and the light field information includes the sphere of each pixel point in the image to be processed. harmonic coefficient.
  • the scene normal information, reflectivity information, and light field information may be displayed in the form of a graph, matrix, vector, etc.
  • the embodiments of the present disclosure do not limit the display method of scene normal information, reflectivity information, and light field information .
  • scene normal information, reflectivity information, and light field information are shown by taking the display manner of the figure as an example.
  • the scene normal information is displayed as a scene normal map, and any pixel in the scene normal map is a unit-length vector, and the unit-length vector is the normal value of the plane where the pixel is located in the image to be processed.
  • the emissivity information is displayed as a reflectance map, and the pixel value of any pixel in the reflectance map corresponds to the reflectance value of the pixel in the image to be processed.
  • the light field information is shown as a light field map, and the light field map is represented as 9 spherical harmonic coefficients of each pixel in the image to be processed.
  • attribute analysis may be performed on the image to be processed to obtain scene normal information, reflectivity information, and light field information of the image to be processed.
  • step S22 rendering processing is performed according to the scene attribute information to obtain a rendered image.
  • the rendered image may be obtained by performing rendering processing on the scene attribute information of the obtained image to be processed by the renderer.
  • the renderer can implement the shading rendering process by a basic matrix multiplication operation, and the renderer acts as a pixel shader, and the shading for each pixel is calculated separately.
  • the image processor (Graphics Processing Unit, GPU) can be seamlessly used to accelerate the integration during the training process of the neural network for performing attribute analysis on the image to be processed.
  • the rendering process can be implemented by the following formula (1).
  • B is used to represent a rendering mode, and B is not limited in this embodiment of the present disclosure.
  • the hemisphere where it is located performs random direction sampling of the light field to obtain the light intensity information in the corresponding direction, and multiplies it by the dot product of the sampled light direction and the normal direction to obtain the color value and calculates the average color value.
  • step S23 the scene attribute information is optimized according to the to-be-processed image and the rendered image to obtain optimized scene attribute information.
  • the scene attribute information can be optimized according to the rendered image and the to-be-processed image, so that the optimized scene attribute information and the rendered image obtained after rendering are infinitely close to the to-be-processed image.
  • This embodiment of the present disclosure can perform attribute analysis on the image to be processed, obtain scene attribute information of the image to be processed, and perform rendering processing according to the scene attribute information to obtain a rendered image, and then according to the image to be processed and the rendered image , and optimize the scene attribute information to obtain optimized scene attribute information.
  • the accuracy of the scene attribute information of the image to be processed can be improved by optimizing the scene attribute information of the image to be processed.
  • the optimization of the scene attribute information according to the to-be-processed image and the rendered image to obtain optimized scene attribute information may include:
  • the scene attribute information is optimized, and rendering processing is performed according to the optimized scene attribute information, until the rendered image obtained by performing the rendering processing on the optimized scene attribute information is different from the to-be-processed image.
  • the difference of the images is less than or equal to the difference threshold.
  • the difference between the image to be processed and the rendered image can be determined.
  • the image to be processed and the rendered image can be loaded as arrays respectively, and the difference between the elements in the two arrays can be calculated one by one, and the image to be processed and the rendered image can be obtained by calculating the norm of the difference between the elements in the two arrays. difference between images.
  • the difference between the to-be-processed image and the rendered image can be obtained by calculating the distance between the feature vectors of the to-be-processed image and the rendered image.
  • the embodiments of the present disclosure do not limit the manner of determining the difference between the image to be processed and the rendered image, and any manner for determining the difference between the images may be used.
  • the difference threshold is a preset value, and the value can be determined according to requirements
  • the scene attribute information can be optimized, that is, the scene attribute information can be adjusted.
  • the process of optimizing the scene attribute information according to the difference between the rendered image and the to-be-processed image until the optimized scene attribute information is rendered The difference between the resulting rendered image and the image to be processed is less than or equal to the difference threshold.
  • At least one item of scene attribute information can be optimized.
  • the scene normal information, reflectivity information, and light field information of the image to be processed are obtained, and the scene normal information can be obtained through the scene normal information.
  • reflectivity information and light field information are rendered to obtain a corresponding rendered image.
  • the reflectivity information and light field information can be optimized according to the difference between the rendered image and the image to be processed, and the optimized reflectance information, the optimized light field information and the scene normal information can be rendered according to the optimized reflectivity information, the optimized light field information and the scene normal information.
  • Render the image and continue to optimize the reflectivity information and light field information according to the difference between the optimized rendered image and the to-be-processed image, until the difference between the optimized rendered image and the to-be-processed image is less than or equal to the difference threshold .
  • performing attribute analysis on the image to be processed to obtain scene attribute information of the image to be processed may include:
  • a scene attribute estimation network for performing attribute analysis on the image to be processed can be pre-trained, the image to be processed is used as the input of the scene attribute estimation network, and the output of the scene attribute estimation network is the scene attribute information of the image to be processed. .
  • the scene attribute estimation network may include a first estimation network, and the first estimation network may include a first feature extraction encoder, a first decoder, and a second decoder, and the The to-be-processed image is input into the scene attribute estimation network to perform attribute analysis, and the scene attribute information of the to-be-processed image is obtained, which may include:
  • the first estimation network is used to perform attribute analysis on the image to be processed to obtain reflectivity information and light field information of the image to be processed, which may adopt a U-net structure.
  • the first estimation network may include a first feature extraction encoder, a first decoder, and a second decoder, wherein the first decoder can obtain the to-be-processed image by performing attribute analysis on image features of the to-be-processed image
  • the second decoder can obtain the light field information of the image to be processed by performing attribute analysis on the image features of the image to be processed.
  • the first decoder and the second decoder in the first estimation network share a first feature extraction encoder, which is treated by the first feature extraction encoder. After processing the image for feature extraction, after obtaining the first image feature, the first decoder and the second decoder respectively perform attribute analysis on the first image feature to obtain reflectivity information and light field information of the to-be-processed image.
  • the scene attribute estimation network may include a second estimation network, the second estimation network includes a second feature extraction encoder and a third decoder, the Inputting a scene attribute estimation network to perform attribute analysis to obtain scene attribute information of the image to be processed, which may include:
  • Attribute analysis is performed on the second image feature by the third decoder to obtain scene normal information of the network to be processed.
  • the second estimation network is used to perform attribute analysis on the image to be processed to obtain scene normal information of the image to be processed.
  • Global information is crucial for estimating scene normal maps, therefore, the second estimation network uses a U-net structure similar to the first estimation network alone, and the network weights are not shared with the first estimation network.
  • the second estimation network may include a second feature extraction encoder and a third decoder.
  • the second feature extraction encoder performs feature extraction on the image to be processed, and after obtaining the second image feature, the third decoding is performed.
  • the device performs attribute analysis on the second image feature to obtain scene normal information of the image to be processed.
  • the method may further include: training the scene attribute estimation network according to a preset training set, where the training set includes a plurality of sample images, and the sample images have pre-labeled scene attributes information, wherein the sample image is obtained by performing a rendering operation according to the pre-labeled scene attribute information.
  • a large amount of sample data can be generated in advance according to preset scene attribute information. For example, preset light field information, preset light field information, preset The scene normal information and preset reflectivity information are obtained, and then a sample image is generated through rendering processing according to the preset scene attribute information.
  • the scene attribute information used to generate the sample image is used as the annotation information of the sample image.
  • a training set for training the scene attribute estimation network can be constructed.
  • the scene attribute estimation network may include a first estimation network and a second estimation network, the reflectance and the illumination coefficient are closely related, and are highly dependent on the local information of the image, and the global information is important for estimating the scene method.
  • Line information is very important, so the first decoder and the second decoder in the first estimation network share a first feature extraction encoder, and the second estimation network uses a second feature extraction encoder alone.
  • the first estimation network may include a first feature extraction encoder, a first decoder, and a second decoder
  • the second estimation network may include a second feature extraction encoder and a third decoder.
  • the encoder includes six residual blocks, each residual block may have three convolution layers, wherein the convolution step of the first layer is The length is 2, and the convolution stride of the remaining two layers is 1.
  • the first estimation network and the second estimation network use the ReLU activation function uniformly.
  • the decoders (including the first decoder, the second decoder and the third decoder) may use bilinear interpolation and forward convolution. Each decoder has 12 convolutional layers with the same number of feature maps as the encoder, in reverse order. At each decoder block, skip connections are implemented to transmit information through concatenation.
  • the sample images are respectively input into the first feature extraction encoder and the second feature extraction encoder.
  • the first feature extraction encoder performs feature extraction on the sample image
  • the second feature extraction encoder performs feature extraction on the sample image.
  • a second image feature is obtained.
  • the first image features are respectively input into the first decoder and the second decoder
  • the first decoder performs attribute analysis on the first image features
  • the reflectivity information of the sample image can be obtained
  • the second decoder analyzes the first image features.
  • attribute analysis the light field information of the sample image can be obtained.
  • scene normal information of the sample image can be obtained.
  • the network loss of the scene attribute estimation network can be obtained.
  • the network loss does not meet the training accuracy requirements (for example, the network loss is greater than the loss threshold)
  • the network parameters of the scene attribute estimation network can be adjusted through the network loss until the scene attribute estimation network has The network loss meets the training accuracy requirements (for example: the network loss is less than the loss threshold), the training is completed, and the trained scene attribute estimation network is obtained.
  • any method for calculating the network loss may be used, for example, a 1-normal form, a 2-normal form, and the like.
  • the scene attribute information of the image to be processed can be estimated through the scene attribute estimation network, including the reflectivity information, the scene normal information and the light field information.
  • the reconstruction effect is also more accurate, and the obtained spatially varying light field information and scene normal information can be used for heavy lighting, material replacement, etc., providing possibilities for many applications of augmented reality.
  • the image to be processed is input to the scene attribute estimation network.
  • the first feature extraction encoder can perform feature extraction on the image to be processed to obtain first image features
  • the second feature extraction encoder can perform feature extraction on the to-be-processed image to obtain second image features.
  • the first decoder performs attribute estimation on the first image feature to obtain reflectivity information of the to-be-processed image
  • the second decoder performs attribute estimation on the first image feature to obtain the light field information of the to-be-processed image.
  • the third decoder performs attribute estimation on the second image feature, and can obtain the scene normal information of the image to be processed (in FIG. 3, the reflectivity information is displayed in the form of a reflectivity map, the light field information is displayed in the form of a light field map, Normal information is displayed in the form of scene normal maps).
  • the rendered image can be obtained by rendering through the reflectivity information, light field information and scene normal information of the image to be processed. Further, the reflectivity information and the light field information of the to-be-processed image can be optimized through the difference between the rendered image and the to-be-processed image, and the optimized reflectivity information and the optimized light-field information can be obtained.
  • the method may further include:
  • At least one item of the optimized scene attribute information can be changed or replaced, so as to realize the re-illumination or texture replacement of the image to be processed, etc. operate.
  • operations such as relighting or texture replacement may be performed on other images according to at least one item of optimized scene attribute information of the image to be processed.
  • the first scene attribute information may be at least one item of the optimized scene attribute information
  • the second scene attribute information is used to compare the third scene attribute information other than the first scene attribute information in the optimized scene attribute information.
  • the scene attribute information to be replaced by the scene attribute information, the second scene attribute information may be preset information, or may be information obtained by adjusting and changing the third scene attribute information.
  • the electronic device can implement AR applications such as heavy lighting operation and texture replacement through the following steps.
  • the electronic device may collect a single indoor scene image through an image collection device.
  • the electronic device may collect an indoor scene image at noon to obtain the to-be-processed image A shown in FIG. 5A and the to-be-processed image B shown in FIG. 5B .
  • the electronic device can use the scene attribute estimation network to process the to-be-processed image A and the to-be-processed image B, respectively.
  • the scene normal information, light field information and reflectivity information of the to-be-processed image A and the to-be-processed image B are obtained.
  • the reflectance information of the image A to be processed may be displayed in the form of the reflectance map A shown in FIG. 6A .
  • the scene normal information of the image A to be processed can be displayed in the form of the scene normal map A shown in FIG. 7A .
  • the light field information of the image A to be processed can be displayed in the form of the light field map A shown in FIG. 8A .
  • the reflectance information of the image B to be processed can be displayed in the form of the reflectance map B shown in FIG. 6B .
  • the scene normal information of the image B to be processed can be displayed in the form of the scene normal map A shown in FIG. 7B .
  • the light field information of the image B to be processed can be displayed in the form of the light field map B shown in FIG. 8B .
  • the optimized scene attribute information of the image A to be processed may be displayed in the form of the scene rendering diagram A shown in FIG. 9A .
  • the optimized scene attribute information of the to-be-processed image B can be displayed by the scene rendering diagram B shown in FIG. 9B .
  • the scene normal information and reflectivity information of the image A to be processed may be determined as the first scene attribute information.
  • the light field information in the night scene (which can be obtained by adjusting the light field information of the image A to be processed, or directly performing attribute analysis on the image captured in the night scene to obtain the light field information) can be used as the second scene attribute information, Rendering the scene normal information and reflectivity information of the image A to be processed and the light field information in the nighttime scene, the reillumination map of the image A to be processed after relighting can be obtained, and the image of the scene at night can be obtained by simulating.
  • the re-illumination map can be referred to as shown in FIG. 10A .
