CN107563964A - The quick joining method of large area array sub-meter grade night scene remote sensing image - Google Patents
The quick joining method of large area array sub-meter grade night scene remote sensing image Download PDFInfo
- Publication number
- CN107563964A CN107563964A CN201710722702.4A CN201710722702A CN107563964A CN 107563964 A CN107563964 A CN 107563964A CN 201710722702 A CN201710722702 A CN 201710722702A CN 107563964 A CN107563964 A CN 107563964A
- Authority
- CN
- China
- Prior art keywords
- mrow
- image
- msub
- rpc
- night scene
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 claims abstract description 46
- 238000012545 processing Methods 0.000 claims abstract description 32
- 238000012937 correction Methods 0.000 claims abstract description 23
- 238000005304 joining Methods 0.000 claims abstract description 15
- 238000012952 Resampling Methods 0.000 claims abstract description 12
- 238000000265 homogenisation Methods 0.000 claims abstract description 10
- 230000000149 penetrating effect Effects 0.000 claims abstract description 10
- 238000006243 chemical reaction Methods 0.000 claims description 19
- 238000003384 imaging method Methods 0.000 claims description 19
- 238000001914 filtration Methods 0.000 claims description 12
- 230000009466 transformation Effects 0.000 claims description 12
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 claims description 9
- 230000005855 radiation Effects 0.000 claims description 8
- AYFVYJQAPQTCCC-GBXIJSLDSA-N L-threonine Chemical compound C[C@@H](O)[C@H](N)C(O)=O AYFVYJQAPQTCCC-GBXIJSLDSA-N 0.000 claims description 7
- 238000000605 extraction Methods 0.000 claims description 5
- 230000015572 biosynthetic process Effects 0.000 claims description 4
- 101100182247 Caenorhabditis elegans lat-1 gene Proteins 0.000 claims description 3
- 101150061388 LON1 gene Proteins 0.000 claims description 3
- 238000009826 distribution Methods 0.000 claims description 3
- 238000005755 formation reaction Methods 0.000 claims description 3
- 235000013399 edible fruits Nutrition 0.000 claims 1
- 230000003287 optical effect Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- VMXUWOKSQNHOCA-UKTHLTGXSA-N ranitidine Chemical compound [O-][N+](=O)\C=C(/NC)NCCSCC1=CC=C(CN(C)C)O1 VMXUWOKSQNHOCA-UKTHLTGXSA-N 0.000 description 3
- 238000012892 rational function Methods 0.000 description 3
- 238000003672 processing method Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 244000207740 Lemna minor Species 0.000 description 1
- 235000006439 Lemna minor Nutrition 0.000 description 1
- 235000001855 Portulaca oleracea Nutrition 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000006073 displacement reaction Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000009738 saturating Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 230000002463 transducing effect Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Image Processing (AREA)
Abstract
The present invention relates to a kind of quick joining method of large area array sub-meter grade night scene remote sensing image, comprise the following steps：Step 1：Relative detector calibration；Step 2：Remove isolated noise；Step 3：Based on RPC without control block adjustment；Step 4：Light and color homogenization processing；Step 5：Correction and image resampling are just penetrated based on RPC.The present invention uses the quick joining method of large area array sub-meter grade night scene remote sensing image, by carrying out relative detector calibration, image denoising to raw video, just penetrating correction based on RPC without control block adjustment, light and color homogenization based on RPC, and the processing such as image resampling obtains large-scale night scene remote sensing splicing image, and the quick processing of algorithm is speeded up to by GPU, both the accuracy of algorithm had been ensure that, processing speed is greatly improved again, algorithm is simple and easy, is easily directly applied in project treatment.
Description
Technical field
The present invention relates to field of remote sensing image processing, more particularly to a kind of large area array sub-meter grade night scene remote sensing image it is quick
Joining method.
Background technology
Image joint is also known as image mosaic, be by two width or several there is the image joint of certain degree of overlapping together, structure
Into the technical process of a width overall image.In remote sensing image application, to obtain larger range of remote sensing image, it usually needs will
Breadth is smaller or image from different sensors is by handling and being stitched together.Only by simply splicing, can make
There is obvious geometry to misplace and radiate difference into stitching portion.Geometry dislocation is due to the space for splicing image with respect to position
Put that relation is incorrect or image regional area has the reasons such as certain geometric distortion and caused, and it is due to acquisition shadow to radiate difference
The reason such as the season of picture or sensor type difference causes.Therefore, in image mosaic, the geometry knot inlayed between image should be eliminated
Structure misplaces and radiation difference could preferably meet the application of image in practice.
Night scene image is the surface data that remote sensing satellite obtains at night, utilizes the low optical scanning function of sensor, Neng Gouyou
Effect ground detects urban lighting, or even to put etc. the low-intensity sent visible for small-scale settlement place, wagon flow, fishing boat light, fire
Optical emitter, the footprint of human lives is caught under the background of dark.Night scene remote sensing image is widely used in social economy
Many research fields such as parameter estimation, regional development research, urbanization monitoring, light pollution, can objectively react social economy
Variation tendency.However, night scene image is limited to the factors such as sensor manufacturing process, satellite orbital altitude when obtaining, individual
Image area coverage is limited, and individual image of such as long star 12K × 5K of light satellite video 03 0.92 meter of resolution ratio of super large area array covers
The surface area of lid is about 50 square kilometres, it is seen that it is main that individual large area array sub-meter grade image can not disposably obtain covering city
The broad range of data in region, it is therefore desirable to which multiple remote sensing in the range of multiple spot imaging task acquisition imaging task are carried out by satellite
Image, the image that multiple spot imaging pattern obtains spliced to obtain further according to the image overlap between neighbouring imaging final distant
Feel image, but the posture of satellite and differ when each task obtains image, it is different saturating that this will make it that every image has
Depending on distortion；Sensor is needed using higher gain and longer time for exposure during imaging, and this will be inevitably in image
Middle to introduce the highlighted noise such as noise and colourity noise, night scene image is integrally dark in addition, is unfavorable for visual interpretation and overlay region
The search of domain splicing line.Therefore the night scene remote sensing image for needing to obtain for different imaging tasks carries out distortion rectification, and image is gone
Make an uproar, the processing of image light and color homogenization；The RPC coefficients of single frames night scene image will be carried out based on RPC's there is certain error
Without control block adjustment, block adjustment process needs to carry out overlapping region sift feature point extractions with matching, in order to improve
Processing speed employs the sift matching algorithms based on GPU, and image is due to being center imaging, and there is certain in imaging process
Shooting angle, therefore also need to that image is carried out just to penetrate correction based on RPC, to the image resampling after correction can with
To final seamless spliced image.
