CN106327452B - A kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area - Google Patents
A kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area Download PDFInfo
- Publication number
- CN106327452B CN106327452B CN201610665330.1A CN201610665330A CN106327452B CN 106327452 B CN106327452 B CN 106327452B CN 201610665330 A CN201610665330 A CN 201610665330A CN 106327452 B CN106327452 B CN 106327452B
- Authority
- CN
- China
- Prior art keywords
- image
- fragmentation
- data
- data collection
- cloudless
- 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
- 238000013467 fragmentation Methods 0.000 title claims abstract description 57
- 238000006062 fragmentation reaction Methods 0.000 title claims abstract description 57
- 238000010189 synthetic method Methods 0.000 title claims abstract description 14
- 238000013480 data collection Methods 0.000 claims abstract description 65
- 238000000034 method Methods 0.000 claims abstract description 57
- 230000005855 radiation Effects 0.000 claims abstract description 50
- 238000012937 correction Methods 0.000 claims abstract description 30
- 238000009826 distribution Methods 0.000 claims abstract description 22
- 230000008569 process Effects 0.000 claims abstract description 20
- AGVAZMGAQJOSFJ-WZHZPDAFSA-M cobalt(2+);[(2r,3s,4r,5s)-5-(5,6-dimethylbenzimidazol-1-yl)-4-hydroxy-2-(hydroxymethyl)oxolan-3-yl] [(2r)-1-[3-[(1r,2r,3r,4z,7s,9z,12s,13s,14z,17s,18s,19r)-2,13,18-tris(2-amino-2-oxoethyl)-7,12,17-tris(3-amino-3-oxopropyl)-3,5,8,8,13,15,18,19-octamethyl-2 Chemical compound [Co+2].N#[C-].[N-]([C@@H]1[C@H](CC(N)=O)[C@@]2(C)CCC(=O)NC[C@@H](C)OP(O)(=O)O[C@H]3[C@H]([C@H](O[C@@H]3CO)N3C4=CC(C)=C(C)C=C4N=C3)O)\C2=C(C)/C([C@H](C\2(C)C)CCC(N)=O)=N/C/2=C\C([C@H]([C@@]/2(CC(N)=O)C)CCC(N)=O)=N\C\2=C(C)/C2=N[C@]1(C)[C@@](C)(CC(N)=O)[C@@H]2CCC(N)=O AGVAZMGAQJOSFJ-WZHZPDAFSA-M 0.000 claims abstract description 11
- 238000010606 normalization Methods 0.000 claims description 47
- 230000015572 biosynthetic process Effects 0.000 claims description 27
- 238000003786 synthesis reaction Methods 0.000 claims description 27
- 239000013598 vector Substances 0.000 claims description 21
- 230000003595 spectral effect Effects 0.000 claims description 18
- 238000001228 spectrum Methods 0.000 claims description 18
- 239000012634 fragment Substances 0.000 claims description 14
- 230000002708 enhancing effect Effects 0.000 claims description 12
- 230000004927 fusion Effects 0.000 claims description 9
- 239000012141 concentrate Substances 0.000 claims description 8
- 238000005192 partition Methods 0.000 claims description 8
- 238000007476 Maximum Likelihood Methods 0.000 claims description 6
- 238000012417 linear regression Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 6
- 238000013139 quantization Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000005303 weighing Methods 0.000 claims description 2
- 239000004744 fabric Substances 0.000 claims 1
- 230000001568 sexual effect Effects 0.000 claims 1
- 230000008901 benefit Effects 0.000 abstract description 4
- 238000001514 detection method Methods 0.000 description 19
- 230000000694 effects Effects 0.000 description 14
- 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 9
- 238000013329 compounding Methods 0.000 description 8
- 238000000746 purification Methods 0.000 description 8
- 238000005516 engineering process Methods 0.000 description 5
- 238000005286 illumination Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 4
- 230000011218 segmentation Effects 0.000 description 4
- 230000004075 alteration Effects 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000007796 conventional method Methods 0.000 description 2
- 238000013502 data validation Methods 0.000 description 2
- 238000003384 imaging method Methods 0.000 description 2
- 238000002310 reflectometry Methods 0.000 description 2
- 230000002194 synthesizing effect Effects 0.000 description 2
- 230000002123 temporal effect Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 206010012373 Depressed level of consciousness Diseases 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000003702 image correction Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 230000002045 lasting effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 230000003014 reinforcing effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000001308 synthesis method Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a kind of fragmentation remote sensing image synthetic methods towards cloudy rain area, comprising the following steps: obtains more stars in cloudy rain area, the original image data collection of multidate；The radiation difference correction of sensor and extrinsicfactor is carried out to obtain the consistent first image data collection of radiation feature to original image data collection；The cloud layer and its shade distribution for obtaining the first image data collection obtain the cloudless fragmentation of data collection in cloudy rain area by mask process；Reference data subset is chosen from cloudless fragmentation of data collection, absent region is studied for a second time courses one has flunked using other cloudless fragmentation of data collection, obtains the complete cloudless resultant image in cloudy rain area.This method takes full advantage of the previous high cloud amount image for being considered as hash and is synthesized, the remote sensing observations space-time covering frequence in cloudy rain area is effectively increased, powerful data supporting especially is provided to the demanding application of timeliness for the remote sensing application based on domestic satellite data.
Description
Technical field
The present invention relates to remote sensing image field, in particular to a kind of fragmentation remote sensing image synthesis towards cloudy rain area
Method and device.
Background technique
East China and southern province area by maritime monsoon or the torrid zone, subtropical climatic condition due to being influenced, often
Year cloudy rainy day gas, especially clear sky weather is more rare in crop (vegetation) Growing season, and big face is shown as on remote sensing image
Long-pending cloud shadow covering, since cloud cover causes earth's surface information that can not be received by remote sensor, so that depending on optical satellite
The remote sensing applications such as resource investigation, ecological environmental protection and the crop remote sensing monitoring of remote sensing can not obtain the cloudless image of crucial phase
Data.With being constantly progressive for China's satellite technology, the satellite sequential transmissions of a variety of different spatial resolutions and temporal resolution
And come into operation, satellite frequency of passing by greatly improves, to realize that the lasting observation of earth's surface information provides potentiality.In this background
Under, how from more stars, multidate to have that excavate useful information in cloud image as far as possible to serve remote sensing application be that current remote sensing is answered
With the problem of middle urgent need to resolve.