  • the scene normal information and light field information of the to-be-processed image B can be determined as the first scene attribute information, and the reflectivity information in the cloudy scene (which can be obtained by adjusting the reflectivity information of the to-be-processed image B, or Directly perform attribute analysis on the image captured in the cloudy scene, and obtain the emissivity information) as the second scene attribute information, and render the scene normal information and light field information of the image B to be processed and the reflectivity information in the cloudy scene.
  • the texture replacement map can be referred to as shown in FIG. 10B .
  • the embodiment of the present disclosure can estimate and obtain the inherent scene attribute information of the image to be processed in the real scene, including reflectivity information, scene normal information and light field information.
  • the user can give different scene attribute information to obtain the scene rendering image corresponding to the real scene, so as to realize re-lighting or texture replacement.
  • the virtual objects may be a virtual game character or furniture during decoration, so as to achieve the effect of virtual and real fusion.
  • the embodiment of the present disclosure can obtain the inherent scene attribute information of the image to be processed, including reflectivity information, scene normal information and light field information, and the decomposition of the image to be processed is more thorough, thereby making the reconstruction effect based on the scene attribute information more accurate.
  • the obtained scene attribute information can be used for heavy lighting, material replacement, etc., providing possibilities for many applications of augmented reality.
  • the present disclosure also provides image processing apparatuses, electronic devices, computer-readable storage media, and programs, all of which can be used to implement any image processing method provided by the present disclosure.
  • image processing apparatuses electronic devices, computer-readable storage media, and programs, all of which can be used to implement any image processing method provided by the present disclosure.
  • FIG. 11 shows a block diagram of an image processing apparatus according to an embodiment of the present disclosure. As shown in FIG. 11 , the apparatus includes:
  • the analysis part 61 can be configured to perform attribute analysis on the image to be processed to obtain scene attribute information of the image to be processed;
  • the first rendering part 62 can be configured to perform rendering processing according to the scene attribute information to obtain a rendered image
  • the optimization part 63 may be configured to optimize the scene attribute information according to the to-be-processed image and the rendered image to obtain optimized scene attribute information.
  • This embodiment of the present disclosure can perform attribute analysis on the image to be processed, obtain scene attribute information of the image to be processed, and perform rendering processing according to the scene attribute information to obtain a rendered image, and then according to the image to be processed and the rendered image , and optimize the scene attribute information to obtain optimized scene attribute information.
  • the accuracy of the scene attribute information of the image to be processed can be improved by optimizing the scene attribute information of the image to be processed.
  • the optimization part 63 may also be configured as:
  • the scene attribute information is optimized, and rendering processing is performed according to the optimized scene attribute information, until the rendered image obtained by performing the rendering processing on the optimized scene attribute information is different from the to-be-processed image.
  • the difference of the images is less than or equal to the difference threshold.
  • the scene attribute information may include at least one item of scene normal information, reflectivity information, and light field information.
  • the analysis part 61 may also be configured as:
  • the scene attribute estimation network includes a first estimation network, and the first estimation network includes a first feature extraction encoder, a first decoder, and a second decoder, and the analysis part 61 , which can also be configured as:
  • the scene attribute estimation network includes a second estimation network
  • the second estimation network includes a second feature extraction encoder and a third decoder
  • the analysis part 61 may also be configured for:
  • Attribute analysis is performed on the second image feature by the third decoder to obtain scene normal information of the to-be-processed image.
  • the apparatus may further include:
  • the training part can be configured to train the scene attribute estimation network according to a preset training set, the training set includes a plurality of sample images, and the sample images have pre-labeled scene attribute information, wherein the sample images are It is obtained by performing a rendering operation according to the pre-marked scene attribute information.
  • the apparatus may further include:
  • the determining part may be configured to determine at least one item of first scene attribute information from the optimized scene attribute information
  • the second rendering part may be configured to perform rendering processing on the to-be-processed image according to the second scene attribute information and the first scene attribute information to obtain a scene rendering image.
  • the functions or modules included in the apparatus provided in the embodiments of the present disclosure may be used to execute the methods described in the above method embodiments, and the implementation of the above method embodiments may refer to the descriptions of the above method embodiments. Repeat.
  • Embodiments of the present disclosure further provide a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the foregoing method is implemented.
  • the computer-readable storage medium may be a non-volatile computer-readable storage medium.
  • An embodiment of the present disclosure further provides an electronic device, including: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to invoke the instructions stored in the memory to execute the above method.
  • Embodiments of the present disclosure also provide a computer program product, including computer-readable codes.
  • a processor in the device executes the image processing method for implementing the image processing method provided by any of the above embodiments. instruction.
  • Embodiments of the present disclosure further provide another computer program product for storing computer-readable instructions, which, when executed, cause the computer to perform the operations of the image processing method provided by any of the foregoing embodiments.
  • the electronic device may be provided as a terminal, server or other form of device.
  • FIG. 12 shows a block diagram of an electronic device 800 according to an embodiment of the present disclosure.
  • electronic device 800 may be a mobile phone, computer, digital broadcast terminal, messaging device, game console, tablet device, medical device, fitness device, personal digital assistant, etc. terminal.
  • an electronic device 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input/output (I/O) interface 812, a sensor component 814 , and the communication component 816 .
  • the processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations.
  • the processing component 802 can include one or more processors 820 to execute instructions to perform all or some of the steps of the methods described above.
  • processing component 802 may include one or more modules that facilitate interaction between processing component 802 and other components.
  • processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.
  • Memory 804 is configured to store various types of data to support operation at electronic device 800 . Examples of such data include instructions for any application or method operating on electronic device 800, contact data, phonebook data, messages, pictures, videos, and the like. Memory 804 may be implemented by any type of volatile or nonvolatile storage device or combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.
  • SRAM static random access memory
  • EEPROM electrically erasable programmable read only memory
  • EPROM erasable Programmable Read Only Memory
  • PROM Programmable Read Only Memory
  • ROM Read Only Memory
  • Magnetic Memory Flash Memory
  • Magnetic or Optical Disk Magnetic Disk
  • Power supply assembly 806 provides power to various components of electronic device 800 .
  • Power supply components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 800 .
  • Multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user.
  • the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user.
  • the touch panel includes one or more touch sensors to sense touch, swipe, and gestures on the touch panel. The touch sensor may not only sense the boundaries of a touch or swipe action, but also detect the duration and pressure associated with the touch or swipe action.
  • the multimedia component 808 includes a front-facing camera and/or a rear-facing camera. When the electronic device 800 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each of the front and rear cameras can be a fixed optical lens system or have focal length and optical zoom capability.
  • Audio component 810 is configured to output and/or input audio signals.
  • audio component 810 includes a microphone (MIC) that is configured to receive external audio signals when electronic device 800 is in operating modes, such as calling mode, recording mode, and voice recognition mode.
  • the received audio signal may be further stored in memory 804 or transmitted via communication component 816 .
  • audio component 810 also includes a speaker for outputting audio signals.
  • the I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which may be a keyboard, a click wheel, a button, or the like. These buttons may include, but are not limited to: home button, volume buttons, start button, and lock button.
  • Sensor assembly 814 includes one or more sensors for providing status assessment of various aspects of electronic device 800 .
  • the sensor assembly 814 can detect the on/off state of the electronic device 800, the relative positioning of the components, such as the display and the keypad of the electronic device 800, the sensor assembly 814 can also detect the electronic device 800 or one of the electronic device 800 Changes in the position of components, presence or absence of user contact with the electronic device 800 , orientation or acceleration/deceleration of the electronic device 800 and changes in the temperature of the electronic device 800 .
  • Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact.
  • Sensor assembly 814 may also include a light sensor, such as a complementary metal oxide semiconductor (CMOS) or charge coupled device (CCD) image sensor, for use in imaging applications.
  • CMOS complementary metal oxide semiconductor
  • CCD charge coupled device
  • the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.
  • Communication component 816 is configured to facilitate wired or wireless communication between electronic device 800 and other devices.
  • the electronic device 800 may access a wireless network based on a communication standard, such as wireless network (WiFi), second generation mobile communication technology (2G) or third generation mobile communication technology (3G), or a combination thereof.
  • the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel.
  • the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication.
  • the NFC module may be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.
  • RFID radio frequency identification
  • IrDA infrared data association
  • UWB ultra-wideband
  • Bluetooth Bluetooth
  • electronic device 800 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A programmed gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation is used to perform the above method.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs digital signal processing devices
  • PLDs programmable logic devices
  • FPGA field programmable A programmed gate array
  • controller microcontroller, microprocessor or other electronic component implementation is used to perform the above method.
  • a non-volatile computer-readable storage medium such as a memory 804 comprising computer program instructions executable by the processor 820 of the electronic device 800 to perform the above method is also provided.
  • FIG. 13 shows a block diagram of an electronic device 1900 according to an embodiment of the present disclosure.
  • the electronic device 1900 may be provided as a server.
  • electronic device 1900 includes a processing component 1922, which further includes one or more processors, and a memory resource, represented by memory 1932, for storing instructions executable by processing component 1922, such as applications.
  • An application program stored in memory 1932 may include one or more modules, each corresponding to a set of instructions.
  • the processing component 1922 is configured to execute instructions to perform the above-described methods.
  • the electronic device 1900 may also include a power supply assembly 1926 configured to perform power management of the electronic device 1900, a wired or wireless network interface 1950 configured to connect the electronic device 1900 to a network, and an input output (I/O) interface 1958 .
  • the electronic device 1900 can operate based on an operating system stored in the memory 1932, such as a Microsoft server operating system (Windows ServerTM), a graphical user interface based operating system (Mac OS XTM) introduced by Apple, a multi-user multi-process computer operating system (UnixTM). ), Free and Open Source Unix-like Operating System (LinuxTM), Open Source Unix-like Operating System (FreeBSDTM) or similar.
  • a non-volatile computer-readable storage medium such as memory 1932 comprising computer program instructions executable by processing component 1922 of electronic device 1900 to perform the above-described method.
  • a computer program comprising computer readable code, when the computer readable code is executed in an electronic device, a processor in the electronic device executes the image processing configured to implement the above-mentioned image processing The steps of a method embodiment.
  • the present disclosure may be a system, method and/or computer program product.
  • the computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of the present disclosure.
  • a computer-readable storage medium may be a tangible device that can hold and store instructions for use by the instruction execution device.
  • the computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing.
  • Non-exhaustive list of computer readable storage media include: portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM) or flash memory), static random access memory (SRAM), portable compact disk read only memory (CD-ROM), digital versatile disk (DVD), memory sticks, floppy disks, mechanically coded devices, such as printers with instructions stored thereon Hole cards or raised structures in grooves, and any suitable combination of the above.
  • RAM random access memory
  • ROM read only memory
  • EPROM erasable programmable read only memory
  • flash memory static random access memory
  • SRAM static random access memory
  • CD-ROM compact disk read only memory
  • DVD digital versatile disk
  • memory sticks floppy disks
  • mechanically coded devices such as printers with instructions stored thereon Hole cards or raised structures in grooves, and any suitable combination of the above.
  • Computer-readable storage media are not to be construed as transient signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (eg, light pulses through fiber optic cables), or through electrical wires transmitted electrical signals.
  • the computer readable program instructions described herein may be downloaded to various computing/processing devices from a computer readable storage medium, or to an external computer or external storage device over a network such as the Internet, a local area network, a wide area network, and/or a wireless network.
  • the network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers.
  • a network adapter card or network interface in each computing/processing device receives computer-readable program instructions from a network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing/processing device .
  • Computer program instructions for carrying out operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or instructions in one or more programming languages.
  • Source or object code written in any combination, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as the "C" language or similar programming languages.
  • the computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server implement.
  • the remote computer may be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (eg, using an Internet service provider through the Internet connect).
  • LAN local area network
  • WAN wide area network
  • custom electronic circuits such as programmable logic circuits, field programmable gate arrays (FPGAs), or programmable logic arrays (PLAs) can be personalized by utilizing state information of computer readable program instructions.
  • Computer readable program instructions are executed to implement various aspects of the present disclosure.
  • These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer or other programmable data processing apparatus to produce a machine that causes the instructions when executed by the processor of the computer or other programmable data processing apparatus , resulting in means for implementing the functions/acts specified in one or more blocks of the flowchart and/or block diagrams.
  • These computer readable program instructions can also be stored in a computer readable storage medium, these instructions cause a computer, programmable data processing apparatus and/or other equipment to operate in a specific manner, so that the computer readable medium on which the instructions are stored includes An article of manufacture comprising instructions for implementing various aspects of the functions/acts specified in one or more blocks of the flowchart and/or block diagrams.
  • Computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other equipment to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other equipment to produce a computer-implemented process , thereby causing instructions executing on a computer, other programmable data processing apparatus, or other device to implement the functions/acts specified in one or more blocks of the flowcharts and/or block diagrams.
  • each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more functions for implementing the specified logical function(s) executable instructions.
  • the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved.
  • each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations can be implemented in dedicated hardware-based systems that perform the specified functions or actions , or can be implemented in a combination of dedicated hardware and computer instructions.
  • the computer program product can be implemented in hardware, software or a combination thereof.
  • the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK) and the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Graphics (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Generation (AREA)