The smoothness of gradation of image curved surface should be taken during image mosaic into account, the definition of image is taken into account again, for remote sensing
The stitching algorithm of image mainly includes the technology such as Image registration and synthesis, and Yang Li duckweeds, Lin Guangfa, Chen Youfei exist within 2009《Remote sensing
Technology and application》On deliver《Shading approach is studied before different aspect SPOT5 image mosaics》, its use numerical value adjustment,
The image processing method such as grid editor filling and feature information extraction, classification, the patch atural object widely different to tone on image
Handled with irregular scrappy atural object, reached the purpose for reducing tone difference, there is stronger applicability.Li De in 2006
Benevolence, Wang Mi, Pan Junzhen exist《Wuhan University Journal (information science version)》In deliver《The automatic dodging of optical remote sensing image
And application》, paper is for the image that optical remote sensing obtains in the presence of between several images in width image inside and regional extent
Color energy imbalance, it is proposed that automatic dodging method and handling process, by the even optical software of image to principle and stream
Cheng Jinhang is realized, and achieves good effect with reference to practical engineering application.Liu Xiaolong exists within 2001《Remote sensing journal》Upper hair
Table《The embedding technique for the digital orthoimage corrected based on Image Matching edge fit》, a kind of digital orthoimage of paper
Embedding technique, including preconditioning technique, embedding technique, inlay flow and application prospect, wherein preconditioning technique includes color
Coloured silk balance, Image Matching and edge fit correct technology, and color balance is corrected for radiancy, and Image Matching is corrected for several with edge fit
The correction of what difference.But the processing side of the few Bonding Problems for large area array sub-meter grade night scene remote sensing image of existing method
Method.
The content of the invention
The invention solves technical problem of the prior art, there is provided a kind of large area array sub-meter grade night scene remote sensing image it is fast
Fast joining method.
In order to solve the above-mentioned technical problem, technical scheme is specific as follows：
A kind of quick joining method of large area array sub-meter grade night scene remote sensing image, comprises the following steps：
Step 1：Relative detector calibration
The processing of correction is normalized to the quantized value of original night remote sensing image each pixel spoke monochrome information response,
The each detecting element response difference of sensor is reduced or eliminated, makes response uniformity of the detecting element to spoke brightness；Utilize phase
Relative detector calibration is carried out to the result of radiation calibration；
Step 2：Remove isolated noise
First, initial data is isolated to the information of tri- wave bands of R, G, B, medium filtering is carried out to each wave band, i.e.,
Imed(R, G, B)=medfilt (Iori) (1)
Wherein, IoriFor initial data, ImedFor the image of medium filtering；
Thereafter, isolated noise medium filtering image will be filtered out and carry out binary conversion treatment, i.e.,
Ibw(R, G, B)=im2bw (Imed(R,G,B),thre) (2)
Wherein, IbwFor binary image, thre is the threshold value of binary conversion treatment；
Binary image is multiplied point by point with initial data, i.e.,
Idenoise(i, j)=Ibw(i,j)×Iori(i,j) (3)
Wherein, IdenoiseFor the image after denoising, Idenoise(i, j) is the gray value of the i-th row of image jth row；
Step 3：Based on RPC without control block adjustment
Sift feature point extractions and matching based on GPU are carried out to the overlapping region of multiple spot imaging task neighbour's close-range image,
Multiframe night scene data are carried out using matching result to eliminate error present in RPC without control block adjustment based on RPC；Step
Rapid four：Light and color homogenization processing
The even light of Mask is carried out to night scene image and the even color based on Wallis conversion is handled；
Using the even light of Mask differential techniques, obtained using Gassian low-pass filter method after the background video of raw video using original
Image carries out phase reducing with background video, obtains the image of illuminance distribution, and its phase reducing is using the public affairs shown in formula 6
Formula；
Iout(x, y)=Iin(x,y)-Iback(x,y)+offset (6)
Offset is grayscale shift amount in formula 6；
According to the maximum f of raw video gray levelmax, minimum value fmin, average valueAnd result image greyscale is maximum
Value gmax, minimum value gmin, average valueCarry out divided linear strength；Stretching formula is provided by formula 7：
G (x, y) is even light result image in formula 7, g'(x, y) it is to the image after even light result image stretch processing；
Wallis conversion can be expressed as：
G (x, y) and f (x, y) is respectively the gray value of raw video and Wallis transformation results images in formula 8；mgAnd mfPoint
Not Wei raw video local gray level average and standard deviation；sgAnd sfRespectively result image local gray level average and standard deviation
The desired value of difference；C ∈ [0,1] are the extension constant of image variance；B ∈ [0,1] are the luminance factor of image, as b → 1, shadow
As average is forced to mf, as b → 0, image average is forced to mg；
Step 5：Correction and image resampling are just penetrated based on RPC
The night scene image obtained to multiple spot imaging is carried out just penetrating correction based on RPC, and weight is carried out to the image after correction
Sampling obtains final splicing image.
In the above-mentioned technical solutions, in step 5, it is as follows that just penetrating based on RPC corrects step：
1) angle point object coordinates are calculated
The initial object coordinates of correspondence provided according to the angle point image space coordinate of image four and regularization coefficient are just being calculated with RPC respectively
Geographical coordinates is obtained, and solves initial affine transformation coefficient, then object space point coordinates corresponding to picpointed coordinate is calculated based on RPC；
2) result image is built
Longitude and latitude minimum, maximum are worth to image covering in the ground areas object space plane coordinates according to corresponding to being imaged multiple spot
Scope (lat0~lat1, lon0~lon1), setting orthography resolution ratio gsd, you can calculate image size (W, H)：
3) pixel-by-pixel Ergodic Maps to raw video
The each pixel (s, l) of orthography can be calculated by projection formula and arrive object space plane coordinates initial value (lat, lon),
Formula is as follows：
The elevation H of (lat, lon) position is obtained according to dem data, RPC model formations is substituted into and picpointed coordinate is calculated
(x,y)；
4) interpolation gray value and assignment
The picpointed coordinate (x, y) obtained by inverse in 3), the interpolation gray scale on raw video；After calculating gray scale p, assignment
Give result image (s, l) position, final output night scene remote sensing splicing image.