Existing cloud removing technology is based primarily upon single type sensor, such as the multi-temporal remote sensing shadow of Landsat satellite
Picture selects the 1 phase cloudless image of the same area as with reference to base map, by the radiation normalization and cloud detection for having cloud image,
To have cloud image and there is relatively uniform radiation feature simultaneously with reference to image, is calculating cloud distribution, and then utilize
The cloudless pixel with reference in image is to there is the obnubilation cover area in cloud image to be replaced, to achieve the purpose that cloud.Technology
Process as shown in Figure 1, specifically includes the following steps:
First to having cloud image and carrying out relative radiometric normalization with reference to image, make the two that there is approximate hue and luminance
Equal radiation features.Relative radiometric normalization weights Multivariate alteration detection converter technique (Iteration Re-weight using iteration
Multivariate Alteration Detection, abbreviation IR-MAD), i.e., it is special using the Scale invariant of Multivariate alteration detection
Property obtain the constant pixel of radiation in same two image of region (also known as pseudo- invariant features, such as building, road, desert, with plant
Different, these ground classes have relatively stable reflectivity, do not change at any time generally)；Based on these pseudo- invariant features
Pixel pair, being calculated using orthogonal regression method is had cloud image and with reference to the regression equation of each wave band of image, then utilizes the recurrence
Equation by wave band carries out relative radiometric normalization to there is cloud image.
Then, cloud image carries out automatic cloud using the method that Threshold segmentation is combined with cluster to after radiation normalization
Detection, determines cloud distribution.Since cloud layer is high reflection in each wave band of image, highlight regions are shown as, with other
Object difference is obvious, carries out cloud detection using this characteristic of cloud layer.Specific steps are as follows: 1. by the similar pixel of Cluster merging come complete
At image block；2. marking off the kind in cloud sector in image using threshold value according to the brightness value distributed area in cloud sector and atural object pixel
The seed region of subregion and atural object completes preliminary cloud detection.3. being utilized again based on the seed region of threshold method detection
The thought of cluster carries out refinement clustering recognition to the region not yet identified.
Finally, returning the method combined with cloud sector domain morphological dilations using linear, image cloud sector domain is carried out
Replacement is filled up, and cloud effect is played.Detailed process is as follows: the cloudless of areal different times is referred to image and target figure by 1.
Corresponding cloudless region progress linear regression fit is obtained as in further slackening the SPECTRAL DIVERSITY in two images
Brightness to reference image normalizes image.2. using flat disc structure to there is the cloud sector domain in cloud image to expand, with
Brightness normalization image carries out fusion replacement, finally obtains Scattered Clouds Or Better image.
Prior art disadvantage are as follows:
1. the prior art is filled up merely with the replacement that the multi-temporal data of single satellite source carries out cloud image mostly,
It is difficult to meet the application demand of higher timeliness in data abundance.
2. the prior art finds image by using distinct methods in the relative radiometric normalization processing of multidate image
The pseudo- invariant features point of centering establishes regression equation, carries out globalization correction to whole scape image, and there is no to different sensors sheet
The external factor difference such as the radiation difference of body and illumination, atmospheric conditions distinguishes, and does not also account for the spoke between different land types
Penetrate property difference.
3. the prior art is realized with reference to image to the substitute cloud removing of target image based on cloudless, wherein as referring to
The acquisition of cloudless image is the prerequisite of this method.However in the cloudy rain area in south, to meet cloudless image covering very
Difficulty limits effective application of this method.
4. the prior art only realizes the detection to cloud distribution using the cloud detection method of optic based on cluster, and for distant
For feeling image, cloud layer and its shadow region are usually an important factor for causing data information to lack.This method can not be to by cloud
Effectively identification and substitute are realized in caused shadow region, this puts the application effect for the method that also limits.
The information disclosed in the background technology section is intended only to increase the understanding to general background of the invention, without answering
When being considered as recognizing or imply that the information constitutes the prior art already known to those of ordinary skill in the art in any form.
Summary of the invention
The purpose of the present invention is to provide a kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area,
To overcome the disadvantage that the abundance of prior art multi-temporal data is lower, radiation characteristic is poor.
Another object of the present invention is to provide a kind of fragmentation remote sensing image synthetic method towards cloudy rain area and
Device causes lacking for data information missing with reference to image capturing hardly possible, cloud layer and its shadow region to overcome the prior art cloudless
Point.
To achieve the above object, according to an aspect of the present invention, a kind of fragmentation remote sensing towards cloudy rain area is provided
Image synthesis method, comprising the following steps:
S101: more stars in cloudy rain area, the original image data collection of multidate are obtained；
S102: the radiation difference correction of sensor and extrinsicfactor is carried out to obtain radiation to the original image data collection
The consistent first image data collection of feature；
S103: the cloud layer and its shade distribution of the first image data collection are obtained, institute is obtained by mask process
State the cloudless fragmentation of data collection in cloudy rain area；
S104: reference data subset is chosen from the cloudless fragmentation of data collection, utilizes other cloudless fragmentation of data collection
Absent region is studied for a second time courses one has flunked, the complete cloudless resultant image in the cloudy rain area is obtained.
To achieve the above object, according to a further aspect of the invention, it is distant to provide a kind of fragmentation towards cloudy rain area
Feel image synthesizing device, specifically include:
Data acquisition module, for obtaining more stars in cloudy rain area, the original image data collection of multidate；
Data correction module, for carrying out the radiation difference school of sensor and extrinsicfactor to the original image data collection
Just to obtain the consistent first image data collection of radiation feature；
Data mask module, for obtaining the cloud layer and its shade distribution of the first image data collection, by covering
Film process obtains the cloudless fragmentation of data collection in the cloudy rain area；
Module is integrated in repairing, for choosing reference data subset from the cloudless fragmentation of data collection, utilizes other nothings
Cloud fragmentation of data collection studies for a second time courses one has flunked absent region, obtains the complete cloudless resultant image in the cloudy rain area.
Compared with prior art, the invention has the following beneficial effects:
The present invention makes full use of cloud shadow image in more stars, multi-temporal data commbined foundations, leads to especially in conventional treatment
The useful pixel region in high cloud amount image being often rejected realizes cloudless shadow by the Land use systems of fragmentation valid data
As synthesis, the remote sensing observations space-time coverage in cloudy rain area is effectively improved.
The present invention considers between sensor, external factor and different land types in relative radiometric normalization processing
Radiation characteristic difference carries out automanual high-precision radiant correction by atural object classification.Meanwhile all having to cloud layer and its shade
Effect identification is extracted, and the deficiency only detected to cloud layer in conventional method is overcome, and in advance without the complete of target area
It is cloudless to be used as substitute data source with reference to image, the application effect and application range of method are improved, to promote China is domestic to defend
The landing application of sing data and the south China remote sensing monitoring application in cloudy rain area throughout the year provide strong data and support.
Other features and advantages of the present invention will be illustrated in the following description, also, partly becomes from specification
It obtains it is clear that understand through the implementation of the invention.The objectives and other advantages of the invention can be by written explanation
Specifically noted structure is achieved and obtained in book, claims and attached drawing.
Below by drawings and examples, technical scheme of the present invention will be described in further detail.