Abstract

本公开涉及一种图像处理方法及装置、电子设备和存储介质，所述方法包括：对待处理图像进行属性分析，得到所述待处理图像的场景属性信息；根据所述场景属性信息进行渲染处理，得到渲染图像；根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。本公开实施例可提高待处理图像的场景属性信息的精度。

Description

图像处理方法及装置、电子设备、存储介质及计算机程序
相关申请的交叉引用
本公开要求在2021年01月19日提交中国专利局、申请号为202110070584.X、申请名称为“图像处理方法及装置、电子设备和存储介质”的中国专利申请的优先权，该中国专利申请的全部内容在此以全文引入的方式引入本公开。
技术领域
本公开涉及计算机技术领域，尤其涉及一种图像处理方法及装置、电子设备、存储介质及计算机程序。
背景技术
逆向绘制与正向绘制不同，逆向绘制从图像出发，旨在估计图像场景中的基本元素，在游戏和增强现实(AR，Augmented Reality)等图形学和视觉领域普遍存在逆向绘制应用。
发明内容
本公开提出了一种用于进行图像处理的技术方案。
根据本公开实施例的一方面，提供了一种图像处理方法，包括：
对待处理图像进行属性分析，得到所述待处理图像的场景属性信息；
根据所述场景属性信息进行渲染处理，得到渲染图像；
根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。
在一种可能的实现方式中，所述根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息，包括：
确定所述待处理图像与所述渲染图像的差异；
在所述差异大于差异阈值时，对所述场景属性信息进行优化，并根据优化后的场景属性信息进行渲染处理，直至优化后的场景属性信息进行渲染处理得到的渲染图像，与所述待处理图像的差异小于或等于差异阈值。
在一种可能的实现方式中，所述场景属性信息包括场景法线信息、反射率信息、及光场信息中的至少一项。
在一种可能的实现方式中，所述对待处理图像进行属性分析，得到所述待处理图像的场景属性信息，包括：
将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息。
在一种可能的实现方式中，所述场景属性估计网络包括第一估计网络，所述第一估计网络包括第一特征提取编码器、第一解码器及第二解码器，所述将所述待处理图像输 入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息，包括：
将所述待处理图像输入所述第一特征编码器中进行特征提取，得到第一图像特征；
通过所述第一解码器对所述第一图像特征进行属性分析，得到所述待处理图像的反射率信息，及通过所述第二解码器对所述第一图像特征进行属性分析，得到所述待处理图像的光场信息。
在一种可能的实现方式中，所述场景属性估计网络包括第二估计网络，所述第二估计网络包括第二特征提取编码器及第三解码器，所述将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息，包括：
将所述待处理图像输入所述第二特征编码器中进行特征提取，得到第二图像特征；
通过所述第三解码器对所述第二图像特征进行属性分析，得到所述待处理图像的场景法线信息。
在一种可能的实现方式中，所述方法还包括：
根据预设的训练集训练所述场景属性估计网络，所述训练集中包括多个样本图像，所述样本图像具有预标注的场景属性信息，其中，所述样本图像为根据所述预标注的场景属性信息进行渲染操作得到的。
在一种可能的实现方式中，所述方法还包括：
从优化后的场景属性信息中确定至少一项第一场景属性信息；
根据第二场景属性信息及所述第一场景属性信息，对所述待处理图像进行渲染处理，得到场景渲染图像。
根据本公开实施例的一方面，提供了一种图像处理装置，包括：
分析部分，被配置为对待处理图像进行属性分析，得到所述待处理图像的场景属性信息；
第一渲染部分，被配置为根据所述场景属性信息进行渲染处理，得到渲染图像；
优化部分，被配置为根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。
在一种可能的实现方式中，所述优化部分，还被配置为：
确定所述待处理图像与所述渲染图像的差异；
在所述差异大于差异阈值时，对所述场景属性信息进行优化，并根据优化后的场景属性信息进行渲染处理，直至优化后的场景属性信息进行渲染处理得到的渲染图像，与所述待处理图像的差异小于或等于差异阈值。
在一种可能的实现方式中，所述场景属性信息包括场景法线信息、反射率信息、及光场信息中的至少一项。
在一种可能的实现方式中，所述分析部分，还被配置为：
将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息。
在一种可能的实现方式中，所述场景属性估计网络包括第一估计网络，所述第一估 计网络包括第一特征提取编码器、第一解码器及第二解码器，所述分析部分，还被配置为：
将所述待处理图像输入所述第一特征编码器中进行特征提取，得到第一图像特征；
通过所述第一解码器对所述第一图像特征进行属性分析，得到所述待处理图像的反射率信息，及通过所述第二解码器对所述第一图像特征进行属性分析，得到所述待处理图像的光场信息。
在一种可能的实现方式中，所述场景属性估计网络包括第二估计网络，所述第二估计网络包括第二特征提取编码器及第三解码器，所述分析部分，还被配置为：
将所述待处理图像输入所述第二特征编码器中进行特征提取，得到第二图像特征；
通过所述第三解码器对所述第二图像特征进行属性分析，得到所述待处理图像的场景法线信息。
在一种可能的实现方式中，所述装置还包括：
训练部分，被配置为根据预设的训练集训练所述场景属性估计网络，所述训练集中包括多个样本图像，所述样本图像具有预标注的场景属性信息，其中，所述样本图像为根据所述预标注的场景属性信息进行渲染操作得到的。
在一种可能的实现方式中，所述装置还包括：
确定部分，被配置为从优化后的场景属性信息中确定至少一项第一场景属性信息；
第二渲染部分，被配置为根据第二场景属性信息及所述第一场景属性信息，对待处理图像进行渲染处理，得到场景渲染图像。
根据本公开实施例的一方面，提供了一种电子设备，包括：处理器；用于存储处理器可执行指令的存储器；其中，所述处理器被配置为调用所述存储器存储的指令，以执行上述方法。
根据本公开实施例的一方面，提供了一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述方法。
根据本公开实施例的一方面，提供了一种计算机程序，包括计算机可读代码，当计算机可读代码在计算机设备中运行时，计算机设备中的处理器执行用于实现上述方法。
本公开实施例可以对待处理图像进行属性分析，得到所述待处理图像的场景属性信息，并根据所述场景属性信息进行渲染处理，得到渲染图像，进而根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。根据本公开实施例提供的图像处理方法及装置、电子设备和存储介质，可以通过对待处理图像的场景属性信息进行优化，提高待处理图像的场景属性信息的精度。
应当理解的是，以上的一般描述和后文的细节描述仅是示例性和解释性的，而非限制本公开。根据下面参考附图对示例性实施例的详细说明，本公开的其它特征及方面将变得清楚。
附图说明
此处的附图被并入说明书中并构成本说明书的一部分，这些附图示出了符合本公开的实施例，并与说明书一起用于说明本公开的技术方案。
图1A示出根据本公开实施例的图像处理方法的示意图；
图1B示出根据本公开实施例的电子设备应用场景示意图一；
图1C示出根据本公开实施例的电子设备应用场景示意图二；
图2示出根据本公开实施例的图像处理方法的流程图；
图3示出根据本公开实施例的图像处理方法的示意图；
图4示出根据本公开实施例的图像处理方法的示意图；
图5A示出根据本公开实施例的待处理图像A；
图5B示出根据本公开实施例的待处理图像B；
图6A示出根据本公开实施例的待处理图像A对应的反射率图A；
图6B示出根据本公开实施例的待处理图像B对应的反射率图B；
图7A示出根据本公开实施例的待处理图像A对应的场景法线贴图A；
图7B示出根据本公开实施例的待处理图像B对应的场景法线贴图B；
图8A示出根据本公开实施例的待处理图像A对应的光场图A；
图8B示出根据本公开实施例的待处理图像B对应的光场图B；
图9A示出根据本公开实施例的待处理图像A对应的场景渲染图A；
图9B示出根据本公开实施例的待处理图像B对应的场景渲染图B；
图10A示出根据本公开实施例的待处理图像A对应的重光照图；
图10B示出根据本公开实施例的待处理图像B对应的纹理替换图；
图11示出根据本公开实施例的图像处理装置的框图；
图12示出根据本公开实施例的一种电子设备800的框图；
图13示出根据本公开实施例的一种电子设备1900的框图。
具体实施方式
以下将参考附图详细说明本公开的各种示例性实施例、特征和方面。附图中相同的附图标记表示功能相同或相似的元件。尽管在附图中示出了实施例的各种方面，但是除非特别指出，不必按比例绘制附图。
在这里专用的词“示例性”意为“用作例子、实施例或说明性”。这里作为“示例性”所说明的任何实施例不必解释为优于或好于其它实施例。
本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中术语“至少一种”表示多种中的任意一种或多种中的至少两种的任意组合，例如，包括A、B、C中的至少一种，可以表示包括从A、B和C构成的集合中选择的任意一个或多个元素。
另外，为了更好地说明本公开，在下文的实施方式中给出了众多的细节。本领域技术人员应当理解，没有某些细节，本公开同样可以实施。在一些实例中，对于本领域技术人员熟知的方法、手段、元件和电路未作详细描述，以便于凸显本公开的主旨。
在游戏和AR等图形学和视觉处理领域普遍存在逆向绘制应用。特别的是，在复杂场景下，由于图像场景中的基本元素(例如自然场景的光照，材质反射率和场景几何)之间的高度耦合，电子设备拆解图像中的上述基本元素时会存在明显的歧义，使得实现高保真和高效的逆向绘制以及重建存在较大的挑战性。
本公开实施例提供了一种图像处理方法，参照图1A，可以对待处理图像进行属性分析后，得到待处理图像的场景属性信息，并根据场景属性信息进行渲染处理，得到对应的渲染图像。进而根据渲染图像与待处理图像之间的差异对待处理图像的场景属性信息进行优化，若优化后的场景属性信息再次渲染后，得到的渲染图像与待处理图像之间的差异大于差异阈值，则继续对优化后的场景属性信息进行优化，直至优化后的场景属性信息经过渲染得到的渲染图像，与待处理图像之间的差异小于或等于差异阈值，可以认为此时得到的优化后的场景属性信息是最精准的，以实现高保真和高效的逆向绘制。
本公开实施例提供的图像处理方法可以由终端设备或服务器等电子设备执行，终端设备可以为用户设备(User Equipment，UE)、移动设备、用户终端、终端、蜂窝电话、无绳电话、个人数字处理(Personal Digital Assistant，PDA)、手持设备、计算设备、车载设备、可穿戴设备等，所述方法可以通过处理器调用存储器中存储的计算机可读指令的方式来实现。或者，可通过服务器执行所述方法。
下面介绍图像处理方法的执行主体实施为电子设备的应用场景。