In the above-mentioned technical solutions, in step 1), RPC just calculations obtain object space point coordinates and comprised the following steps：
(1) object space plane coordinates initial value (Lon, Lat) is given according to initial affine transformation figure parameters for picpointed coordinate；
(2) dem data is read according to given object space plane coordinates initial value and obtains height value H, corresponding picture is solved using RPC
Point coordinates, new radiation conversion coefficient is solved using picture point and object space point coordinates；
(3) object space plane coordinates is given according to new affine transformation figure parameters for picpointed coordinate, and it is corresponding to read DEM
Height value, solve and finish if the height value solved twice is less than threshold value, otherwise repeatedly said process, calculated until twice
Untill object space elevation difference out is less than threshold value.
In the above-mentioned technical solutions, in step 4), the interpolation uses bilinear interpolation method, and formula is as follows：P=p (i,
j)*(1-dx)*(1-dy)+p(i+1,j)*dx*(1-dy)+p(i+1,j+1)*dx*dy+p(i,j+1)*(1-dx)*dy。
The present invention has following beneficial effect：
The present invention is based on existing image mosaic Processing Algorithm, and the attitude of satellite becomes when taking into full account the imaging of optical satellite multiple spot
Change larger, night scene video imaging object is hot spot and the imaging characteristicses of isolated noise be present, and large area array high-resolution image number
According to the characteristics of big is measured, to raw video progress relative detector calibration, image denoising, based on RPC without control block adjustment, even light
Even color, correction just penetrated based on RPC, and the processing such as image resampling obtains large-scale night scene remote sensing splicing image.
The quick joining method of large area array sub-meter grade night scene remote sensing image provided by the invention, by being carried out to raw video
Relative detector calibration, image denoising, correction just penetrated based on RPC without control block adjustment, light and color homogenization based on RPC, and
The processing such as image resampling obtain large-scale night scene remote sensing splicing image, and have speeded up to the quick of algorithm by GPU
Processing, both ensure that the accuracy of algorithm, and had greatly improved processing speed again, algorithm is simple and easy, easily in project treatment
In directly apply.
Brief description of the drawings
The present invention is described in further detail with reference to the accompanying drawings and detailed description.
Fig. 1 is the flow chart of the quick joining method of large area array sub-meter grade night scene remote sensing image of the present invention.
Fig. 2 and Fig. 3 be night scene image joint of the present invention before and after effect contrast figure, wherein Fig. 2 be splicing before image, Fig. 3
For spliced image.
Embodiment
The original color image that the present invention obtains sensor carries out relative radiation calibration and obtains the consistent night scene shadow of radiation
Picture, image denoising is carried out for isolated noise, carried out for denoising night scene image based on RPC without control block adjustment, for
Orthography carries out light and color homogenization processing, and just penetrates correction based on RPC, finally several night scene images consistent to color search
Splicing line obtains preferable night scene splicing image.
The present invention is described in detail below in conjunction with the accompanying drawings.
Referring to Fig. 1：A kind of quick joining method of large area array sub-meter grade night scene remote sensing image, comprises the following steps：
Step 1：Relative detector calibration
Relative detector calibration is also known as the normalized of sensor detecting element, is that each pixel spoke monochrome information is rung
The processing of correction is normalized in the quantized value (DN) answered, and reduces or eliminates each detecting element response difference of sensor, make spy
Survey response uniformity of the element to spoke brightness.Relative detector calibration is carried out using the result of relative radiometric calibration.
Step 2：Remove isolated noise
First, initial data is isolated to the information of tri- wave bands of R, G, B, medium filtering is carried out to each wave band, i.e.,
Imed(R, G, B)=medfilt (Iori) (1)
Wherein, IoriFor initial data, ImedFor the image of medium filtering.Because the highlighted noise in image is isolated makes an uproar
Point, therefore the algorithm of medium filtering can filter out the highlighted noise in image well.
Thereafter, isolated noise medium filtering image will be filtered out and carry out binary conversion treatment, i.e.,
Ibw(R, G, B)=im2bw (Imed(R,G,B),thre) (2)
Wherein, IbwFor binary image, thre is the threshold value of binary conversion treatment.By binary conversion treatment, can isolate
Background noise and foreground information.
Binary image is multiplied point by point with initial data, i.e.,
Idenoise(i, j)=Ibw(i,j)×Iori(i,j) (3)
Wherein, IdenoiseFor the image after denoising, Idenoise(i, j) is the gray value of the i-th row of image jth row.Now obtain
Denoising image not only remove image dark place ambient noise and isolated highlighted noise, while remain the brighter place of image
High-frequency information.
Step 3：Based on RPC without control block adjustment
Rational function model is that picpointed coordinate (r, c) is expressed as with corresponding topocentric coordinates (X, Y, Z) as independent variable
Multinomial ratio, i.e.,
(r in formulan,cn) and (Xn,Yn,Zn) represent that pixel coordinate (r, c) and topocentric coordinates (X, Y, Z) contract through translation respectively
Normalized coordinates after putting, between value is -1.0~+1.0, wherein polynomial coefficient is referred to as the coefficient of rational function
(rational function coefficient, RPC), can be established between image coordinate system and earth axes by RPC and closed
System.
In the case that control participates on no ground, satellite attitude parameters and the RPC moulds of orbit parameter generation are directly utilized
Type often contains larger systematic error, have impact on image positioning precision so that more scape images just penetrate correction after overlapping region without
Method overlaps exactly, it is therefore desirable to is imaged the night scene image obtained for multiple spot and carries out based on RPC without control regional network
Adjustment.Need to carry out Feature Points Matching for night scene image overlap area during adjustment, using with anti-rotation consistency
Sift is matched, in order to realize that the quick splicing of night scene image, can be by routine based on CPU using the sift matchings based on GPU
Sift match times shorten about 140 times.