Detailed description of the invention
Attached drawing is used to provide further understanding of the present invention, and constitutes part of specification, with reality of the invention
It applies example to be used to explain the present invention together, not be construed as limiting the invention.In the accompanying drawings:
Fig. 1 is the flow chart of existing remote sensing image synthetic method.
Fig. 2 is the flow chart of the fragmentation remote sensing image synthetic method according to the present invention towards cloudy rain area.
Fig. 3 is multi-source multidate image relative radiometric normalization flow chart according to the present invention.
Fig. 4 a is Landsat8-OLI reference sensor clear sky image, and Fig. 4 b is GF1-WFV1 target (wait correct) sensor
Clear sky image.
Fig. 5 a is with reference to Landsat8 image, and Fig. 5 b is GF1-WFV1 image to be corrected, and Fig. 5 c is that IR method radiates normalizing
Change image, Fig. 5 d is context of methods radiation normalization image.
Fig. 6 is the flow chart that the detection of cloud shadow is generated with image valid data according to the present invention.
Fig. 7 a and Fig. 7 b is that there are cloud image and cloudless image in areal according to the present invention.
Fig. 8 a is cloud layer reinforcing effect, and Fig. 8 b is shadow enhancement effect, Fig. 8 c cloud layer region binarization segmentation effect, Fig. 8 d
Shadow region binarization segmentation effect.
Fig. 9 a is raw video, Fig. 9 b cloud shadow testing result, Fig. 9 c image valid data, and black portions are countless in Fig. 9 c
According to region.
Figure 10 is the Image compounding flow chart based on valid data according to the present invention.
Figure 11 a- Figure 11 h is the cloudy image for studying area according to the present invention.
Figure 12 is the Image compounding effect based on valid data according to the present invention.
Figure 13 is the structure chart of the fragmentation remote sensing image synthetic method according to the present invention towards cloudy rain area.
Figure 14 is the structure chart of Data correction module according to the present invention.
Figure 15 is the structure chart of coefficient acquisition submodule according to the present invention.
Figure 16 is the structure chart of the first correction module according to the present invention.
Figure 17 is the structure chart of data mask module according to the present invention.
Figure 18 is the structure chart that module is integrated in repairing according to the present invention.
Specific embodiment
With reference to the accompanying drawing, specific embodiments of the present invention will be described in detail, it is to be understood that guarantor of the invention
Shield range is not limited by the specific implementation.
In conventional image process flow, the image generally by cloud amount greater than 50% is considered as useless image data and gives up,
So as to cause the missing of image element information effective in the cloudless region of image.Difficulty is obtained for southern cloudy rain area satellite image,
It is difficult to meet the application problem of a wide range of cloudless image covering, the invention proposes more stars, the united methods of multidate image to mention
High data cover degree, and Image compounding is carried out by the way of based on the cloudless valid data of fragmentation, thus with meeting cloudy rain
Application demand of the area to cloudless image data.
As shown in Fig. 2, a kind of fragmentation remote sensing image towards cloudy rain area of specific embodiment according to the present invention
Synthetic method, comprising the following steps:
Step S101: more stars in cloudy rain area, the original image data collection of multidate are obtained；
Step S102: carrying out the radiation difference correction of sensor and extrinsicfactor to original image data collection, eliminates pass more
Because of radiance difference caused by sensor performance, altitude of the sun, Atmospheric Absorption and scattering etc. between sensor and multidate image,
Obtain the consistent first image data collection of radiation feature；
Step S103: the detection of cloud shadow is carried out to the image data of normalization via radiation, obtains the first image data collection
Cloud layer and its shade distribution, the cloudless fragmentation of data collection in cloudy rain area is obtained by mask process；
Step S104: the requirement according to practical application to sensor of interest and center phase is selected from cloudless fragmentation of data collection
Reference data subset is taken, absent region is studied for a second time courses one has flunked using other cloudless fragmentation of data collection, obtains the complete of cloudy rain area
Cloudless resultant image.
The present invention is by relative radiometric normalization, the detection of automation cloud shadow, the Image compounding based on fragmentation valid data
Etc. key technologies, Cong Duoxing, having for multidate extract available pixel region in cloud image data, make full use of each pixel
Information synthesizes the cloudless image in certain time window, is provided with force data support for the remote sensing monitoring application in cloudy rain area.
Above-mentioned steps S102 is as shown in figure 3, specifically include:
Step S201: being based on clear sky image, obtains sensor spectrum normalization coefficient in such a way that classification returns；Specifically
Are as follows:
In this step, on the basis of being selected as the respective sensor with reference to image, other sensors are target to be corrected, choosing
Take the two that there is the clear sky image of overlapping region and phase close (being no more than 1 week) to relative correction is carried out, at this time it is believed that greatly
Gas and ground mulching variation influence are small, and the radiation difference between two images is mainly from sensor itself.The spectrum of sensor is returned
One, which changes process, has relative independentability, therefore being chosen in spatial and temporal distributions for clear sky image pair can be different from image number to be corrected
According to collection.
Firstly, carrying out radiation calibration to image, by DN value, (DN value (Digital Number) is remote sensing image picture element brightness
Value, the gray value of the atural object of record) spoke brightness is converted to, keep different images pixel value dimension having the same horizontal, eliminates and pass
Influence of quantization series (pixel locating depth) difference to fitting precision between sensor.
Secondly, the overlay region image of the clear sky image of other sensors and the clear sky image of reference sensor is carried out respectively
The supervised classification of the major class such as vegetation, settlement place, bare area, water body, and chosen automatically at random in the class of various regions respectively sufficient amount of
Sample point；
Finally, according to the sample point set of acquisition, in two images different-waveband and classification establish equation of linear regression,
Seek regression coefficient i.e. spectral normalization coefficient.
(usually 1-2) performance is relatively stable within a certain period of time for satellite sensor, therefore, the spectrum normalizing of sensor
Constant can be regarded as within the period by changing coefficient, and is used in the form of a lookup table, to improve parameter reusability and applicable model
It encloses.Table 1 is the regression model for carrying out sensor calibration to GF-WFV1 on the basis of Landsat8-LIO sensor, regression equation
Coefficient is the spectral normalization coefficient of accordingly class.
GF-WFV1 regression model of the table 1 on the basis of Landsat8-LIO
Step S202: being based on sensor spectrum normalization coefficient, using the sample method classified again of transmitting to more stars, it is more when
The original image data collection of phase carries out semi-automatic classification, carries out sensor radiation difference correction to sorted image data collection；
In the step, in Classification in Remote Sensing Image, sample quality is closely related with nicety of grading, for the height for guaranteeing sample information
" fidelity " carries out the selection of sample generally from image to be sorted.When same sample is applied to more scape images, due to atmosphere
The difference of situation, illumination and phase be easy to cause the phenomenon that different spectrum of jljl, same object different images, brings larger uncertainty to classification.
