CN112634143A - Image color correction model training method and device and electronic equipment - Google Patents
Image color correction model training method and device and electronic equipment Download PDFInfo
- Publication number
- CN112634143A CN112634143A CN201910905790.0A CN201910905790A CN112634143A CN 112634143 A CN112634143 A CN 112634143A CN 201910905790 A CN201910905790 A CN 201910905790A CN 112634143 A CN112634143 A CN 112634143A
- Authority
- CN
- China
- Prior art keywords
- color correction
- image color
- sample data
- training
- dimensional rgb
- 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.)
- Pending
Links
- 238000012937 correction Methods 0.000 title claims abstract description 201
- 238000012549 training Methods 0.000 title claims abstract description 175
- 238000000034 method Methods 0.000 title claims abstract description 55
- 230000006870 function Effects 0.000 claims description 45
- 238000004590 computer program Methods 0.000 claims description 9
- 238000005192 partition Methods 0.000 claims description 8
- 230000001131 transforming effect Effects 0.000 claims description 5
- 230000011218 segmentation Effects 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 238000013528 artificial neural network Methods 0.000 description 6
- 238000000638 solvent extraction Methods 0.000 description 6
- 238000012545 processing Methods 0.000 description 5
- 238000000611 regression analysis Methods 0.000 description 5
- 238000004422 calculation algorithm Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012417 linear regression Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 238000010238 partial least squares regression Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000007792 addition Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T19/00—Manipulating 3D models or images for computer graphics
- G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10024—Color image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2219/00—Indexing scheme for manipulating 3D models or images for computer graphics
- G06T2219/20—Indexing scheme for editing of 3D models
- G06T2219/2012—Colour editing, changing, or manipulating; Use of colour codes
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Architecture (AREA)
- Computer Graphics (AREA)
- Computer Hardware Design (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The application discloses an image color correction model training method and device, a computer readable storage medium and electronic equipment. The method comprises the following steps: carrying out interval division on three-dimensional RGB sample data according to a preset mode to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data; training an image color correction network through the multiple sections of the three-dimensional RGB sample data, and updating parameters of the image color correction network to obtain the image color correction model. The training speed of the image color correction network and the accuracy of the trained image color correction model can be improved.
Description
Technical Field
The invention relates to the field of machine learning, in particular to an image color correction model training method and device, a computer readable storage medium and electronic equipment.
Background
With the widespread use of digital image devices, the uniformity of image color reproduction has gradually become a hot spot of current research. The color information of the image is often an important basis for image analysis, so that the research on the color correction technology capable of truly reflecting the color of the observed object has important research significance. The goal of color correction is to investigate how to describe the intrinsic color of an object under various lighting conditions. At present, color correction is applied to a plurality of image processing scenes such as medical images, remote sensing images, mural images, license images and the like, and with the improvement of image technology, the color correction can be more widely applied to the daily life of people in the future.
At present, under the influence of ambient light and optical elements of a camera, an original image signal obtained by the camera has a certain error compared with a real scene. In order to solve the problem, some methods for image color correction have been proposed in succession, such as a traditional color correction method based on partial least squares regression, a traditional neural network color correction method, and the like. The traditional color correction method based on partial least squares regression can better solve the problems of multiple correlation among independent variables and relatively few samples, but the precision is difficult to meet the practical application requirement. The traditional color correction method based on the neural network has larger parameter quantity, occupies more multipliers and is not beneficial to hardware realization.
Disclosure of Invention
The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides an image color correction model training method and device, a computer readable storage medium and electronic equipment.
According to an aspect of the present application, there is provided an image color correction model training method, including: carrying out interval division on three-dimensional RGB sample data according to a preset mode to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data; training an image color correction network through the multiple sections of the three-dimensional RGB sample data, and updating parameters of the image color correction network to obtain the image color correction model.
According to a second aspect of the present application, there is provided an image color correction model training apparatus, comprising: the division module is used for carrying out interval division on three-dimensional RGB sample data according to a preset mode to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data; and the model training module is used for respectively training the image color correction network through the multiple sections of the three-dimensional RGB sample data and updating parameters of the image color correction network to obtain the image color correction model.
According to a third aspect of the present application, there is provided a computer-readable storage medium storing a computer program for executing any one of the image color correction model training methods described above.
According to a fourth aspect of the present application, there is provided an electronic apparatus comprising: a processor; a memory for storing the processor-executable instructions; the processor is configured to execute any one of the image color correction model training methods.
The technical scheme provided by the embodiment of the application can at least bring the following beneficial effects:
by carrying out interval division on the three-dimensional RGB sample data and respectively training the image color correction network by utilizing a plurality of interval sections obtained by the interval division, the training speed of the image color correction network and the accuracy of the trained image color correction model can be improved.
Drawings
The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.
Fig. 1 is a flowchart illustrating an image color correction model training method according to an exemplary embodiment of the present application.
Fig. 2 is a flowchart illustrating an image color correction model training method according to another exemplary embodiment of the present application.
Fig. 3 is a flowchart illustrating an image color correction model training method according to another exemplary embodiment of the present application.
Fig. 4 is a block diagram of an image color correction model training apparatus according to an exemplary embodiment of the present application.
Fig. 5 is a block diagram of an image color correction model training apparatus according to another exemplary embodiment of the present application.
Fig. 6 is a schematic structural diagram of an electronic device according to an exemplary embodiment of the present application.
Detailed Description
Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.