Using acquisition point-to-multipoint of the same name imaging obtain night scene image carry out based on image space without control block adjustment,
Error existing for determining single scape image RPC models to be spliced simultaneously, it ensure that and eliminate after error corresponding to the image of overlapping region
Atural object there's almost no relative geographical position deviation.
Step 4：Light and color homogenization processing
Due to the image of the time of image capturing, shooting angle, external light source, atmospheric attenuation and some other factor,
Different degrees of difference on color be present in the night scene image that can cause to obtain, thus night scene image is carried out the even light of Mask and
The even color processing of Wallis.
Using the even light of Mask differential techniques, obtained using Gassian low-pass filter method after the background video of raw video using original
Image carries out phase reducing with background video, you can obtains the image of illuminance distribution, its phase reducing can use the institute of formula 6
The formula shown.
Iout(x, y)=Iin(x,y)-Iback(x,y)+offset (6)
Offset is grayscale shift amount in formula 6.
In order to increase adjacent thin portion contrast, the overall contrast of whole image is improved, it is necessary to divide the image after processing
Section linear stretch processing.According to the maximum f of raw video gray levelmax, minimum value fmin, average valueAnd result image
Gray scale maximum gmax, minimum value gmin, average valueCarry out divided linear strength.Stretching formula is provided by formula 7：
G (x, y) is even light result image in formula 7, g'(x, y) it is to the image after even light result image stretch processing.This
Kind divided linear strength need not add additional parameter, and image greyscale dynamic range after processing can be restored into raw video
In tonal range.
Need to carry out based on Wallis conversion colors equilibrium treatment side to adjust the color balance between night scene remote sensing image
Method, Wallis become transducing and suppress noise again while original sub-meter grade image local contrast is strengthened, and have local auto-adaptive work(
Energy.Wallis conversion can be expressed as：
G (x, y) and f (x, y) is respectively the gray value of raw video and Wallis transformation results images in formula 8；mgAnd mfPoint
Not Wei raw video local gray level average and standard deviation (variance)；sgAnd sfRespectively result image local gray level average and
The desired value of standard deviation；C ∈ [0,1] are the extension constant of image variance；B ∈ [0,1] are the luminance factor of image, when b → 1
When, image average is forced to mf, as b → 0, image average is forced to mg。
Step 5：Correction and image resampling are just penetrated based on RPC
Projected centered on large area array night scene remote sensing image, and certain inclination angle during image capturing be present, in order to eliminate due to
Deformation caused by image inclination and hypsography etc. carries out just penetrating based on RPC, it is necessary to be imaged the night scene image obtained to multiple spot
Correct, and resampling is carried out to the image after correction and obtains final splicing image.Just penetrating wherein based on RPC corrects step such as
Under：
1) angle point object coordinates are calculated
The initial object coordinates of correspondence provided according to the angle point image space coordinate of image four and regularization coefficient are just being calculated with RPC respectively
To geographical coordinates, and solve initial affine transformation coefficient.Hereafter object space point corresponding to picpointed coordinate is calculated based on RPC again to sit
Mark.Following steps can specifically be used：
(1) object space plane coordinates initial value (Lon, Lat) is given according to initial affine transformation figure parameters for picpointed coordinate；
(2) dem data is read according to given object space plane coordinates initial value and obtains height value H, corresponding picture is solved using RPC
Point coordinates, new radiation conversion coefficient is solved using picture point and object space point coordinates；
(3) object space plane coordinates is given according to new affine transformation figure parameters for picpointed coordinate, and it is corresponding to read DEM
Height value, solve and finish if the height value solved twice is less than threshold value, otherwise repeatedly said process, calculated until twice
Untill object space elevation difference out is less than threshold value.
2) result image is built
Longitude and latitude minimum, maximum are worth to image covering in the ground areas object space plane coordinates according to corresponding to being imaged multiple spot
Scope (lat0~lat1, lon0~lon1), setting orthography resolution ratio gsd, you can calculate image size (W, H)：
3) pixel-by-pixel Ergodic Maps to raw video
The each pixel (s, l) of orthography can be calculated by projection formula and arrive object space plane coordinates initial value (lat, lon),
Formula is as follows：
The elevation H of (lat, lon) position is obtained according to dem data, RPC model formations is substituted into and image space coordinate is calculated
(x,y)。
4) interpolation gray value and assignment
The picpointed coordinate (x, y) obtained by inverse in 3), the interpolation gray scale on raw video.Interpolation is divided into closest to interpolation
Method, three kinds of methods of bilinear interpolation method and bi-cubic interpolation method are public using bilinear interpolation method in order to take into account efficiency and precision
Formula is as follows：
P=p (i, j) * (1-dx) * (1-dy)+p (i+1, j) * dx* (1-dy)+p (i+1, j+1) * dx*dy+p (i, j+1) *
(1-dx)*dy
After calculating gray scale p, result image (s, l) position is assigned to, the remote sensing of final output night scene splices image (referring to figure
2 and 3).
The quick joining method of large area array sub-meter grade night scene remote sensing image provided by the invention, by being carried out to raw video
Relative detector calibration, image denoising, correction just penetrated without control block adjustment, light and color homogenization, based on RPC based on RPC, and
The processing such as image resampling obtain large-scale night scene remote sensing splicing image, and have speeded up to the quick of algorithm by GPU
Processing, both ensure that the accuracy of algorithm, and had greatly improved processing speed again, algorithm is simple and easy, easily in project treatment
In directly apply.
The satellite launched with reference to Chang Guang satellite technologies Co., Ltd --- exemplified by the star of video 03, describe night scene shadow in detail
The denoising of picture and enhancing processing method.
The star of video 03 uses main square as 3200mm video cameras, and the resolution ratio of substar is 0.92m, the single frames night scene of collection
Image size is 12000 × 5000 pixels.For the star of video 03 on April 1st, 2017,5 points were taken the photograph London night scene, shooting point longitude and latitude for 43 minutes
It is 51.4628 ° to spend for -0.179 ° of longitude, latitude.The big face of the present invention is illustrated for this multiple spot night scene video imaging task
The quick joining method of battle array sub-meter grade night scene remote sensing image, this method comprise the following steps：
Step 1：Relative radiation calibration
Relative detector calibration is carried out to night scene image according to relative radiometric calibration result.
Step 2：Isolated noise denoising
First, initial data is isolated to the information of tri- wave bands of R, G, B, R, G, B size are respectively 3000 × 1250
Pixel, 6000 × 2500 pixels and 3000 × 1250 pixels, medium filtering is carried out using formula (1) to each wave band, gone
Except the filtered image I of highlighted noisemed(R,G,B)。
Thereafter, isolated noise medium filtering image I will be filtered outmed(R, G, B) carries out binary conversion treatment, root according to formula (2)
According to the view data of acquisition, it is 6 to set binary-state threshold thre, so as to obtain binary image Ibw.Pass through binary conversion treatment, energy
Background noise and foreground information is enough effectively separated.By binary image and initial data according to formula (3) carry out by
Point is multiplied, and obtains the image I after denoisingdenoise。
The denoising image now obtained not only removes the ambient noise of the dark place of image and isolated highlighted noise, protects simultaneously
The high-frequency information at the brighter place of image is stayed.
Step 3：Based on RPC without control block adjustment
Sift feature point extractions and matching based on GPU are carried out to the overlapping region of multiple spot imaging task neighbour's close-range image,
Multiframe night scene data are carried out using matching result to eliminate error present in RPC without control block adjustment based on RPC.
Step 4：Light and color homogenization processing
The dodging based on Mask principles is carried out for multiframe night scene image and the even color based on Wallis conversion is handled
Obtain the consistent night scene remote sensing image of color.
Step 5：Correction and resampling are just penetrated based on RPC
The image point displacement as caused by imaging angle and hypsography, and root are eliminated by the correction processing of just penetrating based on RPC
Obtain splicing image size according to the object space range computation of image, and for splicing image node-by-node algorithm its in single scape night scene image
Position simultaneously carries out bilinear interpolation and obtains the night scene remote sensing image of final multiple spot splicing.
Obviously, above-described embodiment is only intended to clearly illustrate example, and is not the restriction to embodiment.It is right
For those of ordinary skill in the art, can also make on the basis of the above description it is other it is various forms of change or
Change.There is no necessity and possibility to exhaust all the enbodiments.And the obvious change thus extended out or
Among changing still in the protection domain of the invention.
Claims (4)
1. a kind of quick joining method of large area array sub-meter grade night scene remote sensing image, it is characterised in that comprise the following steps：
Step 1：Relative detector calibration
The processing of correction is normalized to the quantized value of original night remote sensing image each pixel spoke monochrome information response, reduced
Or each detecting element response difference of sensor is eliminated, make response uniformity of the detecting element to spoke brightness；Utilize relative spoke
The result for penetrating calibration carries out relative detector calibration；
Step 2：Remove isolated noise
First, initial data is isolated to the information of tri- wave bands of R, G, B, medium filtering is carried out to each wave band, i.e.,
Imed(R, G, B)=medfilt (Iori) (1)
Wherein, IoriFor initial data, ImedFor the image of medium filtering；
Thereafter, isolated noise medium filtering image will be filtered out and carry out binary conversion treatment, i.e.,
Ibw(R, G, B)=im2bw (Imed(R,G,B),thre) (2)
Wherein, IbwFor binary image, thre is the threshold value of binary conversion treatment；
Binary image is multiplied point by point with initial data, i.e.,
Idenoise(i, j)=Ibw(i,j)×Iori(i,j) (3)
Wherein, IdenoiseFor the image after denoising, Idenoise(i, j) is the gray value of the i-th row of image jth row；
Step 3：Based on RPC without control block adjustment
Sift feature point extractions and matching based on GPU are carried out to the overlapping region of multiple spot imaging task neighbour's close-range image, utilized
Matching result carries out eliminating error present in RPC without control block adjustment based on RPC to multiframe night scene data；
Step 4：Light and color homogenization processing
The even light of Mask is carried out to night scene image and the even color based on Wallis conversion is handled；
Using the even light of Mask differential techniques, raw video is utilized after obtaining the background video of raw video using Gassian low-pass filter method
Phase reducing is carried out with background video, obtains the image of illuminance distribution, its phase reducing is using the formula shown in formula 6；
Iout(x, y)=Iin(x,y)-Iback(x,y)+offset (6)
Offset is grayscale shift amount in formula 6；
According to the maximum f of raw video gray levelmax, minimum value fmin, average valueAnd result image greyscale maximum
gmax, minimum value gmin, average valueCarry out divided linear strength；Stretching formula is provided by formula 7：
<mrow>
<msup>