The characteristics of in conjunction with timing image to the same area continuous observation, proposes a kind of semi-automatic classification method based on sample transmitting,
The candidate samples position that the classification spatial position that early period, classification obtained is classified as next phase, and again for new phase image
The calculating of sample characteristics is carried out, optimizes and classifies again.Therefore for multi-source, multidate image, need to only one be carried out to reference image
The supervised classification of secondary artificial sampling can realize the automatic assorting process of full dataset.
It is specific as follows:
Firstly, the magnitude of the image picture element of original image data collection and spectral normalization coefficients match is fixed radiate
Mark；Data basis of the calibration results as subsequent processes.
Image is referred to secondly, choosing the optimal image of the quality of data from calibration image sequence and being used as, remaining is then wait entangle
Positive image, and sample selection and maximum likelihood classification are carried out to reference image, obtain the classification results for referring to image.
Further, on the basis of current reference image and its classification results, image is corrected to next expectation and carries out sample sieve
Choosing and Sample purification are to obtain universal class very originally；
Maximum likelihood classification is carried out finally, very originally treating using universal class and correcting image, and according to the type of sensor, it is right
The corresponding pixel set of image to be corrected of all categories carries out sensor spectrum normalization.
The automatic classification transmitted by sample effectively increases the place of multi-source, multidate image classification and radiation normalization
Manage efficiency.Wherein, the reapective features that screening sample and Sample purification combine forward and backward two phases image carry out refining for sample, right
It reduces computation complexity and improves again nicety of grading and play a crucial role.
(1) screening sample
Screening sample is to calculate and obtain from image overlap area and is each based on early period is with reference to image and its classification results
The representative Pixel domain position collection of ground class, as next candidate samples spatial position for expecting to correct image.It is general and
Speech, similar atural object have similar spectral signature, are in integrated distribution in the n dimension spectral space that n image wave band is constituted, from classification
The closer pixel in center has higher representativeness.The method of screening sample is to obtain current reference image first and wait correct
The overlay region range of image is constrained in overlay region using classifying image as class scope, refers to image by classification to when early period
Calculating averaged spectrum vector ties up the class center of spectral space as such in n；Then, it is measurement with Euclidean distance, calculates every
All pixel spectral vectors and obtain the m nearest apart from class center by sort algorithm to the distance of class center in one kind
Typical pixel sample of a pixel as such；Finally the corresponding set of spatial locations of all categories pixel sample is passed to wait entangle
Positive image, the spatial distribution as candidate samples.The selection of m value wants depending on the circumstances, if overlay region same ground species
The form of expression of type is complex, can use larger m value to obtain sufficient amount of sample pixel.This paper m value takes corresponding classification picture
The 20% of first sum.
In formula, ED is Euclidean distance of the pixel to class center, xi, yiRespectively represent pixel and class center in the i-th wave band
Gray value, n be image wave band number.
(2) Sample purification
Sample purification carries out recalculating for sample characteristics for image to be corrected, and rejects in candidate samples due to ground species
Noise pixel caused by type variation, Yun Ying covering etc., improves sample purity.Sample purification is carried out using variance method of purification herein.
It combines image to be corrected to read the pixel spectral vector of candidate samples position by class first, asks calculate every a kind of put down on this basis
Equal spectral vector xi；Secondly, the variance var (X) of spectral vector and averaged spectrum vector by each pixel in class solution class is (see public affairs
Formula 2), and mean of variance is asked to all pixel variances in classWherein var (X) describes the different of pixel and generic
Matter degree, value is smaller, shows that the difference of the pixel and classification is lower, andThen reflect being averaged for pixel gray value in classification
Dispersion degree can be used as the reference of the heterogeneous degree of pixel.Finally, taking threshold value for each ground classIn all kinds of
Var (X) > TvarPixel be considered as heterogeneous pixel and rejected, finally obtain the pure sample of all ground class this pixel set.
In formula, xiWithFor respectively pixel spectral vector and averaged spectrum vector in the gray value of i wave band, n is spectral vector
Dimension, i.e. image wave band number.
Step S203: on the basis of classification, the side of selection automatically the PIFs constrained based on NDVI difference value histogram and classification
Method realizes the radiation normalization of the extrinsicfactor of image data collection.
In the step, in conjunction with the sensor spectral normalization feature based on classification, propose based on NDVI difference value histogram
It is basic herein with pseudo- invariant features point (Pseudo Invariant Features, PIFs) automatically selecting method of classification constraint
On construct image to be corrected and the equation of linear regression with reference to each wave band in image, realize the radiation normalization treated and correct image
Correction.Shown in the calculating of NDVI and its difference such as formula (3), (4).
Δ NDVI=NDVIr-NDVIt (4)
In formula, R_nir and R_red are respectively image near-infrared and infrared band reflectance value；NDVIrAnd NDVItRespectively
With reference to the NDVI image of image and image to be corrected.
Image after sensor relative correction, the cities and towns more stable for reflectivity and bare area pixel be believed that its only by
The entire effect of illumination and atmosphere and even variation, therefore, no matter original image NDVI due to object spectrum multiplicity and in unimodal or
Multi-modal, cities and towns and bare area classification NDVI difference show as relatively stable and concentrate, and approximate can be indicated with normal distribution.
Stable radiation point is located near the mean μ of Δ NDVI histogram, and is located at distribution map two sides by the unstable point of noise jamming.It will
PIFs of the point as stable radiation within the scope of μ ± c σ, wherein σ is the standard deviation of Δ NDVI, and c is to determine stable point section
Constant, take c=1 herein.Using PIFs as sample point, the equation of linear regression such as formula (5) is established for each wave band, according to most
Small two multiply principle, calculate the optimal coefficient k of each wave bandi、bi, treat and correct image progress Volatile material.
pri=ki×pti+bi(i=1 ..., n) (5)
In formula, priAnd ptiRespectively indicate the i-th wave band with reference to image and image to be corrected, kiAnd biFor fitting coefficient, n is
Wave band number.
Fig. 4 a and Fig. 4 b are respectively Landsat8-OLI the and GF1-WFV1 two sensors in identical imaging time and region
Clear sky image, it can be seen that image compares obviously in the radiation characteristics such as contrast, brightness, reflects satellite sensor and is responding
There are systematic divergences for characteristic, band setting etc..Also illustrate to be divided into the relative radiometric normalization of image for biography simultaneously
The radiant correction of sensor itself and for the external factor such as illumination two processes of radiation normalization have necessity.Fig. 5 a and figure
5b be respectively on May 13rd, 2015 Landsat8-OLI and May 14 GF-WFV1 the same area raw video data, two
Person's radiation feature difference is obvious, chooses the higher Landsat8-OLI data of clarity as reference, carries out opposite spoke to the latter
Penetrate normalization.Fig. 5 c and Fig. 5 d are respectively using common the image Return Law (Image Regression, IR) and this paper
The calibration result of method, it is seen that radiation normalization method of the present invention can be more effectively to different sensors and phase
Radiation difference between image is corrected.