Summary of the application
The acquired image color has a great relationship with the acquisition environment, and the image color obtained under different acquisition environments is different, so an image color correction model needs to be provided for performing color correction on the acquired image.
In view of the above problems, embodiments of the present application provide a method and an apparatus for training an image color correction model, which can improve the training speed of the image color correction network and the accuracy of the trained image color correction model by performing interval division on three-dimensional RGB sample data and training the image color correction network by using a plurality of intervals obtained through the interval division.
The regression analysis prediction method is characterized in that on the basis of analyzing the correlation between independent variables and dependent variables of various phenomena, a regression equation between the variables is established, the regression equation is used as a prediction model, and the correlation between the dependent variables is predicted according to the quantity change of the independent variables in a prediction period, so that the regression analysis prediction method is an important prediction method. The traditional regression analysis prediction method can be used for printer color correction, and has the advantages of less required storage space, high calculation speed and easy realization. However, the accuracy of the conventional regression analysis prediction method is not ideal, and the sample data amount is generally small, so that the generalization capability of the conventional regression model is poor, and thus the image color correction effect cannot be well achieved.
In an embodiment of the application, the image color correction model obtained by combining the machine learning technology with the regression analysis has higher accuracy and model generalization capability, and the obtained image color correction model has low calculation complexity and is beneficial to realizing color correction on the image to be processed on the terminal equipment.
Exemplary method
Fig. 1 is a flowchart illustrating an image color correction model training method according to an exemplary embodiment of the present application. The embodiment can be applied to an electronic device, which can be, for example, a game host, a laptop portable computer, or a server. As shown in fig. 1, the following steps 110 and 110 are included.
And 110, performing interval division on the three-dimensional RGB sample data according to a preset mode to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data.
The three-dimensional RGB sample data is used for training the image color correction network. In some embodiments, the three-dimensional RGB sample data may be 255 × 255 RGB standard values of an RGB color space, where different RGB standard values correspond to different colors in the RGB color space, but the present application is not limited thereto, and the three-dimensional RGB sample data may further include other RGB values, for example.
In order to obtain a more accurate image color correction model, a large amount of three-dimensional RGB sample data is generally acquired. However, when the image color correction network is trained directly through a large amount of three-dimensional RGB sample data, the training speed of the image color correction network is affected. Therefore, in an embodiment, the three-dimensional RGB sample data is subjected to interval division to obtain a plurality of interval segments corresponding to the three-dimensional RGB sample data. The image color correction network is trained through a plurality of sections of the three-dimensional RGB sample data, so that the training efficiency can be improved, the training error in the training process of the image color correction network is reduced, the accuracy of the trained image color correction model is improved, and the trained image color correction model has a better color correction effect on an image to be corrected.
The embodiment of the application does not limit the partition mode for performing interval partition on the three-dimensional RGB sample data, for example, the three-dimensional RGB sample data may be randomly divided into a plurality of interval sections, or the three-dimensional RGB sample data may be uniformly divided into a plurality of interval sections, or the three-dimensional RGB sample data may be subjected to interval partition in a mode that a training error of a current round of training is minimum, the training error is second minimum or smaller, or the three-dimensional RGB sample data may be subjected to interval partition in a mode that multiple interval partitions are combined.
In an embodiment, the image color correction network is, for example, a regression equation based on adaboost, but this is not limited in this application, and may also be a regression equation based on other machine learning algorithms. The regression equation is a linear regression equation, and in another embodiment, the regression equation may also be a polynomial regression equation, which is not limited in this application.
In an embodiment, the image color correction network is trained through the sub-sample data corresponding to the multiple segments of the three-dimensional RGB sample data, so as to obtain the training error of the image color correction network in each segment, and then the parameters of the image color correction network are updated according to the training error of each segment. For example, when the number of training rounds for training the image color correction network or the training error of the current round satisfies a preset condition, the training is finished, and at this time, the trained image color correction network model can be obtained. For example, the current round training error is calculated according to the training errors of the current blocks, and may be, for example, a weighted sum of the training errors of the current blocks, or an average value, a median value, and the like of the training errors of the current blocks, which is not limited in this embodiment of the application.
In addition, according to the training method of the image color correction model provided by the embodiment of the application, the three-dimensional RGB sample data is subjected to interval division, and the image color correction network is trained by utilizing a plurality of interval sections obtained through the interval division, so that the training speed of the image color correction network and the accuracy of the trained image color correction model can be improved.
Fig. 2 is a flowchart illustrating an image color correction model training method according to another exemplary embodiment of the present application. The embodiment can be applied to an electronic device, which can be, for example, a game host, a laptop portable computer, or a server. As shown in fig. 2, the method includes the following steps 210 to 290.
And 210, transforming the three-dimensional RGB standard value according to the color regression model to obtain a three-dimensional RGB target value corresponding to the three-dimensional RGB standard value.
In an embodiment, the color regression model is a Back Propagation (BP) neural network model, but the present embodiment is not limited thereto, and may also be other neural network models. The color regression model obtained by training the BP neural network will be exemplified below. For example, the process of training the BP neural network may be:
a) acquiring a color card image obtained by shooting a standard color card;
b) preprocessing the color card image to eliminate color deviation caused by exposure and color temperature in the shooting process, wherein the preprocessing comprises black balance correction, white balance correction, Demosaic correction, Gamma correction and the like;
c) extracting the RGB value of each pixel point from the preprocessed color card image to obtain RGB data of the color card image;
d) and inputting the RGB data of the color card image and the RGB data of the standard color card into a back propagation BP neural network for processing to obtain the color regression model. The RGB data of the color chip image is sample data of a back propagation BP neural network, the RGB data of the standard color chip is labeled data of the sample data, and the sample data corresponds to the labeled data one by one, or the RGB data of the color chip image corresponds to the RGB data of the standard color chip one by one.
In one embodiment, the three-dimensional RGB standard values are 255 × 255 RGB standard values of the RGB color space. And transforming the three-dimensional RGB standard value through a color regression model obtained through training to obtain a three-dimensional RGB target value corresponding to the three-dimensional RGB standard value.
The three-dimensional RGB sample data obtained according to the embodiment of the application comprises the three-dimensional RGB standard value and the three-dimensional RGB target value, and the data size of the three-dimensional RGB sample data is expanded, so that the image color correction model trained by the three-dimensional RGB sample data has better generalization capability and accuracy.
And step 230, determining a current value of the selection function based on the current training round number, the total training round number, the previous training error and a preset coefficient.
In some embodiments of the present application, the image color correction network is subjected to multiple rounds of iterative training through the three-dimensional RGB sample data, so as to obtain an image color correction model. Exemplarily, a current value of the selection function in the current round is determined based on the current training round number, the total training round number, the previous round training error and a preset coefficient, the three-dimensional RGB sample data is subjected to interval division according to the current value of the selection function, and the image color correction network is trained based on a plurality of interval obtained by the interval division to obtain the image color correction model.
In one embodiment, the selection function is:
wherein t is the current training round number, n is the total training round number of the image color correction network, loss is the previous training error of the image color correction network, and α, β, and γ are the coefficients of the selection function respectively. By setting the values of the corresponding alpha, beta and gamma, the current value of the selection function can be adjusted, and different modes for carrying out interval division on the three-dimensional RGB sample data are selected.
In one embodiment, α is greater than 0, less than 1; beta is greater than 1; gamma is greater than 0 and less than 1. For example, α may also be greater than γ, for example, γ is greater than 0 and less than 0.2, and α is greater than 0.5, and the specific values of α, β, and γ are not limited in this application embodiment, and a person skilled in the art may set the values thereof according to a specific application.
And 240, when the current value of the selection function meets a preset condition, performing interval division on the three-dimensional RGB sample data according to a preset mode.
In an embodiment, when the current value of the selection function is larger, the three-dimensional RGB sample data is divided into intervals in a random manner, that is, the three-dimensional RGB sample data is randomly divided into a plurality of intervals. In another embodiment, when the current value of the selection function is larger, the three-dimensional RGB sample data is divided into intervals in a uniform manner, that is, the three-dimensional RGB sample data is divided into a plurality of sections uniformly, so that the data amount of the sub-samples in the plurality of sections is equal.
In some embodiments, when the current value of the selection function is greater than a third preset value, the three-dimensional RGB sample data is randomly interval-divided. In some embodiments, when the current value of the selection function is greater than the fourth preset value, the three-dimensional RGB sample data may be further uniformly divided into intervals. The fourth preset value may be equal to the third preset value, or the fourth preset value is not equal to the third preset value, which is not limited in the embodiment of the present application.
In some embodiments of the present application, when a current value of the selection function is small, for example, when the selection function is less than or equal to a first preset value, m division points are selected from the three-dimensional RGB sample data for position division, where the position division divides the three-dimensional RGB sample data into m candidate block data sets according to the m division points, and m is a positive integer; performing regression operation on the m candidate block data sets respectively to obtain regression errors of the regression operation; and acquiring a division point which enables the regression error to be minimum as a preferred division point, and performing interval division on the three-dimensional RGB sample data based on the preferred division point.
For example, assume that the three-dimensional RGB sample data is N-dimensional in R, G, B three dimensions, so that there is N in total in the entire three-dimensional RGB sample data3RGB data. M division points are selected from the three-dimensional RGB sample data for position division, and the three-dimensional RGB sample data can be divided into m candidate block data sets, wherein m is smaller than N. For example, M division points may be selected from the three-dimensional RGB sample data, and M different candidate block data sets may be generated through M position divisions, each of the M position divisions dividing the three-dimensional RGB sample data into M +1 data sets, where M is the number of data sets of the three-dimensional RGB sample data in the current training stage, where M +1 is less than N3. That is, after the position division, the three-dimensional RGB sample data is divided into M +1 data sets at the current round of training, M candidate block data sets each include M +1 data sets, and the RGB data included in each of the M +1 data sets in the M candidate block data sets are different.
In one embodiment, the regression operation is performed on the M candidate partition data sets, that is, on the M different M +1 data sets based on the M different partition points. For example, the regression operation may be a polynomial regression operation or other linear regression operations, which is not limited in the embodiments of the present application. The regression equation in the regression operation may be a linear equation, a quadratic equation, or a g-order equation according to an actual situation, where g is a positive integer greater than 2, which is not limited in the embodiment of the present application. The embodiment of the application respectively performs regression operation on the M +1 data sets, and calculates the regression error of the M +1 data sets after the regression operation according to the input of the regression operation, namely the RGB data values and the current weight of the RGB data in the M +1 data sets, the estimated value of the regression operation and the preset loss function. And obtaining the regression error of the three-dimensional RGB sample data in the p-th position division according to the sum or the average value of the regression errors of the M +1 data sets, wherein p is more than or equal to 1 and less than or equal to M. For example, the predetermined loss function may be a Mean Square Error (MSE), and a minimum value of the predetermined loss function may be calculated by using a gradient descent algorithm or a least Square method, for example, to obtain a regression Error for each of the M +1 data sets.