<mi>g</mi>
<mo>&prime;</mo>
</msup>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mfrac>
<mrow>
<mover>
<mi>f</mi>
<mo>&OverBar;</mo>
</mover>
<mo>-</mo>
<msub>
<mi>f</mi>
<mi>min</mi>
</msub>
</mrow>
<mrow>
<mover>
<mi>g</mi>
<mo>&OverBar;</mo>
</mover>
<mo>-</mo>
<msub>
<mi>g</mi>
<mi>min</mi>
</msub>
</mrow>
</mfrac>
<mo>&CenterDot;</mo>
<mrow>
<mo>(</mo>
<mi>g</mi>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>-</mo>
<msub>
<mi>g</mi>
<mi>min</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msub>
<mi>f</mi>
<mi>min</mi>
</msub>
</mrow>
</mtd>
<mtd>
<mrow>
<msub>
<mi>g</mi>
<mi>min</mi>
</msub>
<mo>&le;</mo>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo><</mo>
<mover>
<mi>g</mi>
<mo>&OverBar;</mo>
</mover>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mfrac>
<mrow>
<msub>
<mi>f</mi>
<mi>max</mi>
</msub>
<mo>-</mo>
<mover>
<mi>f</mi>
<mo>&OverBar;</mo>
</mover>
</mrow>
<mrow>
<msub>
<mi>g</mi>
<mi>max</mi>
</msub>
<mo>-</mo>
<mover>
<mi>g</mi>
<mo>&OverBar;</mo>
</mover>
</mrow>
</mfrac>
<mo>&CenterDot;</mo>
<mrow>
<mo>(</mo>
<mi>g</mi>
<mo>(</mo>
<mrow>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
</mrow>
<mo>)</mo>
<mo>-</mo>
<mover>
<mi>g</mi>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mover>
<mi>f</mi>
<mo>&OverBar;</mo>
</mover>
</mrow>
</mtd>
<mtd>
<mrow>
<mover>
<mi>g</mi>
<mo>&OverBar;</mo>
</mover>
<mo>&le;</mo>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>&le;</mo>
<msub>
<mi>g</mi>
<mi>max</mi>
</msub>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
G (x, y) is even light result image in formula 7, g'(x, y) it is to the image after even light result image stretch processing；
Wallis conversion can be expressed as：
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mo>&lsqb;</mo>
<mi>g</mi>
<mrow>
<mo>(</mo>
<mi>x</mi>
<mo>,</mo>
<mi>y</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<msub>
<mi>m</mi>
<mi>g</mi>
</msub>
<mo>&rsqb;</mo>
<mfrac>
<mrow>
<msub>
<mi>cs</mi>
<mi>f</mi>
</msub>
</mrow>
<mrow>
<msub>
<mi>cs</mi>
<mi>g</mi>
</msub>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>-</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
<msub>
<mi>s</mi>
<mi>f</mi>
</msub>
</mrow>
</mfrac>
<mo>+</mo>
<msub>
<mi>bm</mi>
<mi>f</mi>
</msub>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>-</mo>
<mi>b</mi>
<mo>)</mo>
</mrow>
<msub>
<mi>m</mi>
<mi>g</mi>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
G (x, y) and f (x, y) is respectively the gray value of raw video and Wallis transformation results images in formula 8；mgAnd mfRespectively
The local gray level average and standard deviation of raw video；sgAnd sfRespectively result image local gray level average and standard deviation
Desired value；C ∈ [0,1] are the extension constant of image variance；B ∈ [0,1] are the luminance factor of image, and as b → 1, image is equal
Value is forced to mf, as b → 0, image average is forced to mg；
Step 5：Correction and image resampling are just penetrated based on RPC
The night scene image obtained to multiple spot imaging carries out just penetrating correction based on RPC, and carries out resampling to the image after correction
Obtain final splicing image.
2. the quick joining method of large area array sub-meter grade night scene remote sensing image according to claim 1, it is characterised in that step
In rapid five, it is as follows that just penetrating based on RPC corrects step：
In step 5, it is as follows that just penetrating based on RPC corrects step：
1) angle point object coordinates are calculated
The initial object coordinates of correspondence provided according to the angle point image space coordinate of image four and regularization coefficient are just being obtained with RPC respectively
Geographical coordinates, and initial affine transformation coefficient is solved, then object space point coordinates corresponding to picpointed coordinate is calculated based on RPC；
2) result image is built
Longitude and latitude minimum, maximum are worth to image coverage in the ground areas object space plane coordinates according to corresponding to being imaged multiple spot
(lat0~lat1, lon0~lon1), setting orthography resolution ratio gsd, you can calculate image size (W, H)：
<mrow>
<mtable>
<mtr>
<mtd>
<mrow>
<mi>W</mi>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mi>o</mi>
<mi>n</mi>
<mn>1</mn>
<mo>-</mo>
<mi>l</mi>
<mi>o</mi>
<mi>n</mi>
<mn>0</mn>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mi>g</mi>
<mi>s</mi>
<mi>d</mi>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>H</mi>
<mo>=</mo>
<mrow>
<mo>(</mo>
<mi>l</mi>
<mi>a</mi>
<mi>t</mi>
<mn>1</mn>
<mo>-</mo>
<mi>l</mi>
<mi>a</mi>
<mi>t</mi>
<mn>0</mn>
<mo>)</mo>
</mrow>
<mo>/</mo>
<mi>g</mi>
<mi>s</mi>
<mi>d</mi>
</mrow>
</mtd>
</mtr>
</mtable>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>9</mn>
<mo>)</mo>
</mrow>
</mrow>
3) pixel-by-pixel Ergodic Maps to raw video
The each pixel (s, l) of orthography can be calculated by projection formula and arrive object space plane coordinates initial value (lat, lon), formula
It is as follows：
<mrow>
<mtable>
<mtr>
<mtd>
<mrow>
<mi>l</mi>
<mi>a</mi>
<mi>t</mi>
<mo>=</mo>
<mi>l</mi>
<mo>*</mo>
<mi>g</mi>
<mi>s</mi>
<mi>d</mi>
<mo>+</mo>
<mi>l</mi>
<mi>a</mi>
<mi>t</mi>
<mn>0</mn>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>l</mi>
<mi>o</mi>
<mi>n</mi>
<mo>=</mo>
<mi>s</mi>
<mo>*</mo>
<mi>g</mi>
<mi>s</mi>
<mi>d</mi>
<mo>+</mo>
<mi>l</mi>
<mi>o</mi>
<mi>n</mi>
<mn>0</mn>
</mrow>
</mtd>
</mtr>
</mtable>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>10</mn>
<mo>)</mo>
</mrow>
</mrow>
The elevation H of (lat, lon) position is obtained according to dem data, RPC model formations is substituted into and picpointed coordinate (x, y) is calculated；
4) interpolation gray value and assignment
The picpointed coordinate (x, y) obtained by inverse in 3), the interpolation gray scale on raw video；After calculating gray scale p, knot is assigned to
Fruit image (s, l) position, final output night scene remote sensing splicing image.
3. the quick joining method of large area array sub-meter grade night scene remote sensing image according to claim 2, it is characterised in that step
It is rapid 1) in, RPC is just calculated and is obtained object space point coordinates and comprise the following steps：
(1) object space plane coordinates initial value (Lon, Lat) is given according to initial affine transformation figure parameters for picpointed coordinate；
(2) dem data is read according to given object space plane coordinates initial value and obtains height value H, solving corresponding picture point using RPC sits
Mark, new radiation conversion coefficient is solved using picture point and object space point coordinates；
(3) object space plane coordinates is given according to new affine transformation figure parameters for picpointed coordinate, and read high corresponding to DEM
Journey value, solve and finish if the height value solved twice is less than threshold value, otherwise repeatedly said process, calculated until twice
Object space elevation difference be less than threshold value untill.
4. the quick joining method of large area array sub-meter grade night scene remote sensing image according to claim 2, it is characterised in that step
It is rapid 4) in, the interpolation uses bilinear interpolation method, and formula is as follows：
P=p (i, j) * (1-dx) * (1-dy)+p (i+1, j) * dx* (1-dy)+p (i+1, j+1) * dx*dy+p (i, j+1) * (1-
dx)*dy。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710722702.4A CN107563964B (en) | 2017-08-22 | 2017-08-22 | Rapid splicing method for large-area-array sub-meter-level night scene remote sensing images |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710722702.4A CN107563964B (en) | 2017-08-22 | 2017-08-22 | Rapid splicing method for large-area-array sub-meter-level night scene remote sensing images |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107563964A true CN107563964A (en) | 2018-01-09 |
| CN107563964B CN107563964B (en) | 2020-09-04 |
Family
ID=60976225
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710722702.4A Active CN107563964B (en) | 2017-08-22 | 2017-08-22 | Rapid splicing method for large-area-array sub-meter-level night scene remote sensing images |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107563964B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108550129A (en) * | 2018-04-20 | 2018-09-18 | 北京航天宏图信息技术股份有限公司 | Even color method and device based on geographical template |
| CN108734685A (en) * | 2018-05-10 | 2018-11-02 | 中国矿业大学(北京) | A kind of joining method of UAV system EO-1 hyperion linear array remote sensing image |
| CN109063711A (en) * | 2018-07-06 | 2018-12-21 | 航天星图科技（北京）有限公司 | A kind of satellite image based on LLTS frame just penetrates correct algorithm |
| CN109118421A (en) * | 2018-07-06 | 2019-01-01 | 航天星图科技（北京）有限公司 | A kind of image light and color homogenization method based on Distributed Architecture |
| CN109118429A (en) * | 2018-08-02 | 2019-01-01 | 武汉大学 | A kind of medium-wave infrared-visible light multispectral image rapid generation |
| CN110276280A (en) * | 2019-06-06 | 2019-09-24 | 刘嘉津 | A kind of optical processing method of crop pests image automatic identification |
| CN110660023A (en) * | 2019-09-12 | 2020-01-07 | 中国测绘科学研究院 | Video stitching method based on image semantic segmentation |
| CN110988908A (en) * | 2019-12-19 | 2020-04-10 | 长光卫星技术有限公司 | Quantitative analysis method for influence of spectral shift of optical filter on imaging of spatial optical remote sensor |
| CN112017115A (en) * | 2020-07-09 | 2020-12-01 | 卢凯旋 | Remote sensing image splicing method, device, equipment and storage medium |
| CN112184546A (en) * | 2020-06-10 | 2021-01-05 | 中国人民解放军32023部队 | Satellite remote sensing image data processing method |
| CN112233190A (en) * | 2020-05-19 | 2021-01-15 | 同济大学 | Satellite remote sensing image color balancing method based on block adjustment |
| CN112288650A (en) * | 2020-10-28 | 2021-01-29 | 武汉大学 | A multi-source remote sensing satellite image geometry and semantic integration processing method and system |
| CN112465986A (en) * | 2020-11-27 | 2021-03-09 | 航天恒星科技有限公司 | Method and device for inlaying satellite remote sensing image |
| CN113469899A (en) * | 2021-06-04 | 2021-10-01 | 中国资源卫星应用中心 | Optical remote sensing satellite relative radiation correction method based on radiant energy reconstruction |
| US11430128B2 (en) * | 2018-02-07 | 2022-08-30 | Chang'an University | Geological linear body extraction method based on tensor voting coupled with Hough transformation |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104282006A (en) * | 2014-09-30 | 2015-01-14 | 中国科学院国家天文台 | High-resolution image splicing method based on CE-2 data |
| CN106373088A (en) * | 2016-08-25 | 2017-02-01 | 中国电子科技集团公司第十研究所 | Quick mosaic method for aviation images with high tilt rate and low overlapping rate |
-
2017
- 2017-08-22 CN CN201710722702.4A patent/CN107563964B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104282006A (en) * | 2014-09-30 | 2015-01-14 | 中国科学院国家天文台 | High-resolution image splicing method based on CE-2 data |
| CN106373088A (en) * | 2016-08-25 | 2017-02-01 | 中国电子科技集团公司第十研究所 | Quick mosaic method for aviation images with high tilt rate and low overlapping rate |
Non-Patent Citations (1)
| Title |
|---|
| 汪韬阳等: "卫星遥感影像的区域正射纠正", 《武汉大学学报(信息科学版)》 * |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11430128B2 (en) * | 2018-02-07 | 2022-08-30 | Chang'an University | Geological linear body extraction method based on tensor voting coupled with Hough transformation |
| CN108550129B (en) * | 2018-04-20 | 2019-04-09 | 北京航天宏图信息技术股份有限公司 | Even color method and device based on geographical template |
| CN108550129A (en) * | 2018-04-20 | 2018-09-18 | 北京航天宏图信息技术股份有限公司 | Even color method and device based on geographical template |
| CN108734685A (en) * | 2018-05-10 | 2018-11-02 | 中国矿业大学(北京) | A kind of joining method of UAV system EO-1 hyperion linear array remote sensing image |
| CN108734685B (en) * | 2018-05-10 | 2022-06-03 | 中国矿业大学(北京) | Splicing method for unmanned aerial vehicle-mounted hyperspectral line array remote sensing images |
| CN109063711B (en) * | 2018-07-06 | 2021-10-29 | 中科星图股份有限公司 | Satellite image orthorectification algorithm based on LLTS framework |
| CN109118421A (en) * | 2018-07-06 | 2019-01-01 | 航天星图科技（北京）有限公司 | A kind of image light and color homogenization method based on Distributed Architecture |
| CN109063711A (en) * | 2018-07-06 | 2018-12-21 | 航天星图科技（北京）有限公司 | A kind of satellite image based on LLTS frame just penetrates correct algorithm |
| CN109118429B (en) * | 2018-08-02 | 2023-04-25 | 武汉大学 | Method for rapidly generating intermediate wave infrared-visible light multispectral image |
| CN109118429A (en) * | 2018-08-02 | 2019-01-01 | 武汉大学 | A kind of medium-wave infrared-visible light multispectral image rapid generation |
| CN110276280A (en) * | 2019-06-06 | 2019-09-24 | 刘嘉津 | A kind of optical processing method of crop pests image automatic identification |
| CN110660023A (en) * | 2019-09-12 | 2020-01-07 | 中国测绘科学研究院 | Video stitching method based on image semantic segmentation |
| CN110988908A (en) * | 2019-12-19 | 2020-04-10 | 长光卫星技术有限公司 | Quantitative analysis method for influence of spectral shift of optical filter on imaging of spatial optical remote sensor |
| CN110988908B (en) * | 2019-12-19 | 2023-06-09 | 长光卫星技术股份有限公司 | Quantitative analysis method for imaging influence of spectral shift of optical filter on space optical remote sensor |
| CN112233190A (en) * | 2020-05-19 | 2021-01-15 | 同济大学 | Satellite remote sensing image color balancing method based on block adjustment |
| CN112233190B (en) * | 2020-05-19 | 2023-04-07 | 同济大学 | Satellite remote sensing image color balancing method based on block adjustment |
| CN112184546A (en) * | 2020-06-10 | 2021-01-05 | 中国人民解放军32023部队 | Satellite remote sensing image data processing method |
| CN112184546B (en) * | 2020-06-10 | 2024-03-15 | 中国人民解放军32023部队 | Satellite remote sensing image data processing method |
| CN112017115A (en) * | 2020-07-09 | 2020-12-01 | 卢凯旋 | Remote sensing image splicing method, device, equipment and storage medium |
| CN112288650A (en) * | 2020-10-28 | 2021-01-29 | 武汉大学 | A multi-source remote sensing satellite image geometry and semantic integration processing method and system |
| CN112465986A (en) * | 2020-11-27 | 2021-03-09 | 航天恒星科技有限公司 | Method and device for inlaying satellite remote sensing image |
| CN113469899A (en) * | 2021-06-04 | 2021-10-01 | 中国资源卫星应用中心 | Optical remote sensing satellite relative radiation correction method based on radiant energy reconstruction |
| CN113469899B (en) * | 2021-06-04 | 2023-12-29 | 中国资源卫星应用中心 | Optical remote sensing satellite relative radiation correction method based on radiation energy reconstruction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107563964B (en) | 2020-09-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107563964A (en) | The quick joining method of large area array sub-meter grade night scene remote sensing image | |
| CN110310248B (en) | A kind of real-time joining method of unmanned aerial vehicle remote sensing images and system | |
| CN111080724B (en) | Fusion method of infrared light and visible light | |
| CN110648398B (en) | Real-time ortho image generation method and system based on unmanned aerial vehicle aerial data | |
| CN103413272B (en) | Low spatial resolution multi-source Remote Sensing Images Space Consistency bearing calibration | |
| WO2021120406A1 (en) | Infrared and visible light fusion method based on saliency map enhancement | |
| CN106373088B (en) | The quick joining method of low Duplication aerial image is tilted greatly | |
| TWI322261B (en) | ||
| Grumpe et al. | Construction of lunar DEMs based on reflectance modelling | |
| CN109934788B (en) | Remote sensing image missing data restoration method based on standard remote sensing image | |
| EP2095330A1 (en) | Panchromatic modulation of multispectral imagery | |
| CN102073874A (en) | Geometric constraint-attached spaceflight three-line-array charged coupled device (CCD) camera multi-image stereo matching method | |
| EP2100268A2 (en) | Structured smoothing for superresolution of multispectral imagery based on registered panchromatic image | |
| CN108734685A (en) | A kind of joining method of UAV system EO-1 hyperion linear array remote sensing image | |
| CN111693025B (en) | Remote sensing image data generation method, system and equipment | |
| CN103810706B (en) | A kind of roughness of ground surface participates in the remote sensing images inverted stereo bearing calibration of shadow model | |
| CN108335261B (en) | A kind of Optical remote satellite orthography garland region automatic testing method | |
| US20220020178A1 (en) | Method and system for enhancing images using machine learning | |
| CN102938145A (en) | Consistency regulating method and system of splicing panoramic picture | |
| CN108230326A (en) | Satellite image garland based on GPU-CPU collaborations deforms rapid detection method | |
| CN116245757B (en) | Multi-scene universal remote sensing image cloud restoration method and system for multi-mode data | |
| CN104202538A (en) | Double-registration method for different-exposure images in wide dynamic camera | |
| CN111340895A (en) | Image color uniformizing method based on pyramid multi-scale fusion | |
| CN112016478A (en) | Complex scene identification method and system based on multispectral image fusion | |
| CN104700356B (en) | A Method of Anti-Stereo Correction for Remote Sensing Image |
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 | ||
| CP03 | Change of name, title or address |
Address after: No. 1299, Mingxi Road, Beihu science and Technology Development Zone, Changchun City, Jilin Province Patentee after: Changguang Satellite Technology Co.,Ltd. Address before: 130032 No. 1759, Mingxi Road, Gaoxin North District, Changchun City, Jilin Province Patentee before: CHANG GUANG SATELLITE TECHNOLOGY Co.,Ltd. |
|
| CP03 | Change of name, title or address | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: A Fast Splicing Method for Large Area Array Submeter Level Night Scene Remote Sensing Images Granted publication date: 20200904 Pledgee: Jilin credit financing guarantee Investment Group Co.,Ltd. Pledgor: Changguang Satellite Technology Co.,Ltd. Registration number: Y2024220000032 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right |