Further, as shown in fig. 6, step S103 is specifically included:
Wherein, cloud layer and its shadow region are the principal elements for causing image information to be lost, and method common at present is most
It is handled for cloud layer region and has ignored the detection to cloud shade.The present invention realizes automation, effectively inspection to the two
It surveys, and based on cloud shadow testing result to there is cloud image to carry out validation processing, obtaining the fragmentation only comprising available pixel has
Data are imitated, specifically:
Step S501: to the image data with cloud layer and its shade, based on colour space transformation enhancing cloud layer and its yin
Shadow obtains the cloud layer of enhancing and its partition threshold of shade；
The diffusing reflection feature that cloud layer is high with gray average in remote sensing image, variance is small, and cloud shadow then shows as part
The characteristics of low sum of the grayscale values in region low variance, therefore can be according to cloud layer and shade and ash of other atural objects on remote sensing image
Degree difference identification goes out thin cloud and shadow region.For the intensity contrast for further increasing cloud shadow Yu other atural objects, using being based on color
The cloud shadow detection method of spatial alternation, by the RGB band switching of multispectral image to YCbCr space, and respectively for cloud layer and
Shade carries out Imaging enhanced.Colour space transformation and the calculation formula of cloud shadow enhancing are as follows:
Shadow region enhances formula are as follows:
Is=(Cb+Cr)/Y (7)
Cloud layer region enhances formula:
Ih=Y/Is (8)
The thin cloud of enhancing and shadow character image are stretched to 0-255 range respectively, threshold value choosing is carried out using Otsu method
It takes.This method is based on maximum between-cluster variance thought, that is, chooses suitable threshold value T, so that following formula obtains maximum value:
σb(T)=ω1(T)ω2(T)[μ1(T)-μ2(T)]2 (9)
Wherein
ω1, μ1Respectively indicate percentage and mean value of the gray value 0 to pixel between T；ω2, μ2Respectively indicate gray value
In the percentage and mean value of T to 255 pixel.The segmentation threshold Th of two width images after Yun Ying enhancing is calculated separately according to Otsu method
And Ts.The region of corresponding [0, Th] ∩ [0, Ts] is cloudless shadow zone domain in two width images, by (Th, 255] determine cloud layer region,
(Ts, 255] determine shadow region.In the middle part of the Anhui somewhere have cloud image for cloud shadow detection method is tested
(see Fig. 7 a and Fig. 7 b), Yun Ying enhancing and extraction effect are as shown in Fig. 8 a- Fig. 8 d.
Step S502: carry out Mathematical Morphology expansion and vector quantization respectively to cloud layer and its shadow region with obtain cloud layer and its
The distribution of shade completes cloud shadow detection function.
Step S503: it is based on partition threshold, by the distribution vector of cloud layer and its shade and first image data
Collection carries out mask process to generate cloudless fragmentation data, as Fig. 9 a- Fig. 9 c can be seen that image data validation process is illustrated
Figure.
Finally, as shown in Figure 10, step S104 is specifically included:
In the step, is handled by above-mentioned radiation normalization and the detection of cloud shadow and data validation, form Image compounding
Valid data " raw material ".Based on valid data collection, combining target regional scope and the industry to phase and key data source
Business demand, is made iteratively Data Synthesis, ultimately generates the cloudless resultant image of target area in certain period, specific as follows:
Step S601: made according to the blank grid data that target area range vector generates predefined size and wave band number n
To synthesize base map (all pixels are NULL value)；
Step S602: being constraint with target area, and connected applications are broken from cloudless data to synthesis period and data source requirement
Piece concentrate it is preferential choose in the target area the maximum two pieces of image fragments of area accounting as reference data；
Step S603: carrying out the regression fit by wave band by all pixels in the overlay region to two pieces of image fragments,
It establishes model of fit between the two and is corrected, be consistent brightness and tone between the two, color balance is realized, into one
The two is carried out fusion and inlayed by step；
Step S604: fused data is copied in the synthesis base map of the blank in S601 by wave band；
Step S605: judging the area without data (pixel is NULL value in n wave band) whether base map still has vacant position,
The synthesis of image is completed if no vacancy, otherwise calculates the regional scope of vacancy, and the size according to area of absence is successively from institute
It states cloudless fragmentation of data and concentrates the image fragment for choosing the maximum area of corresponding region；
Step S606: repeating step S603~S605, until target area no data vacancy, finally realizes cloudless image
Synthesis.
It carries out realizing color balance by the way of regression fit using overlay region pixel in step S603, without using common
The methods of Histogram Matching is since Histogram Matching is more suitable for having the globalization color of the more whole picture image of class equal
Weighing apparatus, and the ground class and area in valid data fragment are smaller, while considering that data have been subjected to radiation normalization processing, therefore linear
Homing method is more suitable for the color balance in local cell domain.Figure 11 is that GF1-WFV has cloud image, and Figure 12 is to utilize the method for the present invention
Having cloud image based on GF1-WFV is 2.4 ten thousand Km to area2The effect picture of region progress Image compounding.
The present invention takes full advantage of the previous high cloud amount image for being considered as hash and synthesizes, and effectively increases cloudy rain
The remote sensing observations space-time covering frequence in area, is answered for the remote sensing application based on domestic satellite data is especially demanding to timeliness
With providing powerful data supporting.Simultaneously by the relative radiometric normalization of image data be divided into the correction for sensor differences with
And two processes of radiant correction for the external factor difference such as illumination, atmosphere, while considering the radiation characteristic of different land types
Difference carries out radiation normalization by atural object classification respectively, effectively increases radiation normalization effect.And it realizes while to cloud layer
And effective detection of cloud shade, overcome the deficiency only identified to cloud layer region in common methods.It can be not required in advance cloudless
With reference to the support of image, Image compounding is carried out in a manner of fragmentation, has widened the scope of application of method, is reduced and is applied difficulty.
On the other hand according to the present embodiment, as shown in figure 13, a kind of fragmentation remote sensing towards cloudy rain area is provided
Image synthesizing device specifically includes:
Data acquisition module 10, for obtaining more stars in cloudy rain area, the original image data collection of multidate；
Data correction module 20, for carrying out the radiation difference of sensor and extrinsicfactor to the original image data collection
Correction is to obtain the consistent first image data collection of radiation feature；
Data mask module 30 passes through for obtaining the cloud layer and its shade distribution of the first image data collection
Mask process obtains the cloudless fragmentation of data collection in the cloudy rain area；
Module 40 is integrated in repairing, for choosing reference data subset from the cloudless fragmentation of data collection, using other described
Cloudless fragmentation of data collection studies for a second time courses one has flunked absent region, obtains the complete cloudless resultant image in the cloudy rain area.