And comparing regression errors of the three-dimensional RGB sample data at the m dividing points, acquiring the dividing point with the minimum regression error as a preferred dividing point, and performing interval division on the three-dimensional RGB sample data based on the preferred dividing point.
Illustratively, the preferred division point may be calculated using the following function (1):
wherein (x)i(ii) a ) The estimated value of the regression operation when the division position is p; y isiThe true value input for the regression operation, i.e. the RGB value, w, in the three-dimensional RGB sample dataiIs the weight coefficient of the input of the regression operation, g is the total number of the input of the regression operation, pos is such that
And obtaining the value of the division position p when the minimum value is obtained, wherein p is more than or equal to 1 and less than or equal to m.
In an embodiment, the preferred dividing point selected during each round of image color correction network training may be different, but the embodiment of the present application does not limit this.
In some embodiments of the present application, when the selection function is less than or equal to a second preset value, k sub data chunks are selected from three-dimensional RGB sample data, where the sub data chunks include three-dimensional data; performing regression operation based on the k sub-data blocks respectively to obtain regression errors of the regression operation; and carrying out interval division on the three-dimensional RGB sample data based on the sub data block which enables the regression error to be minimum.
Assuming that the three-dimensional RGB sample data is N-dimensional in R, G, B three dimensions, the three-dimensional RGB sample data is fullThe total number of N in this data3RGB data. K sub-data blocks are selected from the three-dimensional RGB sample data, and k segmentations can be performed on the three-dimensional RGB sample data, wherein k is less than N3. For example, k sub-data blocks may be selected from the three-dimensional RGB sample data through k segmentation, and each segmentation in the k segmentation divides the three-dimensional RGB sample data into L +1 sub-data blocks, where L is the number of sub-data blocks of the three-dimensional RGB sample data in the current training stage, and L +1 is smaller than N3And L is a positive integer. That is, after each segmentation, the three-dimensional RGB sample data is segmented into L +1 sub-data blocks in the current round of training, and k times of segmentation may generate k different sets of L +1 sub-data blocks, that is, the RGB data included in the k different sets of L +1 sub-data blocks are different.
In an embodiment, a regression operation is performed on the k sub-data blocks, that is, the regression operation is performed on k different L +1 sub-data blocks through k times of segmentation. The regression algorithm is not limited in the embodiments of the present application, and the regression operation may be a polynomial regression operation or other linear regression operations, for example. The regression equation in the regression operation can be designed according to actual conditions, and the regression equation is not limited in the embodiment of the application. The embodiment of the application respectively performs regression operation on the L +1 sub-data blocks, and calculates the regression error of the L +1 sub-data blocks after the regression operation according to the input of the regression operation, namely the RGB data value and the current weight of the RGB data in the L +1 sub-data blocks, the estimated value of the regression operation and the preset loss function. And obtaining the regression error of the three-dimensional RGB sample data in k-time position division according to the sum or the mean value of the regression errors of the L +1 sub-data blocks. For example, the predetermined loss function may be a Mean Square Error (MSE), and a minimum value of the predetermined loss function may be calculated by using a gradient descent algorithm or a least Square method, for example, to obtain a regression Error of each L +1 sub-data blocks.
And comparing regression errors of the three-dimensional RGB sample data in k-time segmentation, and performing interval division on the three-dimensional RGB sample data based on the sub-data block with the minimum regression error.
For example, the function (1) may be used to obtain, by a similar calculation method, a segmentation mode that minimizes a regression error of the three-dimensional RGB sample data, and obtain, based on the segmentation mode that minimizes the regression error, a sub-data block that minimizes the regression error, and further perform interval division on the three-dimensional RGB sample data.
It should be noted that the first preset value may be equal to the second preset value, or the first preset value is not equal to the second preset value, which is not limited in the embodiment of the present application. In an embodiment, the first preset value and the second preset value are less than or equal to the third preset value and the fourth preset value.
The expression capacity of polynomial regression operation and linear regression operation on nonlinear data is limited, the three-dimensional RGB sample data are divided into intervals, the similarity of the three-dimensional RGB sub-sample data in each interval is higher than that of the whole three-dimensional RGB sample data, loss in training of the image color correction network can be effectively reduced, the image color correction network can reach the convergence condition faster, and the accuracy of the image color correction model of the embodiment of the application is improved.
In an embodiment, the image color correction network is a regression operation based on machine learning, for example, the image color correction network may be a polynomial regression operation based on adaboost, but the embodiment of the present application does not limit this.
Initially, for example, the weight coefficients of each sample data in the three-dimensional RGB sample data are set to be the same, and a polynomial regression operation based on adaboost is performed on the three-dimensional RGB sample data to obtain a regression error, a weight coefficient of a regression equation in the polynomial regression operation, and a weight coefficient of the three-dimensional RGB sample data. In the subsequent iterative operation process, the three-dimensional RGB sample data is divided into a plurality of sections, and the adaboost-based polynomial regression operation is performed in each section, so that the training error of each section, the weight coefficient of the regression equation of each section, and the weight coefficient of the sub-sample data of each section can be obtained.
In an embodiment, the weighted sum or the average of the training errors of the segments is used as the current training error of the image color correction network, but this is not limited by the embodiment of the present application.
According to the current training error of the image color correction network, updating the parameters of the image color correction network, wherein the parameters can be, for example, the weight coefficients of the regression equation in the image color correction network, the weight coefficients of the three-dimensional RGB sample data, and the like.
In one embodiment, whether to finish the iterative training of the image color correction network may be determined according to the current training round number or the current training error. For example, whether to finish the iterative training of the image color correction network may also be determined according to the current training round number and the current training error. And when the current training round number reaches the preset number and/or the current training error is smaller than the preset threshold, stopping the iterative training of the image color correction network, and executing step 290. When the number of current training rounds is less than the preset number of times and/or the current training error is greater than or equal to the preset threshold, step 230 is executed to perform the next round of training on the image color correction network. The preset times and the preset threshold may be set according to specific situations, which is not limited in the embodiment of the present application.
And 290, stopping updating the parameters, and obtaining the trained image color correction model.
And according to the trained image color correction model, performing color correction on the image to be corrected to obtain a corrected image, so that the display effect of the image is improved.
According to the image color correction model training method, the three-dimensional RGB sample data are subjected to interval division, the image color correction network is trained by utilizing a plurality of intervals obtained through the interval division, and the training speed of the image color correction network and the accuracy of the trained image color correction model can be improved. And the image color correction model obtained by training has low computational complexity and is easy to be deployed in hardware, so that the color correction of the image to be corrected is realized through the image color correction model.
Fig. 3 is a flowchart illustrating an image color correction model training method according to another exemplary embodiment of the present application. The embodiment can be applied to an electronic device, which can be, for example, a game host, a laptop portable computer, or a server. As shown in fig. 3, the method includes the following steps 310 to 390.
And 310, transforming the three-dimensional RGB standard value according to the color regression model to obtain a three-dimensional RGB target value corresponding to the three-dimensional RGB standard value.
And step 320, acquiring the three-dimensional RGB standard value and the three-dimensional RGB target value as three-dimensional RGB sample data.
And 340, when the current value of the selection function meets a preset condition, performing interval division on the three-dimensional RGB sample data according to a preset mode.
And step 360, calculating the current training error of the image color correction network according to the training errors of the sections.
The steps 310 to 360 are similar to the steps 210 to 260, and are not described herein again.
In an embodiment, according to the training error of each segment, the weight coefficient of the regression equation in the image color correction network corresponding to each segment, the weight coefficient of the sub-sample data of each segment, and the like are updated correspondingly. By training the image color correction network for each segment, a plurality of sub-image color correction models can be obtained, and training errors are reduced and training speed is increased.
And 380, judging whether the current training round number reaches a preset number and/or whether the current training error is smaller than a preset threshold value.
In one embodiment, whether to finish the iterative training of the image color correction network may be determined according to the current training round number or the current training error. For example, whether to finish the iterative training of the image color correction network may also be determined according to the current training round number and the current training error. And when the current training round number reaches the preset number and/or the current training error is smaller than the preset threshold, stopping the iterative training of the image color correction network, and executing step 290. When the number of current training rounds is less than the preset number of times and/or the current training error is greater than or equal to the preset threshold, step 230 is executed to perform the next round of training on the image color correction network. The preset times and the preset threshold may be set according to specific situations, which is not limited in the embodiment of the present application.
And correspondingly updating the parameters of the image color correction network corresponding to each interval according to the training errors of each interval, thereby obtaining the sub-image color correction model corresponding to each interval. By combining the plurality of sub-image color correction models, a trained image color correction model can be obtained. For example, the trained image color correction model may be obtained by directly stitching a plurality of sub-image color correction models, but the embodiment of the present application does not limit this.
According to the image color correction model training method, the three-dimensional RGB sample data are subjected to interval division, the image color correction network is trained by utilizing a plurality of intervals obtained through the interval division, and the training speed of the image color correction network and the accuracy of the trained image color correction model can be improved.
And the image color correction model obtained by training has low computational complexity and is easy to be deployed in hardware, so that the color correction of the image to be corrected is realized through the image color correction model.
Exemplary devices
Fig. 4 is a block diagram of an image color correction model training apparatus according to an exemplary embodiment of the present application. The device has the function of implementing the embodiment shown in fig. 1, fig. 2 or fig. 3, and the function may be implemented by hardware, or may be implemented by hardware executing corresponding software. The apparatus may include: a partitioning module 410 and a model training module 420.
The dividing module 410 is configured to perform interval division on the three-dimensional RGB sample data according to a preset manner, so as to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data.
A model training module 420, configured to train an image color correction network through the multiple segments of the three-dimensional RGB sample data, and update parameters of the image color correction network to obtain the image color correction model.
The image color correction model training device provided by the embodiment of the application can improve the training speed of the image color correction network and the accuracy of the trained image color correction model by performing interval division on three-dimensional RGB sample data and respectively training the image color correction network by utilizing a plurality of interval sections obtained by the interval division.
Fig. 5 is a block diagram of an image color correction model training apparatus according to another exemplary embodiment of the present application. The device has the function of implementing the embodiment shown in fig. 1, fig. 2 or fig. 3, and the function may be implemented by hardware, or may be implemented by hardware executing corresponding software. The apparatus may include: a partitioning module 510 and a model training module 520, and the partitioning module 510 includes a determining unit 5110 and a partitioning unit 5120, and the model training module 520 includes a first error calculating unit 5210, a second error calculating unit 5220, and an updating unit 5230.
The dividing module 510 is configured to perform interval division on the three-dimensional RGB sample data according to a preset manner, so as to obtain a plurality of interval segments corresponding to the three-dimensional RGB sample data.