在一种可能的实现方式中，参考图1B所示电子设备应用场景示意图一，电子设备可以包括处理器11和相机12，这样，电子设备11可以通过相机12采集待处理图像，通过处理器12对待处理图像进行分析处理，以确定该待处理图像中优化后的场景属性信息。例如，电子设备可以实施为手机。
在另一种可能的实现方式中，参考图1C所示的一种电子设备应用场景示意图二，电子设备10可以通过接收其他设备13通过网络14传送的待处理图像，这样，电子设备10可以对接收到的待处理图像进行分析处理，从而确定该待处理图像中优化后的场景属性信息。例如，电子设备可以实施为服务器，服务器可以通过网络接收智能手机采集的待处理图像。
图2示出根据本公开实施例的图像处理方法的流程图，如图2所示，所述图像处理方法可以包括：
在步骤S21中，对待处理图像进行属性分析，得到所述待处理图像的场景属性信息。
举例来说，本公开实施例对于待处理图像的格式以及包含内容并没有严格的约束，待处理图像可以包括常见的png，jpg等低动态图像格式。另外，本公开实施例中，待处理可以是通过普通相机拍摄的图像，例如，待处理图像可以是图1B所示的场景中采集到的图像；待处理图像还可以是网络图片，例如，待处理图像可以是图1C所示的场景 中采集到的图像。
举例来说，本公开实施例中的待处理图像可以是室内场景图像，也可以是室外场景图像。
可以对待处理图像进行属性分析，以得到待处理图像的场景属性信息。例如：可以预训练用于估计待处理图像的场景属性信息的神经网络，通过该神经网络对待处理图像进行属性分析，得到待处理图像的场景属性信息，该神经网络可以为深度神经网络，本公开实施例不对该神经网络做限定。
在一种可能的实现方式中，所述场景属性信息可以包括场景法线信息、反射率信息、及光场信息中的至少一项。
其中，场景法线信息包括待处理图像中各像素点所在平面的法线值，反射率信息包括待处理图像中各像素点对应的反射率值，光场信息包括待处理图像各像素点的球谐系数。其中，场景法线信息、反射率信息、及光场信息可以以图、矩阵、向量等方式进行展示，本公开实施例对于场景法线信息、反射率信息、及光场信息的展示方式不作限定。为了使本领域技术人员更好的理解本公开实施例，本公开实施例中以图的展示方式为例，对场景法线信息、反射率信息、及光场信息进行展示。
示例性的，场景法线信息展示为场景法线贴图，场景法线贴图中任一像素点为单位长度矢量，该单位长度矢量为待处理图像中该像素点所在平面的法线值。发射率信息展示为反射率图，反射率图中任一像素点的像素值为待处理图像中该像素点对应的反射率值。光场信息展示为光场图，光场图表示为待处理图像中每一像素点的9个球谐系数。示例性的，可以对待处理图像进行属性分析，得到待处理图像的场景法线信息、反射率信息、及光场信息。
在步骤S22中，根据所述场景属性信息进行渲染处理，得到渲染图像。
举例来说，可以通过渲染器对得到的待处理图像的场景属性信息进行渲染处理，得到渲染图像。示例性的，渲染器可以由基本的矩阵乘法运算来实现着色渲染过程，渲染器即充当像素着色器，针对每个像素点的着色都是单独计算的。这样，可以使得在用于对待处理图像进行属性分析的神经网络的训练过程中，能够无缝地利用图像处理器(Graphics Processing Unit，GPU)加速集成。
示例性的，对待处理图像进行属性分析得到的场景属性信息包括场景法线信息、反射率信息、及光场信息时，可以通过下述公式(一)实现渲染过程。
Figure PCTCN2021103195-appb-000001
其中，
Figure PCTCN2021103195-appb-000002
用于表示渲染图像，
Figure PCTCN2021103195-appb-000003
用于表示反射率信息，
Figure PCTCN2021103195-appb-000004
用于表示光场信息，
Figure PCTCN2021103195-appb-000005
用于表示场景法线信息，B用于表示渲染方式，本公开实施例不对B做限定。在一种可能的渲染方式中，可以对
Figure PCTCN2021103195-appb-000006
所在的半球进行随机方向采样光场得到对应方向的光照强度信息，并乘以采样光线方向和法线方向点积得到颜色值并求平均颜色值。
在步骤S23中，根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。
举例来说，在得到渲染图像后，可以根据渲染图像及待处理图像，对场景属性信息进行优化，以使得优化后的场景属性信息，经渲染后得到的渲染图像与待处理图像无限接近。
本公开实施例可以对待处理图像进行属性分析，得到所述待处理图像的场景属性信息，并根据所述场景属性信息进行渲染处理，得到渲染图像，进而根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。根据本公开实施例提供的图像处理方法，可以通过对待处理图像的场景属性信息进行优化，提高待处理图像的场景属性信息的精度。
在一种可能的实现方式中，所述根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息，可以包括：
确定所述待处理图像与所述渲染图像的差异；
在所述差异大于差异阈值时，对所述场景属性信息进行优化，并根据优化后的场景属性信息进行渲染处理，直至优化后的场景属性信息进行渲染处理得到的渲染图像，与所述待处理图像的差异小于或等于差异阈值。
举例来说，可以确定待处理图像与渲染图像之间的差异。示例性的，可以将待处理图像及渲染图像分别作为数组加载，并逐个计算两个数组中元素之间的差，通过计算两个数组中元素之间的差的范数得到待处理图像与渲染图像之间的差异。或者，可以通过计算待处理图像与渲染图像的特征向量之间的距离，得到待处理图像与渲染图像之间的差异。本公开实施例不对确定待处理图像与渲染图像之间的差异的方式做限定，任一用于确定图像之间差异的方式均可。
在差异大于差异阈值时(差异阈值为预设的数值，取值可以根据需求进行确定)，可以确定当前场景属性信息精准度并不高，故可以对场景属性信息进行优化，即调整场景属性信息中各像素点的值，并循环根据优化后的场景属性信息进行渲染得到渲染图像，根据渲染图像与待处理图像之间的差异对场景属性信息进行优化的过程，直至优化后的场景属性信息渲染后得到的渲染图像与待处理图像之间的差异小于或等于差异阈值。
示例性的，可以对至少一项场景属性信息进行优化，例如：对待处理图像进行属性分析后，得到待处理图像的场景法线信息、反射率信息及光场信息，则可以通过场景法线信息、反射率信息及光场信息进行渲染后，得到对应的渲染图像。可以根据渲染图像及待处理图像之间的差异对反射率信息及光场信息进行优化，并可以根据优化后的反射率信息、优化后的光场信息及场景法线信息进行渲染得到优化后的渲染图像，并根据该优化后的渲染图像及待处理图像之间的差异继续对反射率信息及光场信息进行优化，直至优化后的渲染图像及待处理图像之间的差异小于或等于差异阈值。
这样，通过渲染图像与待处理图像之间的差异优化场景属性信息，可以提高获得的场景属性信息的精度。
在一种可能的实现方式中，所述对待处理图像进行属性分析，得到所述待处理图像 的场景属性信息，可以包括：
将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息。
举例来说，可以预训练用于对待处理图像进行属性分析的场景属性估计网络，将待处理图像作为该场景属性估计网络的输入，该场景属性估计网络的输出即为待处理图像的场景属性信息。
在一种可能的实现方式中，所述场景属性估计网络可以包括第一估计网络，所述第一估计网络可以包括第一特征提取编码器、第一解码器及第二解码器，所述将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息，可以包括：
将所述待处理图像输入所述第一特征编码器中进行特征提取，得到第一图像特征；
通过所述第一解码器对所述第一图像特征进行属性分析，得到所述待处理图像的反射率信息，及通过所述第二解码器对所述第一图像特征进行属性分析，得到所述待处理图像的光场信息。
举例来说，第一估计网络用于对待处理图像进行属性分析，得到待处理图像的反射率信息和光场信息，其可以采用U-net结构。本公开一些实施例中，第一估计网络可以包括第一特征提取编码器、第一解码器及第二解码器，其中第一解码器通过对待处理图像的图像特征进行属性分析，可以得到待处理图像的反射率信息，第二解码器通过对待处理图像的图像特征进行属性分析，可以得到待处理图像的光场信息。
由于反射率和光照系数紧密相关，并且高度依赖于图像的局部信息，故第一估计网络中第一解码器及第二解码器共用一个第一特征提取编码器，由第一特征提取编码器对待处理图像进行特征提取，得到第一图像特征后，分别通过第一解码器及第二解码器对第一图像特征进行属性分析，得到待处理图像的反射率信息和光场信息。
这样一来，可以缓解过度拟合的情况，提高待处理图像的反射率信息和光场信息的精度，还可以显著减小场景属性估计网络的规模。
在一种可能的实现方式中，其所述场景属性估计网络可以包括第二估计网络，所述第二估计网络包括第二特征提取编码器及第三解码器，所述将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息，可以包括：
将所述待处理图像输入所述第二特征编码器中进行特征提取，得到第二图像特征；
通过所述第三解码器对所述第二图像特征进行属性分析，得到所述待处理网络的场景法线信息。
举例来说，第二估计网络用于对待处理图像进行属性分析，得到待处理图像的场景法线信息。全局信息对于估计场景法线贴图至关重要，因此，第二估计网络单独使用与第一估计网络类似的U-net结构，网络权重不与第一估计网络共享。本公开一些实施例中，第二估计网络可以包括第二特征提取编码器及第三解码器，由第二特征提取编码器对待处理图像进行特征提取，得到第二图像特征后，通过第三解码器对第二图像特征进行属 性分析，得到待处理图像的场景法线信息。
这样一来，可以提高获得的待处理图像的场景法线信息的精度。
在一种可能的实现方式中，所述方法还可以包括：根据预设的训练集训练所述场景属性估计网络，所述训练集中包括多个样本图像，所述样本图像具有预标注的场景属性信息，其中，所述样本图像为根据所述预标注的场景属性信息进行渲染操作得到的。
举例来说，本公开实施例中可以根据预设的场景属性信息，预先生成大量的样本数据，例如：可以根据预设的3维场景属性数据通过渲染处理生成预设的光场信息、预设的场景法线信息和预设的反射率信息，进而根据预设的场景属性信息通过渲染处理生成样本图像。其中，用于生成该样本图像的场景属性信息即作为该样本图像的标注信息。根据生成的大量的样本图像可以构建用于训练场景属性估计网络的训练集。
示例性的，参照图3所示，场景属性估计网络可以包括第一估计网络和第二估计网络，反射率和光照系数紧密相关，并且高度依赖于图像的局部信息，而全局信息对于估计场景法线信息至关重要，故第一估计网络中第一解码器及第二解码器共用一个第一特征提取编码器，第二估计网络单独使用一个第二特征提取编码器。
本公开一些实施例中，第一估计网络可以包括第一特征提取编码器、第一解码器和第二解码器，第二估计网络可以包括第二特征提取编码器和第三解码器。示例性的，编码器(包括第一特征提取编码器和第二特征提取编码器)包含六个残差块，每个残差块可以具有三个卷积层，其中第一层的卷积步长为2，其余两层的卷积步长为1。第一估计网络及第二估计网络统一使用ReLU激活函数。解码器(包括第一解码器、第二解码器和第三解码器)可以使用双线性插值和正向卷积。每个解码器具有12个卷积层，特征图数量与编码器相同，顺序相反。在每个解码器块处，实现跳过连接以通过级联传输信息。