In the embodiment preferably, as shown in figure 14, Data correction module 20 specifically includes:
Coefficient acquisition submodule 201 obtains sensor spectrum normalization coefficient for being based on clear sky image；
First correction module 202, for being based on sensor spectrum normalization coefficient, to the initial shadow of more stars, multidate
As data set is classified, sensor radiation difference correction is carried out to sorted image data collection；
Second correction module 203, for being constrained based on NDVI difference value histogram and classification on the basis of classification
PIFs choosing method realizes the radiation normalization of the extrinsicfactor of image data collection.
In the embodiment preferably, as shown in figure 15, coefficient acquisition submodule 201 specifically includes:
Image scaling unit 2011, for choosing the clear sky image of reference sensor and other sensors, to the clear sky
Image is calibrated；
Sample classification unit 2012, for by the weight of the clear sky image of other sensors and the clear sky image of reference sensor
Folded area's image is classified, and chooses the sample point set of sorted plot image respectively；
Coefficient acquiring unit 2013 is directed to the clear sky image of other sensors for the sample point set according to acquisition respectively
Equation of linear regression is established with the different-waveband and classification of the clear sky image of reference sensor to obtain spectral normalization coefficient.
In the embodiment preferably, as shown in figure 16, the first correction module 202 specifically includes:
Image scaling unit 2021, for by the magnitude and spectral normalization of the image picture element of the original image data collection
Coefficients match is to carry out radiation calibration；
Image classification unit 2022 refers to image for choosing, remaining is image to be corrected, and carries out sample to reference image
It chooses and maximum likelihood classification, acquisition refers to the classification results of image；
Purification unit 2023 is screened, for being corrected on the basis of current reference image and its classification results to next expectation
Image carries out screening sample and Sample purification to obtain universal class very originally；
Sensor calibration unit 2024 corrects image progress maximum likelihood classification for very originally treating using universal class, and
According to the type of sensor, sensor spectrum normalization is carried out to the corresponding pixel set of image to be corrected of all categories.
In the embodiment preferably, as shown in figure 17, data mask module 30 specifically includes:
Cloud layer enhances submodule 301, has cloud image data for obtaining, based on colour space transformation enhancing cloud layer and its yin
Shadow；
Threshold values acquisition submodule 302, for obtaining the cloud layer of enhancing and its partition threshold of shade；
Cloud layer acquisition submodule 303, for carrying out Mathematical Morphology expansion and vector quantization respectively to cloud layer and its shadow region
To obtain the distribution of cloud layer and its shade；
Data acquisition submodule 304, for be based on partition threshold, by the distribution vector of cloud layer and its shade with it is described
First image data collection carries out mask process to generate cloudless fragmentation data.
In the embodiment preferably, as shown in figure 18, repairing is integrated module 40 and is specifically included:
Submodule 401 is selected in region, for generating the sky of predefined size and wave band number n according to target area range vector
White raster data is as synthesis base map；
Selection of datum submodule 402 concentrates preferential choose from the cloudless fragmentation of data for being constraint with target area
The maximum two pieces of image fragments of area accounting are as reference data in the target area；
Submodule 403 is inlayed in fusion, for being carried out by all pixels in the overlay region to two pieces of image fragments by wave
The regression fit of section, establishes model of fit between the two and is corrected, be consistent brightness and tone between the two with reality
The two is further carried out fusion and inlayed by existing color balance；
First synthesis submodule 404, for fused data to be copied in blank synthesis base map by wave band；
Second synthesis submodule 405, the area without data whether still having vacant position for judging base map is completed if no vacancy
The synthesis of image, on the contrary the regional scope of vacancy is calculated, and the size according to area of absence is successively from the cloudless fragmentation of data
Concentrate the image fragment for choosing the maximum area of corresponding region；
Wherein, it repeats the fusion and inlays submodule, the first synthesis submodule and the second synthesis submodule
Movement, until target area no data vacancy, realizes the synthesis of cloudless image.
To sum up, the present invention makes full use of cloud shadow image, especially at routine in more stars, multi-temporal data commbined foundations
The useful pixel region in high cloud amount image being usually rejected in reason is realized by the Land use systems of fragmentation valid data
Cloudless Image compounding effectively improves the remote sensing observations space-time coverage in cloudy rain area.
The present invention considers between sensor, external factor and different land types in relative radiometric normalization processing
Radiation characteristic difference carries out automanual high-precision radiant correction by atural object classification.Meanwhile all having to cloud layer and its shade
Effect identification is extracted, and the deficiency only detected to cloud layer in conventional method is overcome, and in advance without the complete of target area
It is cloudless to be used as substitute data source with reference to image, the application effect and application range of method are improved, to promote China is domestic to defend
The landing application of sing data and the south China remote sensing monitoring application in cloudy rain area throughout the year provide strong data and support.
The present invention can there are many various forms of specific embodiments, above by taking Fig. 1-Figure 18 as an example in conjunction with attached drawing to this hair
Bright technical solution gives an example, this is not meant to that specific example applied by the present invention can only be confined to specific process
Or in example structure, those skilled in the art are it is to be appreciated that specific embodiment presented above is a variety of
Some examples in preferred usage, any embodiment for embodying the claims in the present invention should all be wanted in technical solution of the present invention
Within the scope of asking protection.
Finally, it should be noted that the foregoing is only a preferred embodiment of the present invention, it is not intended to restrict the invention,
Although the present invention is described in detail referring to the foregoing embodiments, for those skilled in the art, still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features.
All within the spirits and principles of the present invention, any modification, equivalent replacement, improvement and so on should be included in of the invention
Within protection scope.
Claims (8)
1. a kind of fragmentation remote sensing image synthetic method towards cloudy rain area, which comprises the following steps:
S101: more stars in cloudy rain area, the original image data collection of multidate are obtained；
S102: the radiation difference correction of sensor and extrinsicfactor is carried out to obtain radiation feature to the original image data collection
Consistent first image data collection；
S103: the cloud layer and its shade distribution of the first image data collection are obtained, is obtained by mask process described more
The cloudless fragmentation of data collection in sexual intercourse area；
S104: reference data subset is chosen from the cloudless fragmentation of data collection, using other cloudless fragmentation of data collection to scarce
It loses region to be studied for a second time courses one has flunked, obtains the complete cloudless resultant image in the cloudy rain area；
Step S104 is specifically included:
S601: the blank grid data of predefined size and wave band number n are generated as synthesis bottom according to target area range vector
Figure；
S602: being constraint with target area, concentrates preferential selection area accounting in the target area from the cloudless fragmentation of data
Maximum two pieces of image fragments are as reference data；
Both S603: carrying out the regression fit by wave band by all pixels in the overlay region to two pieces of image fragments, establish
Between model of fit and be corrected, be consistent brightness and tone between the two, realize color balance, further by the two
Fusion is carried out to inlay；
S604: fused data is copied in the synthesis base map of the blank in S601 by wave band；
S605: judging the area without data whether base map still has vacant position, the synthesis of image to be completed if no vacancy, otherwise calculates empty
Scarce regional scope, and the size according to area of absence successively concentrates the maximum for choosing corresponding region from the cloudless fragmentation of data
The image fragment of area；
S606: repeating step S603~S605, until target area no data vacancy, finally realizes the synthesis of cloudless image.