The model training module 520 is configured to train the image color correction network through a plurality of segments of the three-dimensional RGB sample data, and update parameters of the image color correction network to obtain an image color correction model.
In an embodiment, the partitioning module 510 includes a determining unit 5110 and a partitioning unit 5120.
A determining unit 5110, configured to determine a current value of the selection function based on the current training round number, the total training round number, the previous training error, and a preset coefficient.
A dividing unit 5120, configured to perform interval division on the three-dimensional RGB sample data according to a preset manner when the current value of the selection function meets a preset condition.
In an embodiment, the model training module 520 includes a first error calculation unit 5210, a second error calculation unit 5220, and an update unit 5230.
A first error calculation unit 5210, configured to input the multiple segments of the three-dimensional RGB sample data into an image color correction network, respectively, to obtain training errors of the image color correction network in each segment;
a second error calculation unit 5220, configured to calculate a current round training error of the image color correction network according to the training errors of each segment;
an updating unit 5230 is used for updating the parameters of the image color correction network according to the current round of training errors.
In some embodiments provided based on the embodiment shown in fig. 5, the dividing unit 5120 may further be configured to: when the current value of the selection function is smaller than or equal to a first preset value, selecting m division points from three-dimensional RGB sample data for position division, wherein the position division divides the three-dimensional RGB sample data into m candidate block data sets according to the m division points, and m is a positive integer; performing regression operation on the m candidate block data sets respectively to obtain regression errors of the regression operation; and acquiring a division point which enables the regression error to be minimum as an optimal division point, and performing interval division on the three-dimensional RGB sample data based on the optimal division point.
In some embodiments provided based on the embodiment shown in FIG. 5, the preferred division point is chosen differently for each round of image color correction network training.
In some embodiments provided based on the embodiment shown in fig. 5, the dividing unit 5120 may further be configured to: when the current value of the selection function is smaller than or equal to a second preset value, selecting p sub data blocks from three-dimensional RGB sample data, wherein the sub data blocks are three-dimensional data; performing regression operation on the p sub-data blocks respectively to obtain regression errors of the regression operation; and carrying out interval division on the three-dimensional RGB sample data based on the sub-data blocks which enable the regression error to be minimum.
In some embodiments provided based on the embodiment shown in fig. 5, the dividing unit 5120 may further be configured to: when the current value of the selection function is larger than a third preset value, interval division is carried out on the three-dimensional RGB sample data randomly; or when the current value of the selection function is larger than a third preset value, the three-dimensional RGB sample data are uniformly divided into intervals.
In another embodiment of the present application, the model training module 520 includes a third error calculation unit, a second update unit, and a model determination unit (not shown in the figure).
And the third error calculation unit is used for respectively inputting the multiple sections of the three-dimensional RGB sample data into the image color correction network to respectively obtain the training errors of the image color correction network in each section.
The second updating unit is used for correspondingly updating the parameters of the image color correction network corresponding to each section according to the training errors of each section to obtain a plurality of sub-image color correction models;
and the model determining unit is used for obtaining the obtained image color correction model according to the plurality of sub-image color correction models.
In some embodiments provided based on the embodiment shown in fig. 5, the image color correction model training device further includes a model obtaining unit, configured to stop updating the parameters when the number of current training rounds reaches a preset number of times and/or when a training error of the current round is smaller than a preset threshold, and obtain a trained image color correction model.
In some embodiments provided based on the embodiment shown in fig. 5, the image color correction model training apparatus further includes a transformation module and a sample obtaining module (not shown in the figure).
The transformation module is used for transforming the three-dimensional RGB standard value according to the color regression model to obtain a three-dimensional RGB target value corresponding to the three-dimensional RGB standard value;
and the sample acquisition module is used for acquiring the three-dimensional RGB standard value and the three-dimensional RGB target value into three-dimensional RGB sample data.
The image color correction model training device provided by the embodiment of the application can improve the training speed of the image color correction network and the accuracy of the trained image color correction model by performing interval division on three-dimensional RGB sample data and respectively training the image color correction network by utilizing a plurality of interval sections obtained by the interval division. And the image color correction model obtained by training has low computational complexity and is easy to be deployed in hardware, so that the color correction of the image to be corrected is realized through the image color correction model.
Exemplary electronic device
Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 6.
FIG. 6 illustrates a block diagram of an electronic device in accordance with an embodiment of the present application.
As shown in fig. 6, the electronic device 10 includes one or more processors 11 and memory 12.
The processor 11 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.
Memory 12 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, Random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, Read Only Memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer readable storage medium and executed by the processor 11 to implement the image color correction model training methods of the various embodiments of the present application described above and/or other desired functions. Various contents such as an input signal, a signal component, a noise component, etc. may also be stored in the computer-readable storage medium.
In one example, the electronic device 10 may further include: an input device 13 and an output device 14, which are interconnected by a bus system and/or other form of connection mechanism (not shown).
For example, when the electronic device is the first device 100 or the second device 200, the input device 13 may be a microphone or a microphone array as described above for capturing an input signal of a sound source. When the electronic device is a stand-alone device, the input means 13 may be a communication network connector for receiving the acquired input signals from the first device 100 and the second device 200.
The input device 13 may also include, for example, a keyboard, a mouse, and the like.
The output device 14 may output various information including the determined distance information, direction information, and the like to the outside. The output devices 14 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices, among others.
Of course, for simplicity, only some of the components of the electronic device 10 relevant to the present application are shown in fig. 6, and components such as buses, input/output interfaces, and the like are omitted. In addition, the electronic device 10 may include any other suitable components depending on the particular application.
Exemplary computer program product and computer-readable storage Medium
In addition to the methods and apparatus described above, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the training of an image color correction model according to various embodiments of the present application described in the "exemplary methods" section above of this specification.
The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform the steps in the training of image color correction models according to various embodiments of the present application described in the "exemplary methods" section above in this specification.
The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.
The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by those skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".
It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.
Claims (13)
1. An image color correction model training method comprises the following steps:
carrying out interval division on three-dimensional RGB sample data according to a preset mode to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data;
training an image color correction network through the multiple sections of the three-dimensional RGB sample data, and updating parameters of the image color correction network to obtain the image color correction model.
2. The method according to claim 1, wherein the interval division of the three-dimensional RGB sample data according to a preset manner includes:
determining the current value of the selection function based on the current training round number, the total training round number, the previous training error and a preset coefficient;
and when the current value of the selection function meets a preset condition, carrying out interval division on the three-dimensional RGB sample data according to a preset mode.
3. The method according to claim 2, when the selection function meets a preset condition, performing interval division on the three-dimensional RGB sample data according to a preset manner, comprising:
when the current value of the selection function is smaller than or equal to a first preset value, selecting m division points from the three-dimensional RGB sample data for position division, wherein the position division divides the three-dimensional RGB sample data into m candidate block data sets according to the m division points, and m is a positive integer;
performing regression operation on the m candidate block data sets respectively to obtain regression errors of the regression operation;
and acquiring a division point which enables the regression error to be minimum as an optimal division point, and performing interval division on the three-dimensional RGB sample data based on the optimal division point.
4. The method of claim 3, wherein the preferred partition point is chosen differently for each round of image color correction network training.
5. The method according to claim 2, wherein, when the selection function meets a preset condition, performing interval division on the three-dimensional RGB sample data in a preset manner includes:
when the current value of the selection function is smaller than or equal to a second preset value, selecting p sub data blocks from the three-dimensional RGB sample data, wherein the sub data blocks are three-dimensional data;
performing regression operation on the p sub-data blocks respectively to obtain regression errors of the regression operation;
and carrying out interval division on the three-dimensional RGB sample data based on the sub-data blocks which enable the regression error to be minimum.
6. The method according to claim 2, wherein, when the selection function meets a preset condition, performing interval division on the three-dimensional RGB sample data in a preset manner includes:
when the current value of the selection function is larger than a third preset value, interval division is carried out on the three-dimensional RGB sample data randomly; or
And when the current value of the selection function is larger than a third preset value, uniformly carrying out interval division on the three-dimensional RGB sample data.
7. The method of any of claims 1-6, wherein said training an image color correction network through said plurality of segments of said three-dimensional RGB sample data, respectively, to update parameters of said image color correction network comprises:
inputting the multiple sections of the three-dimensional RGB sample data into an image color correction network respectively to obtain training errors of the image color correction network in each section respectively;
calculating the training error of the current round of the image color correction network according to the training error of each interval;
and updating the parameters of the image color correction network according to the training error of the current round.
8. The method of any of claims 1-6, wherein said training an image color correction network through said plurality of segments of said three-dimensional RGB sample data, respectively, to update parameters of said image color correction network comprises:
inputting the multiple sections of the three-dimensional RGB sample data into an image color correction network respectively to obtain training errors of the image color correction network in each section respectively;
correspondingly updating the parameters of the image color correction network corresponding to each interval according to the training errors of each interval to obtain a plurality of sub-image color correction models;
and obtaining the obtained image color correction model according to the plurality of sub-image color correction models.
9. The method of claim 7, further comprising:
and when the number of the current training rounds reaches a preset number and/or the training error of the current round is smaller than a preset threshold value, stopping updating the parameters to obtain a trained image color correction model.
10. The method of claim 1, further comprising:
transforming the three-dimensional RGB standard value according to the color regression model to obtain a three-dimensional RGB target value corresponding to the three-dimensional RGB standard value;
and acquiring the three-dimensional RGB standard value and the three-dimensional RGB target value as the three-dimensional RGB sample data.
11. An image color correction model training device, comprising:
the division module is used for carrying out interval division on three-dimensional RGB sample data according to a preset mode to obtain a plurality of interval sections corresponding to the three-dimensional RGB sample data;
and the model training module is used for respectively training the image color correction network through the multiple sections of the three-dimensional RGB sample data and updating parameters of the image color correction network to obtain the image color correction model.
12. A computer-readable storage medium storing a computer program for executing the image color correction model training method according to any one of claims 1 to 9.
13. An electronic device, the electronic device comprising:
a processor;
a memory for storing the processor-executable instructions;
the processor is configured to perform the image color correction model training method according to any one of claims 1 to 9.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910905790.0A CN112634143A (en) | 2019-09-24 | 2019-09-24 | Image color correction model training method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910905790.0A CN112634143A (en) | 2019-09-24 | 2019-09-24 | Image color correction model training method and device and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112634143A true CN112634143A (en) | 2021-04-09 |
Family
ID=75282792
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910905790.0A Pending CN112634143A (en) | 2019-09-24 | 2019-09-24 | Image color correction model training method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112634143A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116033133A (en) * | 2021-10-22 | 2023-04-28 | 广州视源电子科技股份有限公司 | Color correction method, storage medium and related device |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102768757A (en) * | 2012-06-28 | 2012-11-07 | 北京市遥感信息研究所 | Remote sensing image color correcting method based on image type analysis |
| CN102868891A (en) * | 2012-09-18 | 2013-01-09 | 哈尔滨商业大学 | Multi-angle view video chromatic aberration correction method based on support vector regression |
| US8913827B1 (en) * | 2010-05-10 | 2014-12-16 | Google Inc. | Image color correction with machine learning |
| US20160253787A1 (en) * | 2014-03-25 | 2016-09-01 | Spreadtrum Communications (Shanghai) Co., Ltd. | Methods and systems for denoising images |
| CN106408562A (en) * | 2016-09-22 | 2017-02-15 | 华南理工大学 | Fundus image retinal vessel segmentation method and system based on deep learning |
| CN106791759A (en) * | 2016-12-14 | 2017-05-31 | 南京巨鲨显示科技有限公司 | The bearing calibration of medical display color uniformity and correction system |
| CN107507250A (en) * | 2017-06-02 | 2017-12-22 | 北京工业大学 | A kind of complexion tongue color image color correction method based on convolutional neural networks |
| CN109166127A (en) * | 2018-07-17 | 2019-01-08 | 同济大学 | A kind of wearable plant phenotype sensory perceptual system |
| CN111062876A (en) * | 2018-10-17 | 2020-04-24 | 北京地平线机器人技术研发有限公司 | Method and device for correcting model training and image correction and electronic equipment |
-
2019
- 2019-09-24 CN CN201910905790.0A patent/CN112634143A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8913827B1 (en) * | 2010-05-10 | 2014-12-16 | Google Inc. | Image color correction with machine learning |
| CN102768757A (en) * | 2012-06-28 | 2012-11-07 | 北京市遥感信息研究所 | Remote sensing image color correcting method based on image type analysis |
| CN102868891A (en) * | 2012-09-18 | 2013-01-09 | 哈尔滨商业大学 | Multi-angle view video chromatic aberration correction method based on support vector regression |
| US20160253787A1 (en) * | 2014-03-25 | 2016-09-01 | Spreadtrum Communications (Shanghai) Co., Ltd. | Methods and systems for denoising images |
| CN106408562A (en) * | 2016-09-22 | 2017-02-15 | 华南理工大学 | Fundus image retinal vessel segmentation method and system based on deep learning |
| CN106791759A (en) * | 2016-12-14 | 2017-05-31 | 南京巨鲨显示科技有限公司 | The bearing calibration of medical display color uniformity and correction system |
| CN107507250A (en) * | 2017-06-02 | 2017-12-22 | 北京工业大学 | A kind of complexion tongue color image color correction method based on convolutional neural networks |
| CN109166127A (en) * | 2018-07-17 | 2019-01-08 | 同济大学 | A kind of wearable plant phenotype sensory perceptual system |
| CN111062876A (en) * | 2018-10-17 | 2020-04-24 | 北京地平线机器人技术研发有限公司 | Method and device for correcting model training and image correction and electronic equipment |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116033133A (en) * | 2021-10-22 | 2023-04-28 | 广州视源电子科技股份有限公司 | Color correction method, storage medium and related device |
| CN116033133B (en) * | 2021-10-22 | 2024-05-17 | 广州视源电子科技股份有限公司 | Color correction method, storage medium and related device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11557085B2 (en) | Neural network processing for multi-object 3D modeling | |
| CN109816589B (en) | Method and apparatus for generating cartoon style conversion model | |
| CN105069424B (en) | Quick face recognition system and method | |
| US9292911B2 (en) | Automatic image adjustment parameter correction | |
| US9070044B2 (en) | Image adjustment | |
| CN110245579B (en) | People flow density prediction method and device, computer equipment and readable medium | |
| CN110969200B (en) | Image target detection model training method and device based on consistency negative sample | |
| CN114925748B (en) | Model training and modal information prediction method, related device, equipment and medium | |
| CN111062876A (en) | Method and device for correcting model training and image correction and electronic equipment | |
| US20230021551A1 (en) | Using training images and scaled training images to train an image segmentation model | |
| EP4343616A1 (en) | Image classification method, model training method, device, storage medium, and computer program | |
| KR101981284B1 (en) | Apparatus Processing Image and Method thereof | |
| CN112364916A (en) | Image classification method based on transfer learning, related equipment and storage medium | |
| CN113392241B (en) | Method, device, medium and electronic equipment for identifying definition of well logging image | |
| CN114429641A (en) | Time sequence action detection method and device, storage medium and terminal | |
| CN112634143A (en) | Image color correction model training method and device and electronic equipment | |
| CN117315758A (en) | Facial expression detection method and device, electronic equipment and storage medium | |
| CN116958809A (en) | Remote sensing small sample target detection method for feature library migration | |
| US11688175B2 (en) | Methods and systems for the automated quality assurance of annotated images | |
| WO2022141094A1 (en) | Model generation method and apparatus, image processing method and apparatus, and readable storage medium | |
| CN114912540A (en) | Transfer learning method, device, equipment and storage medium | |
| CN114820755A (en) | Depth map estimation method and system | |
| CN113139540A (en) | Backboard detection method and equipment | |
| WO2022141092A1 (en) | Model generation method and apparatus, image processing method and apparatus, and readable storage medium | |
| WO2023220892A1 (en) | Expanded neural network training layers for convolution |
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 |