样本图像分别输入第一特征提取编码器及第二特征提取编码器中，第一特征提取编码器对样本图像进行特征提取后，得到第一图像特征，第二特征提取编码器对样本图像进行特征提取后，得到第二图像特征。第一图像特征分别输入第一解码器和第二解码器中，由第一解码器对第一图像特征进行属性分析，可以得到样本图像的反射率信息，由第二解码器对第一图像特征进行属性分析，可以得到样本图像的光场信息。第二图像特征输入第三解码器进行属性分析后，可以得到样本图像的场景法线信息。
通过样本图像的标注信息(预设的光场信息、预设的场景法线信息和预设的反射率信息)和样本图像的场景属性信息(光场信息、场景法线信息和反射率信息)，可以得到场景属性估计网络的网络损失，在网络损失未达到训练精度要求时(例如：网络损失大于损失阈值)，可以通过该网络损失调整场景属性估计网络的网络参数，直至场景属性估计网络的网络损失满足训练精度要求(例如：网络损失小于损失阈值)，完成训练，得到训练后的场景属性估计网络。
需要说明的是，本公开实施例对于计算网络损失的方式不做限定方式，任一计算网络损失的方式均可以，例如：1范式、2范式等方式。
这样一来，可以通过场景属性估计网络估计待处理图像的场景属性信息，包括反射 率信息，场景法线信息和光场信息，对于待处理图像的分解更加彻底，进而使得通过场景属性信息进行重建的重建效果也更加准确，并且获得的空间变化的光场信息和场景法线信息，可用于重光照，材质替换等，为增强现实的诸多应用提供了可能性。
为使本领域技术人员更好的理解本公开实施例，以下通过示例对本公开实施例加以说明，本公开实施例仅作为一种示例，而不理解为是一种限定。
参照图3，待处理图像输入场景属性估计网络。第一特征提取编码器可以对待处理图像进行特征提取，得到第一图像特征，第二特征提取编码器可以对待处理图像进行特征提取，得到第二图像特征。第一解码器对第一图像特征进行属性估计，可以得到待处理图像的反射率信息，第二解码器对第一图像特征进行属性估计，可以得到待处理图像的光场信息。第三解码器对第二图像特征进行属性估计，可以得到待处理图像的场景法线信息(图3中反射率信息以反射率图的形式展示、光场信息以光场图的形式展示、场景法线信息以场景法线贴图的形式展示)。
通过待处理图像的反射率信息、光场信息和场景法线信息进行渲染，可以得到渲染图像。进一步的通过渲染图像及待处理图像之间的差异可以对待处理图像的反射率信息、光场信息进行优化，得到优化后的反射率信息和优化后的光场信息。
在一种可能的实现方式中，所述方法还可以包括：
从优化后的场景属性信息中确定至少一项第一场景属性信息；
根据第二场景属性信息及所述第一场景属性信息，对所述待处理图像进行渲染处理，得到场景渲染图像。
举例来说，在得到待处理图像优化后的场景属性信息后，可以通过对优化后的场景属性信息中的至少一项进行变化或者替换处理等，以实现对待处理图像的重光照或者纹理替换等操作。或者，可以根据待处理图像的至少一项优化后的场景属性信息对其他图像进行重光照或者纹理替换等操作。
示例性的，第一场景属性信息可以为优化后的场景属性信息中的至少一项，第二场景属性信息为用于对优化后的场景属性信息中，除第一场景属性信息以外的第三场景属性信息进行替换的场景属性信息，该第二场景属性信息可以为预置的信息，也可以为对第三场景属性信息进行调整变化后得到的信息。
通过渲染器对第二场景属性信息及第一场景属性信息进行渲染处理，可以得到对应的场景渲染图像。
下面，结合示例场景对本公开实施例提供的图像处理方法进行详细阐述。
参考图4所示的流程示意图，在重光照应用场景中，电子设备可以通过以下步骤实现重光操作、纹理替换等AR应用。
S41、获取待处理图像，对待处理图像进行属性分析，得到待处理图像的场景属性信息。
示例性的，电子设备可以通过图像采集装置，采集单张室内场景图像。参考图5A和图5B所示，电子设备可以采集中午的室内场景图像，得到图5A所示的待处理图像A， 以及图5B所示的待处理图像B。
电子设备可以使用场景属性估计网络，分别对待处理图像A和待处理图像B进行处理。得到待处理图像A和待处理图像B的场景法线信息、光场信息和反射率信息。
示例性的，待处理图像A的反射率信息，可以通过图6A所示的反射率图A的形式展示。待处理图像A的场景法线信息，可以通过图7A所示的场景法线贴图A的形式展示。待处理图像A的光场信息，可以通过图8A所示的光场图A的形式展示。
同样地，待处理图像B的反射率信息，可以通过图6B所示的反射率图B的形式展示。待处理图像B的场景法线信息，可以通过图7B所示的场景法线贴图A的形式展示。待处理图像B的光场信息，可以通过图8B所示的光场图B的形式展示。
S42、根据场景属性信息进行渲染处理，得到渲染图像。
S43、基于渲染器对场景属性信息进行优化，得到优化后的场景属性信息。
示例性的，待处理图像A优化后的场景属性信息，可以通过图9A所示的场景渲染图A的形式展示。待处理图像B优化后的场景属性信息，可以通过图9B所示的场景渲染图B展示。
S44、根据优化后的场景属性信息，进行重光照、纹理替换等多种增强现实应用。
示例性的，可以确定待处理图像A的场景法线信息和反射率信息为第一场景属性信息。并可以将夜间场景下的光场信息(可以通过对待处理图像A的光场信息进行调整得到，或者直接对夜间场景下拍摄的图像进行属性分析，得到光场信息)作为第二场景属性信息，对待处理图像A的场景法线信息和反射率信息及夜间场景下的光场信息进行渲染，可以得到待处理图像A进行重光照处理后的重光照图，模拟得到该场景在夜间的图像。其中，重光照图可以参考图10A所示。
另外，可以确定待处理图像B的场景法线信息和光场信息为第一场景属性信息，并可以将阴天场景下的反射率信息(可以通过对待处理图像B的反射率信息进行调整得到，或者直接对阴天场景下拍摄的图像进行属性分析，得到发射率信息)作为第二场景属性信息，对待处理图像B的场景法线信息和光场信息及阴天场景下的反射率信息进行渲染，可以得到待处理图像B进行纹理替换处理后的纹理替换图，模拟得到该场景在阴天的图像。其中，纹理替换图可以参考图10B所示。
本公开实施例可以估计获取待处理图像在现实场景的固有场景属性信息，包括反射率信息，场景法线信息和光场信息。用户可以给定不同的场景属性信息，来获取该现实场景对应的场景渲染图像，实现重光照或者纹理替换。也可以利用估计的场景属性信息将虚拟物体嵌入到真实环境中，虚拟物体可能是某个虚拟游戏角色，可能是装潢时的家具，达到虚实融合的效果。
本公开实施例可以得到待处理图像的固有场景属性信息，包括反射率信息，场景法线信息和光场信息，对待处理图像的分解更加彻底，进而使得根据场景属性信息的重建效果也更加准确。获得的场景属性信息，可用于重光照，材质替换等，为增强现实的诸多应用提供了可能性。
可以理解，本公开提及的上述各个方法实施例，在不违背原理逻辑的情况下，均可以彼此相互结合形成结合后的实施例，限于篇幅，本公开不再赘述。本领域技术人员可以理解，在实施方式的上述方法中，各步骤的执行顺序应当以其功能和可能的内在逻辑确定。
此外，本公开还提供了图像处理装置、电子设备、计算机可读存储介质、程序，上述均可用来实现本公开提供的任一种图像处理方法，相应技术方案和描述和参见方法部分的相应记载，不再赘述。
图11示出根据本公开实施例的图像处理装置的框图，如图11所示，所述装置包括：
分析部分61，可以被配置为对待处理图像进行属性分析，得到所述待处理图像的场景属性信息；
第一渲染部分62，可以被配置为根据所述场景属性信息进行渲染处理，得到渲染图像；
优化部分63，可以被配置为根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。
本公开实施例可以对待处理图像进行属性分析，得到所述待处理图像的场景属性信息，并根据所述场景属性信息进行渲染处理，得到渲染图像，进而根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。根据本公开实施例提供的图像处理装置，可以通过对待处理图像的场景属性信息进行优化，提高待处理图像的场景属性信息的精度。
在一种可能的实现方式中，所述优化部分63，还可以被配置为：
确定所述待处理图像与所述渲染图像的差异；
在所述差异大于差异阈值时，对所述场景属性信息进行优化，并根据优化后的场景属性信息进行渲染处理，直至优化后的场景属性信息进行渲染处理得到的渲染图像，与所述待处理图像的差异小于或等于差异阈值。
在一种可能的实现方式中，所述场景属性信息可以包括场景法线信息、反射率信息、及光场信息中的至少一项。
在一种可能的实现方式中，所述分析部分61，还可以被配置为：
将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息。
在一种可能的实现方式中，所述场景属性估计网络包括第一估计网络，所述第一估计网络包括第一特征提取编码器、第一解码器及第二解码器，所述分析部分61，还可以被配置为：
将所述待处理图像输入所述第一特征编码器中进行特征提取，得到第一图像特征；
通过所述第一解码器对所述第一图像特征进行属性分析，得到所述待处理图像的反射率信息，及通过所述第二解码器对所述第一图像特征进行属性分析，得到所述待处理图像的光场信息。
在一种可能的实现方式中，所述场景属性估计网络包括第二估计网络，所述第二估计网络包括第二特征提取编码器及第三解码器，所述分析部分61，还可以被配置为：
将所述待处理图像输入所述第二特征编码器中进行特征提取，得到第二图像特征；
通过所述第三解码器对所述第二图像特征进行属性分析，得到所述待处理图像的场景法线信息。
在一种可能的实现方式中，所述装置还可以包括：
训练部分，可以被配置为根据预设的训练集训练所述场景属性估计网络，所述训练集中包括多个样本图像，所述样本图像具有预标注的场景属性信息，其中，所述样本图像为根据所述预标注的场景属性信息进行渲染操作得到的。
在一种可能的实现方式中，所述装置还可以包括：
确定部分，可以被配置为从优化后的场景属性信息中确定至少一项第一场景属性信息；
第二渲染部分，可以被配置为根据第二场景属性信息及所述第一场景属性信息，对所述待处理图像进行渲染处理，得到场景渲染图像。在一些实施例中，本公开实施例提供的装置具有的功能或包含的模块可以用于执行上文方法实施例描述的方法，其实现可以参照上文方法实施例的描述，为了简洁，这里不再赘述。
本公开实施例还提出一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现上述方法。计算机可读存储介质可以是非易失性计算机可读存储介质。
本公开实施例还提出一种电子设备，包括：处理器；用于存储处理器可执行指令的存储器；其中，所述处理器被配置为调用所述存储器存储的指令，以执行上述方法。
本公开实施例还提供了一种计算机程序产品，包括计算机可读代码，当计算机可读代码在设备上运行时，设备中的处理器执行用于实现如上任一实施例提供的图像处理方法的指令。
本公开实施例还提供了另一种计算机程序产品，用于存储计算机可读指令，指令被执行时使得计算机执行上述任一实施例提供的图像处理方法的操作。
电子设备可以被提供为终端、服务器或其它形态的设备。
图12示出根据本公开实施例的一种电子设备800的框图。例如，电子设备800可以是移动电话，计算机，数字广播终端，消息收发设备，游戏控制台，平板设备，医疗设备，健身设备，个人数字助理等终端。
参照图12，电子设备800可以包括以下一个或多个组件：处理组件802，存储器804，电源组件806，多媒体组件808，音频组件810，输入/输出(I/O)的接口812，传感器组件814，以及通信组件816。
处理组件802通常控制电子设备800的整体操作，诸如与显示，电话呼叫，数据通信，相机操作和记录操作相关联的操作。处理组件802可以包括一个或多个处理器820来执行指令，以完成上述的方法的全部或部分步骤。此外，处理组件802可以包括一个或多个 模块，便于处理组件802和其他组件之间的交互。例如，处理组件802可以包括多媒体模块，以方便多媒体组件808和处理组件802之间的交互。
存储器804被配置为存储各种类型的数据以支持在电子设备800的操作。这些数据的示例包括用于在电子设备800上操作的任何应用程序或方法的指令，联系人数据，电话簿数据，消息，图片，视频等。存储器804可以由任何类型的易失性或非易失性存储设备或者它们的组合实现，如静态随机存取存储器(SRAM)，电可擦除可编程只读存储器(EEPROM)，可擦除可编程只读存储器(EPROM)，可编程只读存储器(PROM)，只读存储器(ROM)，磁存储器，快闪存储器，磁盘或光盘。
电源组件806为电子设备800的各种组件提供电力。电源组件806可以包括电源管理系统，一个或多个电源，及其他与为电子设备800生成、管理和分配电力相关联的组件。
多媒体组件808包括在所述电子设备800和用户之间的提供一个输出接口的屏幕。在一些实施例中，屏幕可以包括液晶显示器(LCD)和触摸面板(TP)。如果屏幕包括触摸面板，屏幕可以被实现为触摸屏，以接收来自用户的输入信号。触摸面板包括一个或多个触摸传感器以感测触摸、滑动和触摸面板上的手势。所述触摸传感器可以不仅感测触摸或滑动动作的边界，而且还检测与所述触摸或滑动操作相关的持续时间和压力。在一些实施例中，多媒体组件808包括一个前置摄像头和/或后置摄像头。当电子设备800处于操作模式，如拍摄模式或视频模式时，前置摄像头和/或后置摄像头可以接收外部的多媒体数据。每个前置摄像头和后置摄像头可以是一个固定的光学透镜系统或具有焦距和光学变焦能力。
音频组件810被配置为输出和/或输入音频信号。例如，音频组件810包括一个麦克风(MIC)，当电子设备800处于操作模式，如呼叫模式、记录模式和语音识别模式时，麦克风被配置为接收外部音频信号。所接收的音频信号可以被进一步存储在存储器804或经由通信组件816发送。在一些实施例中，音频组件810还包括一个扬声器，用于输出音频信号。
I/O接口812为处理组件802和外围接口模块之间提供接口，上述外围接口模块可以是键盘，点击轮，按钮等。这些按钮可包括但不限于：主页按钮、音量按钮、启动按钮和锁定按钮。
传感器组件814包括一个或多个传感器，用于为电子设备800提供各个方面的状态评估。例如，传感器组件814可以检测到电子设备800的打开/关闭状态，组件的相对定位，例如所述组件为电子设备800的显示器和小键盘，传感器组件814还可以检测电子设备800或电子设备800一个组件的位置改变，用户与电子设备800接触的存在或不存在，电子设备800方位或加速/减速和电子设备800的温度变化。传感器组件814可以包括接近传感器，被配置用来在没有任何的物理接触时检测附近物体的存在。传感器组件814还可以包括光传感器，如互补金属氧化物半导体(CMOS)或电荷耦合装置(CCD)图像传感器，用于在成像应用中使用。在一些实施例中，该传感器组件814还可以包括加速度传感器，陀螺仪传感器，磁传感器，压力传感器或温度传感器。
通信组件816被配置为便于电子设备800和其他设备之间有线或无线方式的通信。电子设备800可以接入基于通信标准的无线网络，如无线网络(WiFi)，第二代移动通信技术(2G)或第三代移动通信技术(3G)，或它们的组合。在一个示例性实施例中，通信组件816经由广播信道接收来自外部广播管理系统的广播信号或广播相关信息。在一个示例性实施例中，所述通信组件816还包括近场通信(NFC)模块，以促进短程通信。例如，在NFC模块可基于射频识别(RFID)技术，红外数据协会(IrDA)技术，超宽带(UWB)技术，蓝牙(BT)技术和其他技术来实现。
在示例性实施例中，电子设备800可以被一个或多个应用专用集成电路(ASIC)、数字信号处理器(DSP)、数字信号处理设备(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、控制器、微控制器、微处理器或其他电子元件实现，用于执行上述方法。
在示例性实施例中，还提供了一种非易失性计算机可读存储介质，例如包括计算机程序指令的存储器804，上述计算机程序指令可由电子设备800的处理器820执行以完成上述方法。
图13示出根据本公开实施例的一种电子设备1900的框图。例如，电子设备1900可以被提供为一服务器。参照图13，电子设备1900包括处理组件1922，其进一步包括一个或多个处理器，以及由存储器1932所代表的存储器资源，用于存储可由处理组件1922的执行的指令，例如应用程序。存储器1932中存储的应用程序可以包括一个或一个以上的每一个对应于一组指令的模块。此外，处理组件1922被配置为执行指令，以执行上述方法。
电子设备1900还可以包括一个电源组件1926被配置为执行电子设备1900的电源管理，一个有线或无线网络接口1950被配置为将电子设备1900连接到网络，和一个输入输出(I/O)接口1958。电子设备1900可以操作基于存储在存储器1932的操作系统，例如微软服务器操作系统(Windows ServerTM)，苹果公司推出的基于图形用户界面操作系统(Mac OS XTM)，多用户多进程的计算机操作系统(UnixTM),自由和开放原代码的类Unix操作系统(LinuxTM)，开放原代码的类Unix操作系统(FreeBSDTM)或类似。
在示例性实施例中，还提供了一种非易失性计算机可读存储介质，例如包括计算机程序指令的存储器1932，上述计算机程序指令可由电子设备1900的处理组件1922执行以完成上述方法。
在示例性实施例中，还提供了一种计算机程序，包括计算机可读代码，当所述计算机可读代码在电子设备中运行时，所述电子设备中的处理器执行配置为实现上述图像处理方法实施例的步骤。
本公开可以是系统、方法和/或计算机程序产品。计算机程序产品可以包括计算机可读存储介质，其上载有用于使处理器实现本公开的各个方面的计算机可读程序指令。
计算机可读存储介质可以是可以保持和存储由指令执行设备使用的指令的有形设备。计算机可读存储介质例如可以是(但不限于)电存储设备、磁存储设备、光存储设备、电磁存储设备、半导体存储设备或者上述的任意合适的组合。计算机可读存储介质 的更详细的例子(非穷举的列表)包括：便携式计算机盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、静态随机存取存储器(SRAM)、便携式压缩盘只读存储器(CD-ROM)、数字多功能盘(DVD)、记忆棒、软盘、机械编码设备、例如其上存储有指令的打孔卡或凹槽内凸起结构、以及上述的任意合适的组合。这里所使用的计算机可读存储介质不被解释为瞬时信号本身，诸如无线电波或者其他自由传播的电磁波、通过波导或其他传输媒介传播的电磁波(例如，通过光纤电缆的光脉冲)、或者通过电线传输的电信号。
这里所描述的计算机可读程序指令可以从计算机可读存储介质下载到各个计算/处理设备，或者通过网络、例如因特网、局域网、广域网和/或无线网下载到外部计算机或外部存储设备。网络可以包括铜传输电缆、光纤传输、无线传输、路由器、防火墙、交换机、网关计算机和/或边缘服务器。每个计算/处理设备中的网络适配卡或者网络接口从网络接收计算机可读程序指令，并转发该计算机可读程序指令，以供存储在各个计算/处理设备中的计算机可读存储介质中。
用于执行本公开操作的计算机程序指令可以是汇编指令、指令集架构(ISA)指令、机器指令、机器相关指令、微代码、固件指令、状态设置数据、或者以一种或多种编程语言的任意组合编写的源代码或目标代码，所述编程语言包括面向对象的编程语言—诸如Smalltalk、C++等，以及常规的过程式编程语言—诸如“C”语言或类似的编程语言。计算机可读程序指令可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中，远程计算机可以通过任意种类的网络—包括局域网(LAN)或广域网(WAN)—连接到用户计算机，或者，可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。在一些实施例中，通过利用计算机可读程序指令的状态信息来个性化定制电子电路，例如可编程逻辑电路、现场可编程门阵列(FPGA)或可编程逻辑阵列(PLA)，该电子电路可以执行计算机可读程序指令，从而实现本公开的各个方面。
这里参照根据本公开实施例的方法、装置(系统)和计算机程序产品的流程图和/或框图描述了本公开的各个方面。应当理解，流程图和/或框图的每个方框以及流程图和/或框图中各方框的组合，都可以由计算机可读程序指令实现。
这些计算机可读程序指令可以提供给通用计算机、专用计算机或其它可编程数据处理装置的处理器，从而生产出一种机器，使得这些指令在通过计算机或其它可编程数据处理装置的处理器执行时，产生了实现流程图和/或框图中的一个或多个方框中规定的功能/动作的装置。也可以把这些计算机可读程序指令存储在计算机可读存储介质中，这些指令使得计算机、可编程数据处理装置和/或其他设备以特定方式工作，从而，存储有指令的计算机可读介质则包括一个制造品，其包括实现流程图和/或框图中的一个或多个方框中规定的功能/动作的各个方面的指令。
也可以把计算机可读程序指令加载到计算机、其它可编程数据处理装置、或其它设 备上，使得在计算机、其它可编程数据处理装置或其它设备上执行一系列操作步骤，以产生计算机实现的过程，从而使得在计算机、其它可编程数据处理装置、或其它设备上执行的指令实现流程图和/或框图中的一个或多个方框中规定的功能/动作。
附图中的流程图和框图显示了根据本公开的多个实施例的系统、方法和计算机程序产品的可能实现的体系架构、功能和操作。在这点上，流程图或框图中的每个方框可以代表一个模块、程序段或指令的一部分，所述模块、程序段或指令的一部分包含一个或多个用于实现规定的逻辑功能的可执行指令。在有些作为替换的实现中，方框中所标注的功能也可以以不同于附图中所标注的顺序发生。例如，两个连续的方框实际上可以基本并行地执行，它们有时也可以按相反的顺序执行，这依所涉及的功能而定。也要注意的是，框图和/或流程图中的每个方框、以及框图和/或流程图中的方框的组合，可以用执行规定的功能或动作的专用的基于硬件的系统来实现，或者可以用专用硬件与计算机指令的组合来实现。
该计算机程序产品可以通过硬件、软件或其结合的方式实现。在一个可选实施例中，所述计算机程序产品体现为计算机存储介质，在另一个可选实施例中，计算机程序产品体现为软件产品，例如软件开发包(Software Development Kit，SDK)等等。
以上已经描述了本公开的各实施例，上述说明是示例性的，并非穷尽性的，并且也不限于所披露的各实施例。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术的改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。