2. the fragmentation remote sensing image synthetic method according to claim 1 towards cloudy rain area, which is characterized in that step
Rapid S102 is specifically included:
S201: being based on clear sky image, obtains sensor spectrum normalization coefficient；
S202: be based on sensor spectrum normalization coefficient, classify to the original image data collection of more stars, multidate, to point
Image data collection after class carries out sensor radiation difference correction；
S203: on the basis of classification, based on the PIFs choosing method that NDVI difference value histogram and classification constrain, image number is realized
According to the radiation normalization of the extrinsicfactor of collection.
3. the fragmentation remote sensing image synthetic method according to claim 2 towards cloudy rain area, which is characterized in that step
Rapid S201 is specifically included:
S301: the clear sky image of reference sensor and other sensors is chosen, the clear sky image of the two is calibrated；
S302: the overlay region image of the clear sky image of other sensors and the clear sky image of reference sensor is classified, point
The sample point set of sorted plot image is not chosen；
S303: according to the sample point set of acquisition, the clear sky image of clear sky image and reference sensor for other sensors
Different-waveband and classification establish equation of linear regression to obtain spectral normalization coefficient.
4. the fragmentation remote sensing image synthetic method according to claim 2 towards cloudy rain area, which is characterized in that step
Rapid S202 is specifically included:
S401: the magnitude of the image picture element of the original image data collection and spectral normalization coefficients match is fixed radiate
Mark；
S402: choosing and refer to image, remaining is image to be corrected, and carries out sample selection and maximum likelihood classification to reference image,
Obtain the classification results for referring to image；
S403: on the basis of current reference image and its classification results, image is corrected to next expectation and carries out screening sample and sample
This purifying is to obtain universal class very originally；
S404: it is very originally treated using universal class and corrects image progress maximum likelihood classification, and according to the type of sensor, to all kinds of
The corresponding pixel set of image not corrected carries out sensor spectrum normalization.
5. the fragmentation remote sensing image synthetic method according to claim 1 towards cloudy rain area, which is characterized in that step
Rapid S103 is specifically included:
S501: acquisition has cloud image data, based on colour space transformation enhancing cloud layer and its shade；
S502: the cloud layer of enhancing and its partition threshold of shade are obtained；
S503: Mathematical Morphology expansion and vector quantization are carried out respectively to cloud layer and its shadow region and is divided with obtaining cloud layer and its shade
Cloth range；
S504: it is based on partition threshold, the distribution vector of cloud layer and its shade and the first image data collection are covered
Film process is to generate cloudless fragmentation data.
6. a kind of fragmentation remote sensing image synthesizer towards cloudy rain area, which is characterized in that specifically include:
Data acquisition module, for obtaining more stars in cloudy rain area, the original image data collection of multidate；
Data correction module, for the original image data collection is carried out the radiation difference correction of sensor and extrinsicfactor with
Obtain the consistent first image data collection of radiation feature；
Data mask module, for obtaining the cloud layer and its shade distribution of the first image data collection, at exposure mask
Reason obtains the cloudless fragmentation of data collection in the cloudy rain area；
Module is integrated in repairing, for choosing reference data subset from the cloudless fragmentation of data collection, utilizes other cloudless numbers
Absent region is studied for a second time courses one has flunked according to fragment collection, obtains the complete cloudless resultant image in the cloudy rain area；
The repairing is integrated module and is specifically included:
Submodule is selected in region, for generating the blank grid of predefined size and wave band number n according to target area range vector
Data are as synthesis base map；
Selection of datum submodule is concentrated from the cloudless fragmentation of data for being constraint with target area and is preferentially chosen at target
The maximum two pieces of image fragments of area accounting are as reference data in region；
Submodule is inlayed in fusion, for carrying out the recurrence by wave band by all pixels in the overlay region to two pieces of image fragments
Fitting, establishes model of fit between the two and is corrected, and is consistent brightness and tone between the two to realize that color is flat
The two is further carried out fusion and inlayed by weighing apparatus；
First synthesis submodule, for fused data to be copied in blank synthesis base map by wave band；
Second synthesis submodule, the area without data whether still having vacant position for judging base map, completing image if no vacancy
Synthesis, on the contrary the regional scope of vacancy is calculated, and the size according to area of absence successively concentrates choosing from the cloudless fragmentation of data
Take the image fragment of the maximum area of corresponding region；
Wherein, the movement that submodule, the first synthesis submodule and the second synthesis submodule are inlayed in the fusion is repeated,
Until target area no data vacancy, realizes the synthesis of cloudless image.
7. the fragmentation remote sensing image synthesizer according to claim 6 towards cloudy rain area, which is characterized in that institute
Data correction module is stated to specifically include:
Coefficient acquisition submodule obtains sensor spectrum normalization coefficient for being based on clear sky image；
First correction module, for being based on sensor spectrum normalization coefficient, to the original image data collection of more stars, multidate
Classify, sensor radiation difference correction is carried out to sorted image data collection；
Second correction module, on the basis of classification, the PIFs constrained based on NDVI difference value histogram and classification to be chosen
Method realizes the radiation normalization of the extrinsicfactor of image data collection.
