CN108009989A - A kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN - Google Patents
A kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN Download PDFInfo
- Publication number
- CN108009989A CN108009989A CN201711362202.0A CN201711362202A CN108009989A CN 108009989 A CN108009989 A CN 108009989A CN 201711362202 A CN201711362202 A CN 201711362202A CN 108009989 A CN108009989 A CN 108009989A
- Authority
- CN
- China
- Prior art keywords
- remote sensing
- microwave remote
- resolution
- sensing image
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 18
- 230000005855 radiation Effects 0.000 claims abstract description 5
- 238000012360 testing method Methods 0.000 claims description 7
- 238000013528 artificial neural network Methods 0.000 claims description 4
- 230000005540 biological transmission Effects 0.000 claims description 4
- 239000011159 matrix material Substances 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 2
- 238000011478 gradient descent method Methods 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 230000008569 process Effects 0.000 claims description 2
- 238000006467 substitution reaction Methods 0.000 claims 2
- 239000000126 substance Substances 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 6
- 238000013480 data collection Methods 0.000 abstract 2
- 238000004364 calculation method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000006872 improvement Effects 0.000 description 3
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
- 230000035515 penetration Effects 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000002759 woven fabric Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
-
- 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
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- 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]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Geophysics And Detection Of Objects (AREA)
Abstract
The invention discloses a kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN, belong to microwave remote sensing and Detection Techniques.Present invention simulated microwave remote sensing flow first, forward modeling produce high-resolution microwave remote sensing images TB and low resolution microwave remote sensing image TA composition data collection；Preprocessed data collection afterwards, produces SRCNN training sets；Five layer depth convolutional neural networks are built based on SRCNN training sets again；Pending low resolution microwave remote sensing image is finally input to the five layer depth convolutional neural networks built, the 4th layer of output of the depth convolutional neural networks is the high-resolution microwave remote sensing images after rebuilding.The present invention can effectively reduce the computation complexity in rebuilding and accurate antenna radiation pattern is not required, and be a kind of new microwave remote sensing image rebuilding method, can the more efficient bright temperature image for rebuilding original scene in real time.
Description
Technical Field
The invention belongs to the technical field of microwave remote sensing and detection, and particularly relates to a microwave remote sensing image super-resolution reconstruction method based on SRCNN.
Background
Microwave remote sensing is an extremely important remote sensing technology, has the characteristics of all-time and all-weather, and has deeper penetration capacity compared with visible light and infrared remote sensing. The satellite microwave remote sensing has many unique advantages of wide coverage space-time range, large detection information quantity, high detection frequency and the like. With the deep application research of satellite-borne radiometers in recent years in China, the role of the satellite-borne microwave radiometer in satellite microwave remote sensing is more important. The satellite-borne microwave radiometer can be used for detecting various elements of atmospheric environment. The weather information such as atmospheric temperature and humidity profile, cloud bottom height, liquid nitrogen and water in cloud, strong convection, tornado, solid water and precipitation rate in cloud can be monitored in the weather field, and powerful support is provided for civil and military weather forecast.
The spatial resolution of a satellite-borne microwave radiometer depends primarily on the half-power beamwidth of the antenna and the height of the satellite. The half-power beam width of the antenna is related to the detection wavelength and the antenna aperture, and the higher the detection frequency of the microwave radiometer is, the larger the antenna aperture is, and the higher the spatial resolution is. However, the low spatial resolution common to satellite-borne radiometers limits the usefulness in inverting surface parameters and weather meteorological data. At present, there are two main measures for improving the spatial resolution of a space-borne radiometer: on one hand, the aperture of the antenna is improved from the perspective of physical technology, but the aperture of the spaceborne radiometer antenna is limited by the space capacity of the rocket, and the limit of the aperture size exists; on the other hand, from the data processing perspective, the existing data and the antenna directional pattern measured in advance are utilized to recover the high-resolution image through the deconvolution technology, and currently, BG and SIR algorithms are generally adopted, but certain errors exist in the algorithms and the algorithms are long in time consumption.
Disclosure of Invention
Aiming at the defects or improvement requirements of the prior art, the invention provides a microwave remote sensing image super-resolution reconstruction method based on SRCNN, aiming at clarifying the microwave remote sensing image super-resolution reconstruction problem from the perspective of supervised learning selection, thereby solving the technical problems that the complexity of the current method is higher and a specific antenna pattern needs to be known.
In order to achieve the above object, the present invention provides a microwave remote sensing image super-resolution reconstruction method based on SRCNN, the method includes:
(1) Simulating a microwave remote sensing process, and forward generating a high-resolution microwave remote sensing image TB and a low-resolution microwave remote sensing image TA to form a data set;
(2) Preprocessing a data set to generate an SRCNN training set;
(3) Constructing a five-layer deep convolutional neural network based on the SRCNN training set;
(4) And inputting the low-resolution microwave remote sensing image to be processed into the constructed five-layer deep convolutional neural network, wherein the fourth-layer output of the deep convolutional neural network is the reconstructed high-resolution microwave remote sensing image.
Further, the step (1) is specifically: and forward generating a high-resolution microwave remote sensing image TB based on a microwave radiation transmission RT mode, and inputting the high-resolution microwave remote sensing image TB into a radiometer forward observation model to simulate and calculate a low-resolution microwave remote sensing image TA actually observed.
Further, the step (2) specifically includes:
(21) Performing pixel interpolation on the TA to obtain TC, so that the TC and the TB have the same pixel number;
(22) Respectively segmenting TB and TC to obtain image sets TD and TE, enabling the image sizes in TD and TE to be Z multiplied by Z, enabling Z to be a preset value, and enabling 35-layer-woven Z-layer-woven fabric to be 45, and preferably enabling Z =40;
(23) And (3) disordering the arrangement sequence of the images in the TD and the TE, selecting the first M images of the TD and the TE respectively as a sample set TDm and a label set TEm in a training data set, and selecting the rest images of the TD and the TE respectively as a sample set TDn and a label set TEN in a test data set.
Further, the step (2) further comprises:
(24) And generating the training data set and the testing data set into files in an HDF5 format.
Further, the step (3) specifically includes:
(31) Designing the structure of the network by using a caffe framework, and setting a weight matrix of the ith layer of the deep convolutional neural network as W i (ii) a Offset is b i I =2,3,4; substituting x into image in TDn i Substituting the image in TEN into y i I =1,2,3, · M; training the deep network, minimizing the output of the fifth layer,
wherein, the first and the second end of the pipe are connected with each other,wherein f (x) = max (0, x); n represents the number of samples in each training set;
(32) Setting hyper-parameters of a five-layer deep convolutional neural network;
(33) Optimizing the loss by adopting a random gradient descent method to ensure that a network weight matrix W i And offset b i Converging, and finishing the construction of the five-layer deep convolution neural network.
Further, the hyper-parameters of the five-layer deep convolutional neural network in the step (32) specifically include: network iteration times, network operation modes and network parameter storage positions.
Further, before the low-resolution microwave remote sensing image is input into the trained five-layer deep convolutional neural network in the step (4), linear interpolation needs to be performed on the low-resolution microwave remote sensing image to enable the number of pixels of the low-resolution microwave remote sensing image to be the same as the number of pixels of the sample label in the SRCNN training set.
Generally, compared with the prior art, the technical scheme of the invention has the following technical characteristics and beneficial effects:
the invention clarifies the problem of super-resolution reconstruction of microwave remote sensing images from the aspect of supervised learning selection: obtaining low-resolution brightness-temperature images with the same number of pixels as the reconstructed images through interpolation, extracting the characteristics of the low-resolution brightness-temperature images by a deep convolution neural network characteristic extraction layer, carrying out nonlinear conversion on a characteristic space by a nonlinear mapping layer, and carrying out filtering operation on the characteristic images in an averaging mode by an image reconstruction layer by aggregating high-resolution characteristic blocks extracted by convolution; the loss layer adopts an Euclidean distance loss function to calculate the mean square error of the reconstructed bright temperature image and the label bright temperature image, and then parameters of each layer of the network are corrected through backward propagation; after multiple iterations, the output of the loss layer is stabilized near a certain value through network convergence, so that a functional relation between the input low-resolution brightness-temperature image and the input super-resolution brightness-temperature image is obtained; the method can effectively reduce the calculation complexity of the image super-resolution reconstruction without knowing a specific antenna directional diagram, and is a novel microwave remote sensing image super-resolution reconstruction method.
Drawings
FIG. 1 is a flow chart of an implementation of a method embodiment of the present invention;
FIG. 2 (a) is an original low fraction sample light temperature image of example 1 of the method of the present invention;
FIG. 2 (b) is a super-resolution brightness-temperature image reconstructed in example 1 of the present invention;
FIG. 3 (a) is an original low fraction sample light temperature image of example 2 of the method of the present invention;
fig. 3 (b) is a super-resolution brightness-temperature image reconstructed in embodiment 2 of the method of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
The microwave remote sensing image super-resolution method provided by the embodiment of the invention is based on the SRCNN, can effectively reduce the calculation complexity of the microwave remote sensing image super-resolution method aiming at the real-time application of the microwave remote sensing image super-resolution and does not need to accurately determine an antenna directional diagram, and is a novel microwave remote sensing image super-resolution method. The embodiment takes a microwave remote sensing image with a frequency of 50.3GHz as an example.
Firstly, generating a high-resolution microwave remote sensing image TB with the frequency of 50.3GHz based on forward modeling of a microwave radiation transmission RT mode, and inputting the high-resolution microwave remote sensing image TB into a radiometer forward observation model to simulate and calculate a low-resolution microwave remote sensing image TA actually observed. Finally, a total of 96 TB and TA groups were generated.
And secondly, linearly interpolating the low-resolution brightness temperature image TA to obtain TC with the same size as the corresponding TB. The size of an image obtained by discriminating and dividing a TC image and a TB image is 40 × 40. Since TB is 239 × 299, a total of 3360 sample luminance temperature images TD and sample labels TE can be obtained. TD and TE were then randomly shuffled and the top 3150 groups were chosen as training data sets and converted to train.h5 in HDF5 format, and the remaining 210 groups were chosen as test data sets and converted to test.h5 in HDF5 format.
And thirdly, writing a network structure file net. The network has a total of 5 layers, an input layer and 3 convolutional layers and a final loss layer. Compiling a super parameter file, namely, a solution, setting the basic learning rate and the operation mode of the network and the storage position of network parameters, and training the network.
And fourthly, writing a network structure mat. And extracting and storing the generated network parameter ca ffemodel file in the third step as an x5.Mat file.
Fifthly, in fig. 2 (a), any low-resolution brightness-temperature image in the test data set is interpolated into images with the same number of pixels as the brightness-temperature images of the sample labels, and the images are used as the input of the deep convolutional neural network. Fig. 2 (b) is a super-resolution reconstructed brightness temperature image, which is output from layer 4 of the network. The mean square error RMSE1 between fig. 2 (a) and the sample label light temperature image was calculated to be 10.6534.
Sixthly, in fig. 3 (a), another low-resolution brightness-temperature image with the frequency of 50.3GHz is interpolated into the image with the same number of pixels as the sample label brightness-temperature image, and the interpolated image is used as the input of the deep convolutional neural network. . Fig. 3 (b) is a super-resolution reconstructed brightness temperature image which is output from the network layer 4. The mean square error RMSE3 between the sample label light temperature image and fig. 3 (a) is 5.1062, and the mean square error RMSE2 between the sample label light temperature image and fig. 2 (b) is 4.8673.
The results show that the spatial resolution of the microwave remote sensing bright temperature image is improved visually and clearly, and the resolution is improved by 60% according to the calculation of the microwave radiation transmission RT mode; from the mean square error value, the mean square error of the brightness temperature image is reduced, namely the observation error is reduced due to the improvement of the spatial resolution.
It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (7)
1. A microwave remote sensing image super-resolution reconstruction method based on SRCNN is characterized by comprising the following steps:
(1) Simulating a microwave remote sensing process, and forward generating a high-resolution microwave remote sensing image TB and a low-resolution microwave remote sensing image TA to form a data set;
(2) Preprocessing a data set to generate an SRCNN training set;
(3) Constructing a five-layer deep convolutional neural network based on the SRCNN training set;
(4) And inputting the low-resolution microwave remote sensing image to be processed into the constructed five-layer deep convolutional neural network, wherein the fourth-layer output of the deep convolutional neural network is the reconstructed high-resolution microwave remote sensing image.
2. The microwave remote sensing image super-resolution reconstruction method according to claim 1, wherein the step (1) specifically comprises: and generating a high-resolution microwave remote sensing image TB based on the forward evolution of the microwave radiation transmission RT mode, and inputting the high-resolution microwave remote sensing image TB into a radiometer forward observation model to simulate and calculate a low-resolution microwave remote sensing image TA actually observed.
3. The microwave remote sensing image super-resolution reconstruction method according to claim 1, wherein the step (2) specifically comprises:
(21) Performing pixel interpolation on the TA to obtain TC, so that the TC and the TB have the same pixel number;
(22) Respectively segmenting TB and TC to obtain image sets TD and TE, so that the image sizes in TD and TE are Z multiplied by Z;
(23) And (3) disordering the arrangement sequence of the images in the TD and the TE, selecting the first M images of the TD and the TE respectively as a sample set TDm and a label set TEm in a training data set, and selecting the rest images of the TD and the TE respectively as a sample set TDn and a label set TEN in a test data set.
4. The microwave remote sensing image super-resolution reconstruction method according to claim 3, wherein the step (2) further comprises:
(24) And generating the training data set and the testing data set into files in an HDF5 format.
5. The microwave remote sensing image super-resolution reconstruction method according to claim 1 or 3, wherein the step (3) specifically comprises:
(31) Designing the structure of the network by using a caffe framework, and setting a weight matrix of the ith layer of the deep convolutional neural network as W i (ii) a Offset is b i I =2,3,4; will TD n Middle image substitution x i Of TE n Middle image substitution y i I =1,2,3, · M; training the deep network, minimizing the output of the fifth layer,
wherein the content of the first and second substances,wherein f (x) = max (0, x); n represents the number of samples in each training set;
(32) Setting hyper-parameters of a five-layer deep convolution neural network;
(33) Optimizing the loss by adopting a random gradient descent method to ensure that a network weight matrix W i And offset b i And (5) converging, and finishing the construction of the five-layer deep convolution neural network.
6. The microwave remote sensing image super-resolution reconstruction method according to claim 5, wherein the hyper-parameters of the five-layer deep convolutional neural network in the step (32) specifically include: network iteration times, network operation modes and network parameter storage positions.
7. The microwave remote sensing image super-resolution reconstruction method according to claim 1, wherein before the low-resolution microwave remote sensing image is input into the trained five-layer deep convolutional neural network in the step (4), linear interpolation is required to be performed on the low-resolution microwave remote sensing image to make the number of pixels of the low-resolution microwave remote sensing image the same as that of the sample labels in the SRCNN training set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711362202.0A CN108009989A (en) | 2017-12-18 | 2017-12-18 | A kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711362202.0A CN108009989A (en) | 2017-12-18 | 2017-12-18 | A kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108009989A true CN108009989A (en) | 2018-05-08 |
Family
ID=62059731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711362202.0A Pending CN108009989A (en) | 2017-12-18 | 2017-12-18 | A kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108009989A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108335263A (en) * | 2018-01-25 | 2018-07-27 | 华中科技大学 | A kind of microwave remote sensing image super-resolution rebuilding method based on VDSR |
| CN108765343A (en) * | 2018-05-29 | 2018-11-06 | Oppo（重庆）智能科技有限公司 | Method, apparatus, terminal and the computer readable storage medium of image procossing |
| CN109117944A (en) * | 2018-08-03 | 2019-01-01 | 北京悦图遥感科技发展有限公司 | A kind of super resolution ratio reconstruction method and system of steamer target remote sensing image |
| CN109299163A (en) * | 2018-11-26 | 2019-02-01 | 武汉大学 | A kind of interpolation method and device of the precipitation data based on convolutional neural networks |
| CN109360152A (en) * | 2018-10-15 | 2019-02-19 | 天津大学 | 3 d medical images super resolution ratio reconstruction method based on dense convolutional neural networks |
| CN109461120A (en) * | 2018-09-19 | 2019-03-12 | 华中科技大学 | A kind of microwave remote sensing bright temperature image reconstructing method based on SRGAN |
| CN109584164A (en) * | 2018-12-18 | 2019-04-05 | 华中科技大学 | Medical image super-resolution three-dimensional rebuilding method based on bidimensional image transfer learning |
| CN110070486A (en) * | 2018-01-24 | 2019-07-30 | 杭州海康威视数字技术股份有限公司 | A kind of image processing method, device and electronic equipment |
| CN110838085A (en) * | 2018-08-15 | 2020-02-25 | Tcl集团股份有限公司 | Super-resolution reconstruction method and device for image and electronic equipment |
| CN112347945A (en) * | 2020-11-10 | 2021-02-09 | 北京航空航天大学 | Noise-containing remote sensing image enhancement method and system based on deep learning |
| US10949949B2 (en) | 2018-09-13 | 2021-03-16 | Samsung Electronics Co., Ltd. | Non-transitory computer-readable medium and method for monitoring a semiconductor fabrication process |
| CN112733394A (en) * | 2020-12-21 | 2021-04-30 | 国家卫星气象中心(国家空间天气监测预警中心) | Atmospheric parameter inversion method and device |
| CN117870872A (en) * | 2024-01-22 | 2024-04-12 | 华中科技大学 | Bright temperature image data set generation method based on satellite-borne multi-beam microwave radiometer |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106127684A (en) * | 2016-06-22 | 2016-11-16 | 中国科学院自动化研究所 | Image super-resolution Enhancement Method based on forward-backward recutrnce convolutional neural networks |
| US9665927B2 (en) * | 2015-06-03 | 2017-05-30 | Samsung Electronics Co., Ltd. | Method and apparatus of multi-frame super resolution robust to local and global motion |
| CN106960416A (en) * | 2017-03-20 | 2017-07-18 | 武汉大学 | A kind of video satellite compression image super-resolution method of content complexity self adaptation |
-
2017
- 2017-12-18 CN CN201711362202.0A patent/CN108009989A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9665927B2 (en) * | 2015-06-03 | 2017-05-30 | Samsung Electronics Co., Ltd. | Method and apparatus of multi-frame super resolution robust to local and global motion |
| CN106127684A (en) * | 2016-06-22 | 2016-11-16 | 中国科学院自动化研究所 | Image super-resolution Enhancement Method based on forward-backward recutrnce convolutional neural networks |
| CN106960416A (en) * | 2017-03-20 | 2017-07-18 | 武汉大学 | A kind of video satellite compression image super-resolution method of content complexity self adaptation |
Non-Patent Citations (4)
| Title |
|---|
| LI GONGWEI ET AL.: "performance evaluation of raditive transfer models of satellite atmospheric microwave sounding for data assimilation", 《2016 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM (IGARSS)》 * |
| 刘乐等: "RTTOV 快速辐射传输模式应用研究进展", 《安徽农业科学》 * |
| 卢昱璇: "基于卷积神经网络的图像超分辨率重建及其视觉改进", 《中国优秀硕士学位论文全文数据库》 * |
| 陈冉冉: "微波辐射计全链路仿真软件接口设计与实现", 《万方数据库》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110070486A (en) * | 2018-01-24 | 2019-07-30 | 杭州海康威视数字技术股份有限公司 | A kind of image processing method, device and electronic equipment |
| CN108335263A (en) * | 2018-01-25 | 2018-07-27 | 华中科技大学 | A kind of microwave remote sensing image super-resolution rebuilding method based on VDSR |
| CN108765343A (en) * | 2018-05-29 | 2018-11-06 | Oppo（重庆）智能科技有限公司 | Method, apparatus, terminal and the computer readable storage medium of image procossing |
| CN108765343B (en) * | 2018-05-29 | 2021-07-20 | Oppo（重庆）智能科技有限公司 | Image processing method, device, terminal and computer readable storage medium |
| CN109117944A (en) * | 2018-08-03 | 2019-01-01 | 北京悦图遥感科技发展有限公司 | A kind of super resolution ratio reconstruction method and system of steamer target remote sensing image |
| CN109117944B (en) * | 2018-08-03 | 2021-01-15 | 北京悦图数据科技发展有限公司 | Super-resolution reconstruction method and system for ship target remote sensing image |
| CN110838085A (en) * | 2018-08-15 | 2020-02-25 | Tcl集团股份有限公司 | Super-resolution reconstruction method and device for image and electronic equipment |
| CN110838085B (en) * | 2018-08-15 | 2021-09-24 | Tcl科技集团股份有限公司 | Super-resolution reconstruction method and device for image and electronic equipment |
| US10949949B2 (en) | 2018-09-13 | 2021-03-16 | Samsung Electronics Co., Ltd. | Non-transitory computer-readable medium and method for monitoring a semiconductor fabrication process |
| CN109461120A (en) * | 2018-09-19 | 2019-03-12 | 华中科技大学 | A kind of microwave remote sensing bright temperature image reconstructing method based on SRGAN |
| CN109360152A (en) * | 2018-10-15 | 2019-02-19 | 天津大学 | 3 d medical images super resolution ratio reconstruction method based on dense convolutional neural networks |
| CN109299163B (en) * | 2018-11-26 | 2020-07-24 | 武汉大学 | Rainfall data interpolation method and device based on convolutional neural network |
| CN109299163A (en) * | 2018-11-26 | 2019-02-01 | 武汉大学 | A kind of interpolation method and device of the precipitation data based on convolutional neural networks |
| CN109584164A (en) * | 2018-12-18 | 2019-04-05 | 华中科技大学 | Medical image super-resolution three-dimensional rebuilding method based on bidimensional image transfer learning |
| CN109584164B (en) * | 2018-12-18 | 2023-05-26 | 华中科技大学 | Medical image super-resolution three-dimensional reconstruction method based on two-dimensional image transfer learning |
| CN112347945A (en) * | 2020-11-10 | 2021-02-09 | 北京航空航天大学 | Noise-containing remote sensing image enhancement method and system based on deep learning |
| CN112347945B (en) * | 2020-11-10 | 2023-01-17 | 北京航空航天大学 | Noise-containing remote sensing image enhancement method and system based on deep learning |
| CN112733394A (en) * | 2020-12-21 | 2021-04-30 | 国家卫星气象中心(国家空间天气监测预警中心) | Atmospheric parameter inversion method and device |
| CN117870872A (en) * | 2024-01-22 | 2024-04-12 | 华中科技大学 | Bright temperature image data set generation method based on satellite-borne multi-beam microwave radiometer |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108009989A (en) | A kind of microwave remote sensing image super-resolution rebuilding method based on SRCNN | |
| Guo et al. | Lidar boosts 3D ecological observations and modelings: A review and perspective | |
| CN108875905B (en) | Direct inversion method for visibility function of atmospheric temperature and humidity profile | |
| CN109063569B (en) | Semantic level change detection method based on remote sensing image | |
| Miyoshi et al. | “Big data assimilation” toward post-petascale severe weather prediction: An overview and progress | |
| Malone et al. | Mapping continuous depth functions of soil carbon storage and available water capacity | |
| CN108445464A (en) | Satellite radar inverting fusion methods of the NRIET based on machine learning | |
| CN109725316B (en) | One-dimensional synthetic aperture microwave radiometer-based sea surface temperature physical inversion method | |
| Song et al. | An improved surface soil moisture downscaling approach over cloudy areas based on geographically weighted regression | |
| CN102073039B (en) | Thermal infrared hyperspectral emissivity simulation method and system | |
| CN108154193B (en) | Long-time sequence precipitation data downscaling method | |
| CN113239830A (en) | Remote sensing image cloud detection method based on full-scale feature fusion | |
| CN109033543A (en) | A kind of heterogeneous area's vegetation fraction estimation method, device and equipment of earth's surface | |
| CN104268869A (en) | Automatic multilevel-resolution remote-sensing image registration method based on particle swarm optimization | |
| CN108335263A (en) | A kind of microwave remote sensing image super-resolution rebuilding method based on VDSR | |
| CN112668615B (en) | Satellite cloud picture prediction method based on depth cross-scale extrapolation fusion | |
| CN116776651B (en) | Method and device for measuring and calculating surface evapotranspiration, electronic equipment and storage medium | |
| CN110969609B (en) | Microwave remote sensing bright temperature image cloud detection method | |
| CN111273376B (en) | Downscaling sea surface net radiation determination method, system, equipment and storage medium | |
| CN109344769A (en) | A kind of photovoltaic plant detection method and system based on remote sensing image | |
| CN109919990B (en) | Forest height prediction method by using depth perception network and parallax remote sensing image | |
| CN110705010A (en) | Remote sensing-based next-day night surface heat island simulation method | |
| CN115877345A (en) | Method and device for supplementing missing data of wind profile radar | |
| CN115222837A (en) | True color cloud picture generation method and device, electronic equipment and storage medium | |
| CN114324410A (en) | Multi-terrain microwave remote sensing soil humidity downscaling method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180508 |
|
| RJ01 | Rejection of invention patent application after publication |