Claims (19)

  1. 一种图像处理方法，包括：
    对待处理图像进行属性分析，得到所述待处理图像的场景属性信息；
    根据所述场景属性信息进行渲染处理，得到渲染图像；
    根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。
  2. 根据权利要求1所述的方法，其中，所述根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息，包括：
    确定所述待处理图像与所述渲染图像的差异；
    在所述差异大于差异阈值时，对所述场景属性信息进行优化，并根据优化后的场景属性信息进行渲染处理，直至优化后的场景属性信息进行渲染处理得到的渲染图像，与所述待处理图像的差异小于或等于差异阈值。
  3. 根据权利要求1或2所述的方法，其中，所述场景属性信息包括场景法线信息、反射率信息、及光场信息中的至少一项。
  4. 根据权利要求1至3中任一项所述的方法，其中，所述对待处理图像进行属性分析，得到所述待处理图像的场景属性信息，包括：
    将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息。
  5. 根据权利要求4所述的方法，其中，所述场景属性估计网络包括第一估计网络，所述第一估计网络包括第一特征提取编码器、第一解码器及第二解码器，所述将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息，包括：
    将所述待处理图像输入所述第一特征编码器中进行特征提取，得到第一图像特征；
    通过所述第一解码器对所述第一图像特征进行属性分析，得到所述待处理图像的反射率信息，及通过所述第二解码器对所述第一图像特征进行属性分析，得到所述待处理图像的光场信息。
  6. 根据权利要求4或5所述的方法，其中，所述场景属性估计网络包括第二估计网络，所述第二估计网络包括第二特征提取编码器及第三解码器，所述将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息，包括：
    将所述待处理图像输入所述第二特征编码器中进行特征提取，得到第二图像特征；
    通过所述第三解码器对所述第二图像特征进行属性分析，得到所述待处理图像的场景法线信息。
  7. 根据权利要求4至6中任一项所述的方法，其中，所述方法还包括：
    根据预设的训练集训练所述场景属性估计网络，所述训练集中包括多个样本图像， 所述样本图像具有预标注的场景属性信息，其中，所述样本图像为根据所述预标注的场景属性信息进行渲染操作得到的。
  8. 根据权利要求1至7中任一项所述的方法，其中，所述方法还包括：
    从优化后的场景属性信息中确定至少一项第一场景属性信息；
    根据第二场景属性信息及所述第一场景属性信息，对所述待处理图像进行渲染处理，得到场景渲染图像。
  9. 一种图像处理装置，包括：
    分析部分，被配置为对待处理图像进行属性分析，得到所述待处理图像的场景属性信息；
    第一渲染部分，被配置为根据所述场景属性信息进行渲染处理，得到渲染图像；
    优化部分，被配置为根据所述待处理图像及所述渲染图像，对所述场景属性信息进行优化，得到优化后的场景属性信息。
  10. 根据权利要求9所述的装置，其中，所述优化部分，还被配置为：
    确定所述待处理图像与所述渲染图像的差异；
    在所述差异大于差异阈值时，对所述场景属性信息进行优化，并根据优化后的场景属性信息进行渲染处理，直至优化后的场景属性信息进行渲染处理得到的渲染图像，与所述待处理图像的差异小于或等于差异阈值。
  11. 根据权利要求9或10所述的装置，其中，所述场景属性信息包括场景法线信息、反射率信息、及光场信息中的至少一项。
  12. 根据权利要求9至11任一项所述的装置，其中，所述分析部分，还被配置为：
    将所述待处理图像输入场景属性估计网络进行属性分析，得到所述待处理图像的场景属性信息。
  13. 根据权利要求12所述的装置，其中，所述场景属性估计网络包括第一估计网络，所述第一估计网络包括第一特征提取编码器、第一解码器及第二解码器；
    所述分析部分，还被配置为：
    将所述待处理图像输入所述第一特征编码器中进行特征提取，得到第一图像特征；
    通过所述第一解码器对所述第一图像特征进行属性分析，得到所述待处理图像的反射率信息，及通过所述第二解码器对所述第一图像特征进行属性分析，得到所述待处理图像的光场信息。
  14. 根据权利要求12或13所述的装置，其中，所述场景属性估计网络包括第二估计网络，所述第二估计网络包括第二特征提取编码器及第三解码器；
    所述分析部分，还被配置为：
    将所述待处理图像输入所述第二特征编码器中进行特征提取，得到第二图像特征；
    通过所述第三解码器对所述第二图像特征进行属性分析，得到所述待处理图像的场景法线信息。
  15. 根据权利要求12至14任一项所述的装置，其中，所述装置还包括：
    训练部分，被配置为根据预设的训练集训练所述场景属性估计网络，所述训练集中包括多个样本图像，所述样本图像具有预标注的场景属性信息，其中，所述样本图像为根据所述预标注的场景属性信息进行渲染操作得到的。
  16. 根据权利要求9至15任一项所述的装置，其中，所述装置还包括：
    确定部分，被配置为从优化后的场景属性信息中确定至少一项第一场景属性信息；
    第二渲染部分，被配置为根据第二场景属性信息及所述第一场景属性信息，对所述待处理图像进行渲染处理，得到场景渲染图像。
  17. 一种电子设备，包括：
    处理器；
    用于存储处理器可执行指令的存储器；
    其中，所述处理器被配置为调用所述存储器存储的指令，以执行权利要求1至8中任意一项所述的方法。
  18. 一种计算机可读存储介质，其上存储有计算机程序指令，所述计算机程序指令被处理器执行时实现权利要求1至8中任意一项所述的方法。
  19. 一种计算机程序，包括计算机可读代码，当所述计算机可读代码在电子设备中运行时，所述计算机设备中的处理器执行用于实现权利要求1至8任一项所述的方法。
PCT/CN2021/103195 2021-01-19 2021-06-29 图像处理方法及装置、电子设备、存储介质及计算机程序 WO2022156150A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110070584.XA CN112785672B (zh) 2021-01-19 2021-01-19 图像处理方法及装置、电子设备和存储介质
CN202110070584.X 2021-01-19

Publications (1)

Publication Number Publication Date
WO2022156150A1 true WO2022156150A1 (zh) 2022-07-28

Family

ID=75757708

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/103195 WO2022156150A1 (zh) 2021-01-19 2021-06-29 图像处理方法及装置、电子设备、存储介质及计算机程序

Country Status (2)

Country Link
CN (1) CN112785672B (zh)
WO (1) WO2022156150A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112785672B (zh) * 2021-01-19 2022-07-05 浙江商汤科技开发有限公司 图像处理方法及装置、电子设备和存储介质
CN113537194A (zh) * 2021-07-15 2021-10-22 Oppo广东移动通信有限公司 光照估计方法、光照估计装置、存储介质与电子设备
CN113436247B (zh) * 2021-07-29 2024-03-01 北京达佳互联信息技术有限公司 一种图像处理方法、装置、电子设备及存储介质
CN116612244B (zh) * 2023-07-21 2023-09-29 之江实验室 一种图像生成方法、装置、存储介质及电子设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180365890A1 (en) * 2015-06-17 2018-12-20 Rosemount Aerospace Inc. System and method for processing captured images
CN109166170A (zh) * 2018-08-21 2019-01-08 百度在线网络技术（北京）有限公司 用于渲染增强现实场景的方法和装置
CN111652960A (zh) * 2020-05-07 2020-09-11 浙江大学 一种基于可微渲染器的从单幅图像求解人脸反射材质的方法
CN112183637A (zh) * 2020-09-29 2021-01-05 中科方寸知微(南京)科技有限公司 一种基于神经网络的单光源场景光照重渲染方法及系统
CN112785672A (zh) * 2021-01-19 2021-05-11 浙江商汤科技开发有限公司 图像处理方法及装置、电子设备和存储介质

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104134230B (zh) * 2014-01-22 2015-10-28 腾讯科技（深圳）有限公司 一种图像处理方法、装置及计算机设备
EP3642800A4 (en) * 2017-07-10 2020-05-20 Samsung Electronics Co., Ltd. POINT CLOUD AND NETWORK COMPRESSION WITH IMAGE / VIDEO CODECS
CN108986199B (zh) * 2018-06-14 2023-05-16 北京小米移动软件有限公司 虚拟模型处理方法、装置、电子设备及存储介质
CN110659581B (zh) * 2019-08-29 2024-02-20 腾讯科技（深圳）有限公司 一种图像处理方法、装置、设备和存储介质
CN110855958B (zh) * 2019-11-18 2022-04-26 RealMe重庆移动通信有限公司 图像调节方法及装置、电子设备、存储介质
CN110910486B (zh) * 2019-11-28 2021-11-19 浙江大学 室内场景光照估计模型、方法、装置、存储介质以及渲染方法
CN111462268B (zh) * 2020-03-31 2022-11-11 北京市商汤科技开发有限公司 图像重建方法及装置、电子设备和存储介质

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180365890A1 (en) * 2015-06-17 2018-12-20 Rosemount Aerospace Inc. System and method for processing captured images
CN109166170A (zh) * 2018-08-21 2019-01-08 百度在线网络技术（北京）有限公司 用于渲染增强现实场景的方法和装置
CN111652960A (zh) * 2020-05-07 2020-09-11 浙江大学 一种基于可微渲染器的从单幅图像求解人脸反射材质的方法
CN112183637A (zh) * 2020-09-29 2021-01-05 中科方寸知微(南京)科技有限公司 一种基于神经网络的单光源场景光照重渲染方法及系统
CN112785672A (zh) * 2021-01-19 2021-05-11 浙江商汤科技开发有限公司 图像处理方法及装置、电子设备和存储介质

Also Published As

Publication number Publication date
CN112785672A (zh) 2021-05-11
CN112785672B (zh) 2022-07-05

Similar Documents

Publication Publication Date Title
WO2022156150A1 (zh) 图像处理方法及装置、电子设备、存储介质及计算机程序
JP7106687B2 (ja) 画像生成方法および装置、電子機器、並びに記憶媒体
TWI718631B (zh) 人臉圖像的處理方法及裝置、電子設備和儲存介質
WO2022188305A1 (zh) 信息展示方法及装置、电子设备、存储介质及计算机程序
TW202032425A (zh) 圖像處理方法及裝置、電子設備和儲存介質
US20220206581A1 (en) Communication interface with haptic feedback response
WO2023051356A1 (zh) 一种虚拟对象的显示方法及装置、电子设备和存储介质
CN111815750A (zh) 对图像打光的方法及装置、电子设备和存储介质
US11989348B2 (en) Media content items with haptic feedback augmentations
US11531400B2 (en) Electronic communication interface with haptic feedback response
US20220317774A1 (en) Real-time communication interface with haptic and audio feedback response
US20220317773A1 (en) Real-time communication interface with haptic and audio feedback response
WO2022212175A1 (en) Interface with haptic and audio feedback response
WO2022212174A1 (en) Interface with haptic and audio feedback response
EP4315001A1 (en) Virtual reality interface with haptic feedback response
WO2022147151A1 (en) Real-time video communication interface with haptic feedback
WO2022147450A1 (en) Communication interface with haptic feedback response
US20220319059A1 (en) User-defined contextual spaces
WO2022151687A1 (zh) 合影图像生成方法、装置、设备、存储介质、计算机程序及产品
WO2022042290A1 (zh) 一种虚拟模型处理方法、装置、电子设备和存储介质
WO2022134416A1 (zh) 视频数据处理方法及装置、电子设备和存储介质
WO2022212144A1 (en) User-defined contextual spaces
US11825276B2 (en) Selector input device to transmit audio signals
US11874960B2 (en) Pausing device operation based on facial movement
US11922587B2 (en) Dynamic augmented reality experience

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2022532031

Country of ref document: JP

Kind code of ref document: A

121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21920531

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21920531

Country of ref document: EP

Kind code of ref document: A1