8. the fragmentation remote sensing image synthesizer according to claim 6 towards cloudy rain area, which is characterized in that institute
Data mask module is stated to specifically include:
Cloud layer enhances submodule, has cloud image data for obtaining, based on colour space transformation enhancing cloud layer and its shade；
Threshold values acquisition submodule, for obtaining the cloud layer of enhancing and its partition threshold of shade；
Cloud layer acquisition submodule, for carrying out Mathematical Morphology expansion and vector quantization respectively to cloud layer and its shadow region to obtain cloud
The distribution of layer and its shade；
Data acquisition submodule, for being based on partition threshold, by the distribution vector of cloud layer and its shade and first shadow
As data set carries out mask process to generate cloudless fragmentation data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610665330.1A CN106327452B (en) | 2016-08-14 | 2016-08-14 | A kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610665330.1A CN106327452B (en) | 2016-08-14 | 2016-08-14 | A kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106327452A CN106327452A (en) | 2017-01-11 |
| CN106327452B true CN106327452B (en) | 2019-05-07 |
Family
ID=57739354
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610665330.1A Active CN106327452B (en) | 2016-08-14 | 2016-08-14 | A kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106327452B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107240084B (en) * | 2017-06-14 | 2021-04-02 | 湘潭大学 | Method and device for removing rain from single image |
| CN109671038B (en) * | 2018-12-27 | 2023-04-28 | 哈尔滨工业大学 | Relative radiation correction method based on pseudo-invariant feature point classification layering |
| CN109859118B (en) * | 2019-01-03 | 2020-10-13 | 武汉大学 | Method and system for removing cloud coverage area through effective mosaic polygon optimization based on quadtree |
| CN109934788B (en) * | 2019-03-22 | 2022-12-30 | 鲁东大学 | Remote sensing image missing data restoration method based on standard remote sensing image |
| CN111402162B (en) * | 2020-03-13 | 2023-11-07 | 北京华云星地通科技有限公司 | Method for processing clear sky data set of satellite remote sensing image |
| CN112017177B (en) * | 2020-09-08 | 2023-12-01 | 科大讯飞股份有限公司 | Reference image determining method, device, equipment and storage medium |
| CN112180406B (en) * | 2020-09-17 | 2022-07-15 | 广东省科学院广州地理研究所 | Method, device and equipment for monitoring data integrity of satellite image |
| CN113077458B (en) * | 2021-04-25 | 2023-09-19 | 北京艾尔思时代科技有限公司 | Cloud and shadow detection method and system in remote sensing image |
| CN113223040B (en) * | 2021-05-17 | 2024-05-14 | 中国农业大学 | Banana estimated yield method and device based on remote sensing, electronic equipment and storage medium |
| CN115410096B (en) * | 2022-11-03 | 2023-01-24 | 成都国星宇航科技股份有限公司 | Satellite remote sensing image multi-scale fusion change detection method, medium and electronic device |
| CN116109478B (en) * | 2023-04-10 | 2023-07-04 | 中国科学院空天信息创新研究院 | Remote sensing image mosaic processing method and device based on comparability |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004001659A1 (en) * | 2001-12-17 | 2003-12-31 | Digitalglobe Inc. | System, method, and apparatus for satellite remote sensing |
| CN102708307A (en) * | 2012-06-26 | 2012-10-03 | 上海大学 | Vegetation index construction method applied to city |
-
2016
- 2016-08-14 CN CN201610665330.1A patent/CN106327452B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2004001659A1 (en) * | 2001-12-17 | 2003-12-31 | Digitalglobe Inc. | System, method, and apparatus for satellite remote sensing |
| CN102708307A (en) * | 2012-06-26 | 2012-10-03 | 上海大学 | Vegetation index construction method applied to city |
Non-Patent Citations (8)
| Title |
|---|
| Automatic Recognition of Cloud Images by Using Visual Saliency Features;Xiangyun Hu 等;《IEEE Geoscience and Remote Sensing Letters》;20150831;第12卷(第8期);第1760-1764页 * |
| 地理时空大数据协同计算技术;骆剑承 等;《地球信息科学》;20160531;第18卷(第5期);正文第3.1节、图3 * |
| 多光谱遥感影像中云影区域的检测与修复;周伟 等;《遥感学报》;20120125;第16卷(第1期);第132-142页 * |
| 多星数据协同的地块尺度作物分类与面积估算方法研究;黄启厅 等;《地球信息科学》;20160531;第18卷(第5期);正文第2.2节 * |
| 多源多时相遥感影像相对辐射归一化方法研究;黄启厅 等;《地球信息科学》;20160531;第18卷(第5期);摘要、正文第2.1-2.2节 * |
| 骆剑承 等.地理时空大数据协同计算技术.《地球信息科学》.2016,第18卷(第5期),第590-598页. * |
| 黄启厅 等.多星数据协同的地块尺度作物分类与面积估算方法研究.《地球信息科学》.2016,第18卷(第5期),第708-717页. * |
| 黄启厅 等.多源多时相遥感影像相对辐射归一化方法研究.《地球信息科学》.2016,第18卷(第5期),第606-614页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106327452A (en) | 2017-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106327452B (en) | A kind of fragmentation remote sensing image synthetic method and device towards cloudy rain area | |
| Thonfeld et al. | Robust change vector analysis (RCVA) for multi-sensor very high resolution optical satellite data | |
| KR101728137B1 (en) | Method for land-cover item images classification by using satellite picture and GIS | |
| Liu et al. | Object-based shadow extraction and correction of high-resolution optical satellite images | |
| Sandmeier et al. | A physically-based model to correct atmospheric and illumination effects in optical satellite data of rugged terrain | |
| CN102982538B (en) | Nature color simulating method of resource satellite multi-spectral image | |
| JP3105840B2 (en) | Earth information supply system | |
| CN112733596A (en) | Forest resource change monitoring method based on medium and high spatial resolution remote sensing image fusion and application | |
| Szantoi et al. | Fast and robust topographic correction method for medium resolution satellite imagery using a stratified approach | |
| Sudharsan et al. | A survey on hyperspectral imaging for mineral exploration using machine learning algorithms | |
| CN110703244A (en) | Method and device for identifying urban water body based on remote sensing data | |
| Zhang et al. | Quantitative assessment of a haze suppression methodology for satellite imagery: Effect on land cover classification performance | |
| Zhou et al. | Restoration of information obscured by mountainous shadows through Landsat TM/ETM+ images without the use of DEM data: A new method | |
| CN102231190B (en) | Automatic extraction method for alluvial-proluvial fan information | |
| Schläpfer et al. | Correction of shadowing in imaging spectroscopy data by quantification of the proportion of diffuse illumination | |
| Vanzella et al. | Multicolor Observations of the Hubble Deep Field South | |
| CN113743303A (en) | Forestry carbon remittance method for forest remote sensing data | |
| CN109389063A (en) | Remote sensing image Strip noise removal method based on wave band correlation | |
| Danoedoro et al. | Integrating spectral indices and geostatistics based on Landsat-8 imagery for surface clay content mapping in Gunung Kidul area, Yogyakarta, Indonesia | |
| Nimbalkar et al. | Optimal band configuration for the roof surface characterization using hyperspectral and LiDAR imaging | |
| Harris et al. | A comparison of different remotely sensed data for classifying bedrock types in Canada’s Arctic: application of the robust classification method and random forests | |
| Ouerghemmi et al. | Urban vegetation mapping by airborne hyperspetral imagery; feasibility and limitations | |
| Sun et al. | A priori surface reflectance-based cloud shadow detection algorithm for Landsat 8 OLI | |
| Rani et al. | The effect of topographic correction on SPOT6 land cover classification in water catchment areas in Bandung Basin, Indonesia | |
| Kao et al. | Calibrated ratio approach for vegetation detection in shaded areas |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |