CN113962897B - Modulation transfer function compensation method and device based on sequence remote sensing image - Google Patents
Modulation transfer function compensation method and device based on sequence remote sensing image Download PDFInfo
- Publication number
- CN113962897B CN113962897B CN202111289811.4A CN202111289811A CN113962897B CN 113962897 B CN113962897 B CN 113962897B CN 202111289811 A CN202111289811 A CN 202111289811A CN 113962897 B CN113962897 B CN 113962897B
- Authority
- CN
- China
- Prior art keywords
- image
- mtf
- sequence
- result
- compensation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 81
- 238000012546 transfer Methods 0.000 title claims abstract description 42
- 230000009466 transformation Effects 0.000 claims abstract description 62
- 238000003384 imaging method Methods 0.000 claims abstract description 32
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 28
- 238000004364 calculation method Methods 0.000 claims abstract description 20
- 238000005457 optimization Methods 0.000 claims abstract description 12
- 238000006731 degradation reaction Methods 0.000 claims description 79
- 230000015556 catabolic process Effects 0.000 claims description 63
- 239000011159 matrix material Substances 0.000 claims description 54
- 230000006870 function Effects 0.000 claims description 48
- 230000008569 process Effects 0.000 claims description 27
- 238000004590 computer program Methods 0.000 claims description 9
- 230000014509 gene expression Effects 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 230000005855 radiation Effects 0.000 claims description 5
- 239000000654 additive Substances 0.000 claims description 3
- 230000000996 additive effect Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 230000007547 defect Effects 0.000 abstract description 9
- 230000006872 improvement Effects 0.000 abstract description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000003287 optical effect Effects 0.000 description 5
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000005259 measurement Methods 0.000 description 3
- 239000000203 mixture Substances 0.000 description 3
- 230000002829 reductive effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 238000005315 distribution function Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000002945 steepest descent method Methods 0.000 description 2
- 230000001360 synchronised effect Effects 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- YZCKVEUIGOORGS-OUBTZVSYSA-N Deuterium Chemical compound [2H] YZCKVEUIGOORGS-OUBTZVSYSA-N 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000002238 attenuated effect Effects 0.000 description 1
- 229910002056 binary alloy Inorganic materials 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000265 homogenisation Methods 0.000 description 1
- 230000000670 limiting effect Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012634 optical imaging Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000013139 quantization Methods 0.000 description 1
- 238000011084 recovery Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Images
Classifications
-
- G06T5/70—
-
- G06T5/73—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/337—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving reference images or patches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/35—Determination of transform parameters for the alignment of images, i.e. image registration using statistical methods
-
- 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
- G06T2207/10016—Video; Image sequence
-
- 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
- G06T2207/20076—Probabilistic image processing
Abstract
The invention relates to a modulation transfer function compensation method and a modulation transfer function compensation device based on a sequence remote sensing image. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the overall improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
Description
Technical Field
The invention relates to the technical field, in particular to a modulation transfer function compensation method and device based on a sequence remote sensing image.
Background
The geostationary orbit optical satellite adopts a staring imaging mode, and can realize minute-level repeated observation on global hot spot areas. However, in the process of imaging from a ground object to an image plane, an optical imaging system of a geostationary orbit is affected by the links such as orbit height, attitude stability, sensor vibration, atmospheric scattering and absorption, atmospheric turbulence, motion and the like, and an MTF (Modulation Transfer Function) of the system is attenuated, so that medium-high frequency information of scene radiation is lost, and an image is blurred. But because the orbit height of the earth's stationary orbit is high, the quality of satellite imaging is not greatly affected by atmospheric turbulence. Some analyses on the influence of the satellite imaging quality show that when the attitude stability of the satellite platform is 5 multiplied by 10 < -4 > (° s), the MTF is reduced by 6 percent, the influence of the vibration of the satellite platform on the MTF is the largest, and the MTF can be reduced by nearly 30 percent when the satellite platform is serious. Therefore, Modulation Transfer Function Compensation (MTFC) is required for the imaging process to meet the basic requirements of the imaging of the remote sensing optical system.
The traditional modulation transfer function compensation mode comprises a frequency domain inverse filtering method based on Fourier transform, a wiener filtering method, a spatial domain iteration method, an information entropy maximization method and the like. The modulation transfer function compensation mode is that firstly, the MTF or PSF (point spread function) of the in-orbit remote sensing imaging system is measured by using the feature of the ground object, and then the image is processed by using the digital image post-processing technology on the ground, so that the MTF compensation is realized, and the purpose of improving the quality of the remote sensing image is achieved. However, because the data volume of the remote sensing image is large, MTF compensation calculation is performed on each frame of image, a large amount of calculation requirements are generated, and the application of the compensation technology in the field of actual processing is limited. On the other hand, complete and accurate MTF or PSF of the in-orbit remote sensing imaging system is difficult to obtain, so that new high-frequency noise is introduced into a reconstructed image, and the higher the MTF compensation is, the larger the introduced noise is, and the overall quality of the image and the dynamic range of the image are reduced. Meanwhile, preprocessing such as homogenization radiation correction and dynamic range modulation and selection of quantization bits, compression algorithm and compression ratio have great influence on realization of image MTF compensation. It can be seen that the above disadvantages exist in the conventional modulation transfer function compensation method.
Disclosure of Invention
Therefore, it is necessary to provide a modulation transfer function compensation method and device based on a sequence remote sensing image, aiming at the defects of the conventional modulation transfer function compensation method.
A modulation transfer function compensation method based on sequence remote sensing images comprises the following steps:
acquiring multi-frame sequence satellite images, and constructing an MTF degradation model according to the degradation process of the sequence satellite images; wherein the degradation process is used for representing the degradation from the high-resolution image to the low-resolution image;
obtaining a geometric deformation distortion matrix of the MTF degradation model after interframe registration through interframe registration of multi-frame sequence satellite images;
performing approximate solution on the MTF degradation model according to the geometric deformation distortion matrix after the reference image and the interframe registration; the reference image is a frame sequence satellite image selected in the interframe registration process;
performing mathematical transformation on the approximate solution result, and establishing a prior probability model according to the mathematical transformation result;
and obtaining a final MTF compensation result image according to iterative calculation of the prior probability model.
According to the modulation transfer function compensation method based on the sequence remote sensing image, an MTF degradation model is established through multi-frame sequence satellite images, a geometric deformation distortion matrix of the MTF degradation model is obtained through combination of inter-frame registration, and the MTF degradation model is approximately solved. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the overall improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
In one embodiment, the MTF degradation model is as follows:
g k ＝H k M k z+n k
wherein H k Fuzzy matrices, M, representing MTF and radiation distortion of imaging systems of satellite imagery of k-th frame sequence k Geometric distortion matrix, n, representing the satellite imagery of the k-th frame sequence k Additive noise, g, representing zero mean gaussian distribution contained in the k-th frame sequence satellite imagery k The image sequence of the k frame is represented by a satellite image, and z is represented by a high-resolution image of the multi-frame satellite image after MTF compensation.
In one embodiment, the process of obtaining the geometric deformation distortion matrix of the MTF degradation model through interframe registration of multi-frame sequence satellite images includes the steps of:
extracting feature points of the multi-frame sequence satellite images;
calculating a feature descriptor according to the extracted feature points to obtain a result image of coarse registration;
constructing a pyramid according to the result image and the reference image, and calculating the mutual correlation information of the result image and the reference image; the reference image is one frame of image in the multi-frame sequence satellite images;
translating pixels between local image blocks of the fine tuning result image and the reference image, and executing global loop iteration;
and executing image transformation according to the transformation matrix of the global loop iteration, finishing the precise registration of the multi-frame sequence satellite image, and determining a geometric deformation distortion matrix in the corresponding MTF degradation model.
In one embodiment, the process of performing a mathematical transformation on the result of the approximation solution and building a prior probability model based on the result of the mathematical transformation includes the steps of:
performing orthogonal wavelet transformation on the approximate solution result to obtain a mathematical transformation result;
and introducing the wavelet domain uHMT model into signal processing of a mathematical transformation result, and constructing a prior probability model.
In one embodiment, before the process of constructing the prior probability model according to the wavelet domain uHMT model, the method further includes the steps of:
and performing mathematical estimation on the vector expression in the mathematical transformation result.
In one embodiment, the mathematical estimate comprises a MAP estimate or a MLE estimate.
In one embodiment, the process of obtaining a final MTF compensation result image according to iterative computation of a prior probability model includes the steps of:
and performing iterative computation in preset iterative counting on the wavelet coefficients in the prior probability model through a first-order optimization algorithm to obtain a final MTF compensation result image.
A modulation transfer function compensation device based on sequence remote sensing images comprises:
the image acquisition module is used for acquiring multi-frame sequence satellite images and constructing an MTF degradation model according to the degradation process of the sequence satellite images; wherein the degradation process is used for representing the degradation from the high-resolution image to the low-resolution image;
the interframe registration module is used for acquiring a geometric deformation distortion matrix of the MTF degradation model after interframe registration through interframe registration of multi-frame sequence satellite images;
the approximate solving module is used for carrying out approximate solving on the MTF degradation model according to the geometric deformation distortion matrix after the reference image and the interframe registration; the reference image is a frame sequence satellite image selected in the interframe registration process;
the model establishing module is used for executing mathematical transformation on the approximate solution result and establishing a prior probability model according to the mathematical transformation result;
and the result compensation module is used for obtaining a final MTF compensation result image according to iterative calculation of the prior probability model.
According to the modulation transfer function compensation device based on the sequence remote sensing image, the MTF degradation model is established through the multi-frame sequence satellite image, the geometric deformation distortion matrix of the MTF degradation model is obtained by combining inter-frame registration, and the MTF degradation model is approximately solved. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the integral improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
A computer storage medium, on which computer instructions are stored, and the computer instructions, when executed by a processor, implement the modulation transfer function compensation method based on the sequence remote sensing image according to any of the above embodiments.
The computer storage medium establishes the MTF degradation model through multi-frame sequence satellite images, calculates the geometric deformation distortion matrix of the MTF degradation model by combining inter-frame registration, and performs approximate solution on the MTF degradation model. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the integral improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
A computer device includes a memory, a processor and a computer program stored in the memory and executable on the processor, and when the processor executes the program, the modulation transfer function compensation method based on the sequence remote sensing image according to any of the embodiments is implemented.
The computer device establishes the MTF degradation model through the multi-frame sequence satellite images, calculates the geometric deformation distortion matrix of the MTF degradation model by combining inter-frame registration, and performs approximate solution on the MTF degradation model. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the overall improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
Drawings
Fig. 1 is a flowchart of a modulation transfer function compensation method based on a sequence remote sensing image according to an embodiment;
FIG. 2 is a schematic diagram illustrating a degradation process from a high resolution image to a low resolution image;
fig. 3 is a flowchart of a modulation transfer function compensation method based on a sequence remote sensing image according to another embodiment;
FIG. 4 is a flowchart of sequential satellite image registration according to an embodiment;
FIG. 5 is a flow chart of MTF compensation according to an exemplary embodiment;
FIG. 6 is a block diagram of a modulation transfer function compensation apparatus based on a sequence remote sensing image according to an embodiment;
FIG. 7 is a schematic diagram of an internal structure of a computer according to an embodiment.
Detailed Description
For better understanding of the objects, technical solutions and technical effects of the present invention, the present invention will be further explained with reference to the accompanying drawings and examples. It is to be noted that the following examples are provided only for illustrating the present invention and are not intended to limit the present invention.
According to the embodiment, sequential frame remote sensing image data acquired by an earth stationary orbit satellite is adopted, firstly, the MTF degradation process from a high-resolution image to a low-resolution image is analyzed, an MTF degradation model is constructed, then a motion compensation matrix in the MTF degradation model is obtained through a method based on joint registration, MTF compensation research of an image is carried out through a maximum posterior probability method based on a universal hidden Markov tree model, the imaging quality of the image is improved, and the MTF of the whole imaging system is improved.
The embodiment of the invention provides a modulation transfer function compensation method based on a sequence remote sensing image.
Fig. 1 is a flowchart of a modulation transfer function compensation method based on a sequence remote sensing image according to an embodiment, and as shown in fig. 1, the modulation transfer function compensation method based on the sequence remote sensing image according to an embodiment includes steps S100 to S104:
s100, acquiring multi-frame sequence satellite images, and constructing an MTF (modulation transfer function) degradation model according to a degradation process of the sequence satellite images; wherein the degradation process is used for representing the degradation from the high-resolution image to the low-resolution image;
s101, obtaining a geometric deformation distortion matrix of the MTF degradation model after interframe registration through interframe registration of multi-frame sequence satellite images;
s102, performing approximate solution on the MTF degradation model according to the geometric deformation distortion matrix after the reference image and the interframe registration; the reference image is a frame sequence satellite image selected in the interframe registration process;
s103, performing mathematical transformation on the approximate solution result, and establishing a prior probability model according to the mathematical transformation result;
and S104, obtaining a final MTF compensation result image according to iterative calculation of the prior probability model.
The multi-frame sequence satellite images are obtained in a sequence frame mode. In the acquisition of the sequence frame image, the problems of platform jitter of a satellite detector, diffraction, defocusing and the like of an (optical) imaging system can be solved, so that the problems of distortion, geometric deformation, motion displacement, optical blurring, motion blurring, sensor blurring and the like of the sequence frame low-resolution image exist.
In one embodiment, the sequence of satellite imagery includes short duration continuous imaging of geostationary orbit satellites to acquire low resolution satellite data of the same region. Multi-frame sequence satellite image including low-resolution sequence frame satellite data g 1 ,g 2 ,g 3 ,...,g k 。
In order to solve the problems existing in the process of acquiring the multi-frame sequence satellite images, an MTF degradation model is established by analyzing the process of high-resolution degradation in the sequence frame images. Fig. 2 is a schematic diagram of a degradation process from a high resolution image to a low resolution image, and as shown in fig. 2, the high resolution image undergoes geometric deformation, blurring and noise in the degradation process to the low resolution image, and a matrix is constructed for each problem process. In one embodiment, as shown in FIG. 2, the MTF degradation model is given by:
g k ＝H k M k z+n k
wherein H k Blurring matrix, M, representing MTF and radiation distortion of an imaging system of a k-th frame sequential satellite imagery k Geometric distortion matrix, n, representing the satellite imagery of the k-th frame sequence k Additive noise, g, representing zero mean gaussian distribution contained in the k-th frame sequence satellite imagery k The image processing method comprises the steps of representing a kth frame sequence satellite image, and representing a high-resolution image of the multi-frame sequence satellite image after MTF compensation.
It should be noted that different satellite imaging systems determine the final high-resolution image according to the multi-frame sequence low-resolution images by considering different influence factors, and the MTF degradation model in the above embodiment is only an embodiment, and does not represent a limitation on all MTF degradation processes. According to different MTF degradation analyses, corresponding MTF degradation models can be determined.
And according to the constructed MTF degradation model, solving a geometric deformation distortion matrix in the MTF degradation process in the sequence satellite images through interframe registration of the multi-frame sequence satellite images. And determining the vector change between the satellite images with different frame sequences by comparing the satellite images with different frame sequences which are registered among frames, and determining a geometric deformation distortion matrix by registering the convergence vector change.
In one embodiment, fig. 3 is a flowchart of a modulation transfer function compensation method based on a remote sensing sequence image according to another embodiment, and as shown in fig. 3, a process of obtaining a geometric deformation distortion matrix of an MTF degradation model through inter-frame registration of a plurality of frame sequence satellite images in step S101 includes steps S200 to S204:
s200, extracting feature points of multi-frame sequence satellite images;
in order to better explain steps S200 to S204, steps S200 to S204 are explained below with a specific application example. Fig. 4 is a flowchart illustrating registration of sequential satellite images according to a specific application example, and as shown in fig. 4, feature points of multiple sequential satellite images can be extracted through convolution calculation. In one embodiment, feature points in each frame sequence satellite image are extracted by feature point extraction through an ORB-based feature point registration algorithm. In one embodiment, feature point extraction is performed through a feature point registration algorithm of SIFT, and feature points in each frame of satellite image are extracted. The registration algorithms of SIFT and ORB have better robustness, and the arithmetic speed of the algorithms is higher.
S201, calculating a feature descriptor according to the extracted feature points to obtain a result image of coarse registration;
as shown in fig. 4, the feature descriptors are calculated according to the feature points, and the key feature points with better matching between different frame-sequential satellite images are extracted according to the feature descriptors to calculate the homography matrix between each frame-sequential satellite image, so as to obtain the result image of coarse registration.
In one embodiment, the similarity measurement is performed by finding the Hamming Distance (Hamming Distance) between the feature descriptors to extract the key feature points with better matching.
In one embodiment, a homography matrix between multi-frame sequence satellite image sequences is solved through a RANSAC (Random Sample Consensus) algorithm, and a result image after coarse registration is obtained.
S202, constructing a pyramid according to the result image and the reference image, and calculating the mutual correlation information of the result image and the reference image; the reference image is one frame of image in the multi-frame sequence satellite images;
as shown in fig. 4, one frame of the multi-frame sequence satellite images is selected as a reference image, and the other sequence satellite images are to-be-registered images. And constructing a pyramid with the reference image according to the result image corresponding to the image to be registered, and calculating the mutual correlation information of the result image and the reference image.
In one embodiment, a gray level registration method is used for establishing similarity measurement between the reference image and the result image according to gray level information of the reference image and the result image, and a transformation model parameter value with the maximum similarity measurement is determined to serve as cross-correlation information.
S203, translating pixels between local image blocks of the fine tuning result image and the reference image, and executing global loop iteration;
and S204, executing image transformation according to the transformation matrix of the global loop iteration, finishing the precise registration image of the multi-frame sequence satellite image, and determining the corresponding geometric deformation distortion matrix in the MTF degradation model.
After a geometric transformation matrix M between the reference frame and the frame to be registered is obtained, the precise registration of the multi-frame satellite images of the sequence can be completed through the transformation matrix, and the geometric deformation distortion matrix Mk in the corresponding MTF degradation model can be calculated by using the solved transformation matrix and the characteristics of the images of the sequence frames.
Because the image registration can provide inter-frame sub-pixel displacement information for MTF compensation, the same scenes in a plurality of obtained sequence frame images are aligned and are in the same coordinate system. The accurate solution of the inter-frame geometric transformation matrix M determines the correct position of non-redundant information, and directly influences the quality of a reconstruction result. In the process, rough registration is carried out on each frame of image and a reference image to obtain approximate position information of the two images, and then refined configuration is carried out on the image subjected to rough registration and the reference image to obtain an accurate transformation matrix, so that the image precision after MTFC compensation is improved.
As shown in fig. 4, a global loop iteration is performed after the pixel translation hinting. In one embodiment, the resulting image is g i The reference image is g r For explanation. Performing translation fine adjustment on image pixel positions between local blocks of the image by adopting E pix (m) error function representing rigid global registration, introducing a smooth constraint function
As constraints for fine tuning, local affine transformations and intensity variations are handled. The error function between the pixels of the sequential satellite images is expressed as:
make E iterate continuously pix And (m) minimizing the value, finally finishing refined registration, and determining a geometric deformation distortion matrix in the corresponding MTF degradation model. Based on the above, an accurate transformation matrix is obtained through the steps, so that the displacement difference after the registration between the images reaches the sub-pixel level, and the calculated MTF-compensated image quality is higher than that of the conventional MTF compensationThe image quality of the method is higher.
And performing approximate solution estimation on the MTF degradation model in the step S100 according to the determined geometric deformation distortion matrix after refined registration. And replacing corresponding parameters of the original MTF degradation model by the determined geometric deformation distortion matrix after refined registration, and performing approximate estimation and solution on other parameters.
In one embodiment, the MTF degradation model is approximated in step S102 by the following equation:
g r ＝H' r M' r z+n' r
wherein, M' r A unit matrix is represented, and can be determined by approximate estimation of the geometric deformation distortion matrix determined in step S101; fuzzy matrix H' r Can pass through H k Solving and determining approximate estimation; n' r Represent the errors due to the blur matrix and the geometric distortion warp matrix, and assume n' r Obeying an independent, identically distributed gaussian distribution with a mean value of zero.
After the approximate solution estimation is completed, a prior probability model is established by performing mathematical transformation on the approximate solution result.
In one embodiment, the mathematical transform comprises a fourier transform or an orthogonal wavelet transform.
In one embodiment, as shown in fig. 3, the process of performing mathematical transformation on the result of the approximation solution in step S103 and establishing a prior probability model according to the result of the mathematical transformation includes steps S300 and S301:
s300, performing orthogonal wavelet transformation on the approximate solution result to obtain a mathematical transformation result;
the traditional MTF compensation method can improve the MTF of the system and simultaneously blur the details of the image. The orthogonal wavelet transform has good time-frequency localization characteristics, the scale function of the orthogonal wavelet transform is changed according to a binary system, the characteristics of the orthogonal wavelet transform expressed in the frequency resolution of a high frequency band and the time resolution of a low frequency band are different, signals can be better depicted, the high frequency part of the signals is randomly finely divided, the signals and noise in each scale are maximally separated, the MTF of the images can be improved, the image noise is removed, meanwhile, the image details can be kept, and the optimal recovery of the original images is obtained. In addition, the orthogonal wavelet transform has higher algorithm convergence rate in the calculation process, reduces the calculation complexity of the algorithm and improves the calculation capability of the algorithm. Based on the method, the result of mathematical transformation is obtained by adopting the method of orthogonal wavelet transformation.
To better explain the process of establishing and calculating the prior probability model, the following is explained with a specific application example, fig. 5 is a flowchart of MTF compensation of a specific application example, as shown in fig. 5, and orthogonal wavelet transform is performed on the approximate solution result, as follows:
wherein the content of the first and second substances,
each represents g r 、z、n r Vector expressions, matrices, arranged in lexicographic order in the wavelet domain
And
are respectively a matrix H k And M k The wavelet of (2).
As a preferred embodiment, the sequential satellite imagery is fully decomposed by Daubechies wavelets.
In one embodiment, as shown in fig. 3, before the process of constructing the prior probability model according to the wavelet domain uHMT model in step S301, step S400 is further included:
and S400, performing mathematical estimation on the vector expression in the mathematical transformation result.
In one embodiment, the mathematical estimate comprises a MAP estimate or a MLE estimate.
As a preferred embodiment, the MA is used for mathematical estimation, as shown in FIG. 5And P estimation. By using
To represent
Expression in the wavelet domain, estimated and assumed using MAP estimation, then
S301, introducing the wavelet domain uHMT model into signal processing of a mathematical transformation result, and constructing a prior probability model.
The wavelet domain is replaced in a space or frequency domain, because the spectrogram of the wavelet coefficient can be described by a Gaussian mixture spectrogram, the probability distribution function of the wavelet coefficient can be well represented on the wavelet domain through the dependency relationship between the Gaussian mixture probability distribution function and the coarse-scale wavelet coefficient, and the HMT can well describe the wavelet coefficients of upper and lower scales on the wavelet domain. Wavelet transformation of real images is often sparse, most wavelet coefficients are small, and real images have a plurality of smooth regions separated by edges. The uHMT model of the wavelet domain can capture key characteristics of the real data wavelet coefficient joint probability density. By utilizing the characteristics of the wavelet coefficient, the wavelet coefficient of the real image is exponentially decreased in scale, and the durability of the wavelet coefficient is exponentially enhanced in smaller scale. Meanwhile, the wavelet domain uHMT model does not need any type of training, is relatively simple, can be estimated by utilizing a sequence frame remote sensing image, and retains almost all key image structures and detail information modeled by complete HMT, and can effectively remove noise contained in the image to obtain a better effect.
As shown in fig. 5, a wavelet domain uHMT model is introduced into signal denoising and signal estimation, and an energy function is constructed to determine a prior probability model. Assuming that the marginal probability density functions of the wavelet coefficients are conditionally independent, then
Can be expressed as:
wherein the content of the first and second substances,
and
representing wavelet coefficients
The probabilities of belonging to the hidden states 1 and 2,
and
subject to gaussian density, it is shown that, in the case of the crypto-state,
the probability of occurrence. N is a radical of hydrogen 2 Expressed as the number of pixels in the original sequence satellite image z.
The marginal probability density function of the wavelet coefficient is regarded as gaussian mixture, and is solved by using a mode that a single gaussian distribution is approximate to mixed gaussian distribution, so that the formula of the minimized parameter expression can be expressed as the following formula:
where α represents a parameter of the noise variance that describes the error caused by the erroneous warping and blurring matrices. Alpha may balance the contribution from the prior probability model with all low resolution images. The denominator K is introduced in order to eliminate the influence of the difference in the number of low-resolution images in different cases. P is a radical of
Of (N) 2 ×N 2 ) Diagonal with the number of pixels. And (4) matrix.
P is in [0,1 ]]Normalized with respect to each other.
Where the parameter a of the noise variance can be used to describe the error caused by the erroneous warping and blurring matrices. The contribution from the prior probability models of all the low resolution images can be balanced regardless of how many frames of low resolution images are used. Therefore, the above-described embodiment does not have to consider the number of frames used for low-resolution images in the algorithm, and it is also easy to select the value of the parameter α of the noise variance through experiments and experience.
Based on this, in one embodiment, as shown in fig. 3, the process of obtaining a final MTF compensation result image according to the iterative computation of the prior probability model in step S104 includes step S500:
and S500, performing iterative computation in preset iterative counting on the wavelet coefficients in the prior probability model through a first-order optimization algorithm to obtain a final MTF compensation result image.
In one embodiment, the first order optimization algorithm uses the steepest descent method. As shown in FIG. 5, an algorithm iteration loop parameter r is set iter When the wavelet coefficient of the original image is 0, the wavelet coefficient is estimated by the steepest descent method, and calculation is carried out
Where d is the falling gradient and a is the step size. The algorithm iterates cyclically until
T is a set threshold value, and therefore the preset iteration count is completed. And updating the MTF compensation result obtained last time in each iteration until the iteration count exceeds the maximum value, finishing the operation of the algorithm and obtaining a final MTF compensation result image.
The steepest descent algorithm is to perform one-dimensional search starting from a certain point along the negative direction of the gradient (minimum point), and select a proper optimal step length to reduce the value of the target function to the maximum extent. The steepest descent algorithm has less requirement on the initial point, when the problem of the embodiment of the invention is processed, the optimal processing result can be obtained only by considering the property of the objective function at a certain point, and meanwhile, a stable and approximate estimation value can be provided by adopting the steepest descent algorithm.
In the modulation transfer function compensation method based on the sequence remote sensing image according to any of the embodiments, the MTF degradation model is established through the multi-frame sequence satellite image, the geometric deformation distortion matrix of the MTF degradation model is solved by combining inter-frame registration, and the MTF degradation model is approximately solved. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the overall improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
The embodiment of the invention also provides a modulation transfer function compensation device based on the sequence remote sensing image.
Fig. 6 is a block diagram of a modulation transfer function compensation apparatus based on a sequence remote sensing image according to an embodiment, and as shown in fig. 6, the modulation transfer function compensation apparatus based on a sequence remote sensing image according to an embodiment includes an image acquisition module 100, an inter-frame registration module 101, an approximation solution module 102, a model establishment module 103, and a result compensation module 104:
the image acquisition module 100 is configured to acquire a multi-frame sequence satellite image and construct an MTF degradation model according to a degradation process of the sequence satellite image; wherein the degradation process is used for representing the degradation from the high-resolution image to the low-resolution image;
the interframe registration module 101 is configured to obtain a geometric deformation distortion matrix of the MTF degradation model after interframe registration through interframe registration of a multi-frame sequence satellite image;
the approximate solving module 102 is configured to perform approximate solving on the MTF degradation model according to the geometric deformation distortion matrix after the reference image and the inter-frame registration; the reference image is a frame sequence satellite image selected in the interframe registration process;
the model establishing module 103 is used for performing mathematical transformation on the approximate solution result and establishing a prior probability model according to the mathematical transformation result;
and the result compensation module 104 is configured to obtain a final MTF compensation result image according to iterative computation of the prior probability model.
According to the modulation transfer function compensation device based on the sequence remote sensing image, the MTF degradation model is established through the multi-frame sequence satellite image, the geometric deformation distortion matrix of the MTF degradation model is obtained by combining inter-frame registration, and the MTF degradation model is approximately solved. Further, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result so as to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the overall improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
The embodiment of the invention also provides a computer storage medium, on which computer instructions are stored, and when the instructions are executed by a processor, the method for compensating the modulation transfer function based on the sequence remote sensing image is implemented according to any one of the embodiments.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware related to instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, the computer program can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database or other medium used in the embodiments provided herein can include non-volatile and/or volatile memory. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), Rambus (Rambus) direct RAM (RDRAM), direct bused dynamic RAM (DRDRAM), and bused dynamic RAM (RDRAM).
Alternatively, the integrated unit of the present invention may be stored in a computer-readable storage medium if it is implemented in the form of a software functional module and sold or used as a separate product. Based on such understanding, the technical solutions of the embodiments of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a terminal, or a network device) to execute all or part of the methods of the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, a RAM, a ROM, a magnetic or optical disk, or various other media that can store program code.
Corresponding to the computer storage medium, in an embodiment, there is also provided a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement any one of the above-described embodiments of the method for modulation transfer function compensation based on sequential remote sensing images.
The computer device may be a terminal, and its internal structure diagram may be as shown in fig. 7. The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to realize a modulation transfer function compensation method based on the sequence remote sensing image. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, a key, a track ball or a touch pad arranged on a shell of the computer equipment, an external keyboard, a touch pad or a mouse and the like.
The computer equipment establishes the MTF degradation model through multi-frame sequence satellite images, calculates a geometric deformation distortion matrix of the MTF degradation model by combining inter-frame registration, and performs approximate solution on the MTF degradation model. Furthermore, mathematical transformation is carried out on the approximate solution result, a prior probability model is established according to the mathematical transformation result to reconstruct the image, algorithm optimization is carried out on the reconstructed image result through iterative calculation, and a final MTF compensation result image is obtained. Based on the method, on the basis of not changing the hardware environment of the imaging system of the sequence satellite images, the definition and the signal-to-noise ratio of the sequence satellite images are improved through the overall improvement after the improved MTF compensation, the noise is effectively controlled, the imaging quality of the sequence satellite images is improved, and the defects of the traditional MTF compensation mode are overcome.
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above examples only show some embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (8)
1. A modulation transfer function compensation method based on sequence remote sensing images is characterized by comprising the following steps:
acquiring multi-frame sequence satellite images, and constructing an MTF (modulation transfer function) degradation model according to the degradation process of the sequence satellite images; wherein the degradation process is used to characterize the degradation of the high resolution image to the low resolution image; the method comprises the following steps that a multi-frame sequence satellite image is obtained in a sequence frame mode;
obtaining a geometric deformation distortion matrix of the MTF degradation model after interframe registration through interframe registration of the multi-frame sequence satellite images; the inter-frame registration is based on feature point and gray level combination;
the process of obtaining the geometric deformation distortion matrix of the MTF degradation model through the interframe registration of the multi-frame sequence satellite images comprises the following steps:
extracting feature points of the multi-frame sequence satellite images;
calculating a feature descriptor according to the extracted feature points to obtain a result image of coarse registration;
constructing a pyramid according to the result image and the reference image, and calculating the mutual correlation information of the result image and the reference image; the reference image is one image in the multi-frame sequence satellite images;
translating and fine-tuning pixels between local image blocks of the result image and the reference image, and executing global loop iteration;
executing image transformation according to the transformation matrix of the global loop iteration, finishing the precise registration image of the multi-frame sequence satellite image, and determining a geometric deformation distortion matrix in a corresponding MTF degradation model;
according to the geometric deformation distortion matrix after the reference image and the interframe registration, the MTF degradation model is approximately solved; the reference image is a frame sequence satellite image selected in the interframe registration process;
performing mathematical transformation on the approximate solution result, and establishing a prior probability model according to the mathematical transformation result;
and obtaining a final MTF compensation result image according to iterative calculation of the prior probability model.
2. The method for compensating the modulation transfer function based on the sequence remote sensing image according to claim 1, wherein the MTF degradation model is as follows:
wherein the content of the first and second substances,H k is shown askThe MTF of the imaging system of the frame-sequential satellite imagery and the blur matrix of the radiation distortion,M k is shown askThe geometric distortion matrix of the frame sequence satellite imagery,n k denotes the firstkThe frame-sequential satellite imagery contains zero-mean gaussian distributed additive noise,g k is shown askThe sequence of frames is a sequence of satellite images,zand the high-resolution images of the multi-frame sequence satellite images after MTF compensation are shown.
3. The method for compensating the modulation transfer function based on the sequence remote sensing image according to claim 1 or 2, wherein the process of performing mathematical transformation on the approximate solution result and establishing a prior probability model according to the mathematical transformation result comprises the steps of:
performing orthogonal wavelet transformation on the approximate solution result to obtain a mathematical transformation result;
and introducing the wavelet domain uhm model into signal processing of the mathematical transformation result to construct the prior probability model.
4. The method for modulation transfer function compensation based on sequence remote sensing images as claimed in claim 3, wherein before the process of constructing the prior probability model according to the wavelet domain uhm model, the method further comprises the steps of:
and performing mathematical estimation on a vector expression in the mathematical transformation result.
5. The method of claim 4, wherein the mathematical estimate comprises a MAP estimate or a MLE estimate.
6. The modulation transfer function compensation method based on the sequence remote sensing image according to claim 3, wherein the process of obtaining a final MTF compensation result image according to iterative calculation of the prior probability model comprises the following steps:
and performing iterative computation in preset iterative counting on the wavelet coefficients in the prior probability model through a first-order optimization algorithm to obtain a final MTF compensation result image.
7. A computer storage medium having computer instructions stored thereon, wherein the computer instructions, when executed by a processor, implement the modulation transfer function compensation method according to any one of claims 1 to 6 based on the sequence remote sensing image.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the method for modulation transfer function compensation based on sequential remote sensing images according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111289811.4A CN113962897B (en) | 2021-11-02 | 2021-11-02 | Modulation transfer function compensation method and device based on sequence remote sensing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111289811.4A CN113962897B (en) | 2021-11-02 | 2021-11-02 | Modulation transfer function compensation method and device based on sequence remote sensing image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113962897A CN113962897A (en) | 2022-01-21 |
| CN113962897B true CN113962897B (en) | 2022-09-02 |
Family
ID=79469173
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111289811.4A Active CN113962897B (en) | 2021-11-02 | 2021-11-02 | Modulation transfer function compensation method and device based on sequence remote sensing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113962897B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114677275B (en) * | 2022-03-11 | 2023-03-24 | 自然资源部国土卫星遥感应用中心 | High-frequency repeated staring imaging space-based remote sensing load on-orbit testing method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111127325A (en) * | 2019-12-31 | 2020-05-08 | 珠海大横琴科技发展有限公司 | Satellite video super-resolution reconstruction method and system based on cyclic neural network |
| CN113128433A (en) * | 2021-04-26 | 2021-07-16 | 刘秀萍 | Video monitoring image enhancement method of color migration matching characteristics |
Family Cites Families (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100510773C (en) * | 2006-04-14 | 2009-07-08 | 武汉大学 | Single satellite remote sensing image small target super resolution ratio reconstruction method |
| US7616826B2 (en) * | 2006-07-28 | 2009-11-10 | Massachusetts Institute Of Technology | Removing camera shake from a single photograph using statistics of a natural image |
| WO2012041492A1 (en) * | 2010-09-28 | 2012-04-05 | MAX-PLANCK-Gesellschaft zur Förderung der Wissenschaften e.V. | Method and device for recovering a digital image from a sequence of observed digital images |
| CN103020898B (en) * | 2012-12-03 | 2018-01-05 | 北京理工大学 | Sequence iris image super resolution ratio reconstruction method |
| CN103020909B (en) * | 2012-12-06 | 2015-02-18 | 清华大学 | Single-image super-resolution method based on multi-scale structural self-similarity and compressive sensing |
| CN105469415A (en) * | 2015-12-28 | 2016-04-06 | 云南师范大学 | Multi-view remote sensing image fusion method |
| CN105550993B (en) * | 2016-01-18 | 2018-11-20 | 中国空间技术研究院 | Super resolution ratio reconstruction method based on multiple transform domain |
| CN106204438A (en) * | 2016-06-28 | 2016-12-07 | 北京航空航天大学 | The super-resolution rebuilding of lack sampling degraded image optimizes restoration methods |
| CN108090869B (en) * | 2017-11-27 | 2021-07-09 | 中国空间技术研究院 | On-satellite super-resolution reconstruction method based on area array CMOS optical camera |
| CN109035139B (en) * | 2018-06-28 | 2022-11-22 | 长光卫星技术股份有限公司 | High-resolution satellite image modulation transfer function compensation method |
| CN108961163A (en) * | 2018-06-28 | 2018-12-07 | 长光卫星技术有限公司 | A kind of high-resolution satellite image super-resolution reconstruction method |
| CN111986134B (en) * | 2020-08-26 | 2023-11-24 | 中国空间技术研究院 | Remote sensing imaging method and device for area-array camera |
| CN112261315B (en) * | 2020-09-07 | 2022-06-17 | 清华大学 | High-resolution calculation imaging system and method based on camera array aperture synthesis |
-
2021
- 2021-11-02 CN CN202111289811.4A patent/CN113962897B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111127325A (en) * | 2019-12-31 | 2020-05-08 | 珠海大横琴科技发展有限公司 | Satellite video super-resolution reconstruction method and system based on cyclic neural network |
| CN113128433A (en) * | 2021-04-26 | 2021-07-16 | 刘秀萍 | Video monitoring image enhancement method of color migration matching characteristics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113962897A (en) | 2022-01-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | Image super-resolution based on structure-modulated sparse representation | |
| US9692939B2 (en) | Device, system, and method of blind deblurring and blind super-resolution utilizing internal patch recurrence | |
| US9230303B2 (en) | Multi-frame super-resolution of image sequence with arbitrary motion patterns | |
| CN110675347B (en) | Image blind restoration method based on group sparse representation | |
| CN107133923B (en) | Fuzzy image non-blind deblurring method based on adaptive gradient sparse model | |
| US10043242B2 (en) | Method and apparatus for synthesis of higher resolution images | |
| CN112241976A (en) | Method and device for training model | |
| US11720999B2 (en) | Method, device and non-transitory computer-readable storage medium for increasing the resolution and dynamic range of a sequence of respective top view images of a same terrestrial location | |
| CN112419380B (en) | Cloud mask-based high-precision registration method for stationary orbit satellite sequence images | |
| CN113962897B (en) | Modulation transfer function compensation method and device based on sequence remote sensing image | |
| Karwowska et al. | Using super-resolution algorithms for small satellite imagery: A systematic review | |
| Wu et al. | A framelet sparse reconstruction method for pansharpening with guaranteed convergence | |
| CN108876829B (en) | SAR high-precision registration method based on nonlinear scale space and radial basis function | |
| Hua et al. | Removing atmospheric turbulence effects via geometric distortion and blur representation | |
| CN116310883B (en) | Agricultural disaster prediction method based on remote sensing image space-time fusion and related equipment | |
| CN109615584B (en) | SAR image sequence MAP super-resolution reconstruction method based on homography constraint | |
| CN107194874B (en) | Super-resolution imaging system and method based on bias image stabilization | |
| Thurnhofer-Hemsi et al. | Super-resolution of 3D magnetic resonance images by random shifting and convolutional neural networks | |
| JP6541454B2 (en) | Image processing apparatus, imaging apparatus, image processing method, image processing program, and storage medium | |
| Amrouche et al. | Statistical destriping of pushbroom-type images based on an affine detector response | |
| Rahimi et al. | Preserving quality in minimum frame selection within multi-frame super-resolution | |
| Chen et al. | Single satellite imagery superresolution based on hybrid nonlocal similarity constrained convolution sparse coding | |
| Ghasemi-Falavarjani et al. | Particle filter based multi-frame image super resolution | |
| Gunturk | Super-resolution imaging | |
| CENGİZ et al. | The Effect of Super Resolution Method on Classification Performance of Satellite Images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |