CN108154471A - A kind of method and apparatus for remote sensing image splicing - Google Patents
A kind of method and apparatus for remote sensing image splicing Download PDFInfo
- Publication number
- CN108154471A CN108154471A CN201711131587.XA CN201711131587A CN108154471A CN 108154471 A CN108154471 A CN 108154471A CN 201711131587 A CN201711131587 A CN 201711131587A CN 108154471 A CN108154471 A CN 108154471A
- Authority
- CN
- China
- Prior art keywords
- image
- effective
- primary election
- effective image
- target
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4038—Scaling the whole image or part thereof for image mosaicing, i.e. plane images composed of plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
Abstract
The object of the present invention is to provide a kind of method and apparatus for remote sensing image splicing.Effective image included in multiple remote sensing images that present invention extraction is obtained, according to multiple effective images, it determines preferably effective image corresponding with target splicing regions and generates and the corresponding initial target image of target splicing regions, then the preferably effective boundary line of the image in the initial target image is extracted, and root is iterated even color to preferably effective image and handles, to obtain the target image corresponding to the target splicing regions.Compared with prior art, the present invention realizes by extracting effective image and determining the modes such as preferably effective image and quickly and efficiently handles remote sensing image, improve the efficiency of remote sensing image splicing;Meanwhile by extracting the processing such as boundary line and the even color of iteration, make two images that the gray scale under different image-forming conditions has differences mutually coordinated consistent on hue and luminance, improve the quality of the even color of image joint.
Description
Technical field
The present invention relates to technical field of remote sensing image processing more particularly to a kind of technologies for remote sensing image splicing.
Background technology
Remote sensing image (Remote Sensing Image) refers to record the film or photograph of various atural object electromagnetic wave sizes,
Such as airphoto and satellite photograph.It is all sat in every remote sensing image comprising corresponding geographical coordinate according to the geography of remote sensing image
Mark can learn the overlapping region of two adjacent remote sensing images;In other words, each pixel in remote sensing image can obtain
Its unique geographical coordinate position.
Remote sensing image splicing (or remote sensing image is inlayed) is two or multiple phases that will be obtained under different image-forming conditions
The remote sensing image with lap is spliced between neighbour, to generate the technology of a high-resolution global image.Remote sensing shadow
The feature of picture mainly includes three kinds, i.e. spectral signature, space characteristics and textural characteristics, therefore, when carrying out remote sensing image splicing,
It needs to account in terms of three from this, so that spliced picture is realized on spectral signature, space characteristics and textural characteristics
It is mutually coordinated.
However, current remote sensing image splicing efficiency in splicing is low, splicing is difficult, particularly in remote sensing shadow
Tone, brightness are inconsistent in the mosaic process of picture so that obtained stitching image is second-rate.Therefore, how remote sensing is improved
The efficiency of image joint improves the quality of image joint so that two width that the gray scale under different image-forming conditions has differences or
Several remote sensing images are mutually coordinated consistent on hue and luminance, become those skilled in the art institute urgent problem to be solved.
Invention content
The object of the present invention is to provide a kind of method and apparatus for remote sensing image splicing.
According to one embodiment of present invention, a kind of method for remote sensing image splicing is provided, wherein, this method packet
Include following steps:
Obtain multiple remote sensing images;
Extract effective image that the remote sensing image is included;
According to multiple effective images, determine and target splicing regions corresponding preferred effectively image and generation and institute
State the corresponding initial target image of target splicing regions;
Extract the preferably effective boundary line of the image in the initial target image;
According to the boundary line, even color is iterated to preferably effective image and is handled, is spliced with obtaining the target
Target image corresponding to region.
Optionally, the step of extracting effective image that every remote sensing image is included includes:
The remote sensing image is progressively scanned, when scanning is at least one Pixel Information, is believed according to the pixel
The position of breath determines and corresponding at least one edge line of the remote sensing image;
According to the edge line, effective image that the remote sensing image is included is extracted.
Optionally it is determined that preferably effective image corresponding with target splicing regions and generation and the target splicing regions
The step of corresponding initial target image, includes:
According to multiple effective images, filter out with the effective image of the corresponding primary election of target splicing regions, and determine
Stitching position corresponding to each effective image of the primary election;
According to the stitching position corresponding to each effective image of primary election, the redundancy in the effective image of the primary election is removed
Image, to obtain preferably effective image;
The preferably effectively image is spliced, with generation and the corresponding initial target figure of the target splicing regions
Picture.
Optionally it is determined that preferably effective image corresponding with target splicing regions and generation and the target splicing regions
The step of corresponding initial target image, includes:
According to multiple effective images, filter out with the effective image of the corresponding primary election of target splicing regions, and determine
Stitching position corresponding to each effective image of the primary election;
The effective image of the primary election is spliced, with generation and the corresponding original object figure of the target splicing regions
Picture;
According to stitching position of each effective image of primary election in the original target image, removing the primary election has
The redundancy image in image is imitated, the initial target figure with the preferably effective image of acquisition and corresponding to the preferred effectively image
Picture.
Optionally, it screens the effective image of primary election and is wrapped the step of determining the stitching position corresponding to the primary election effective image
It includes:
According to effective image and the coordinate position relationship of target splicing regions, screened from image effective multiple described
Go out with the effective image of the corresponding primary election of target splicing regions, and determine each effective image of primary election corresponding to splice bits
It puts.
Optionally, the step of obtaining preferably effective image includes:
According to longitude station, Position Latitude and the intersection location corresponding to each effective image of primary election, determine each
Stitching position corresponding to a effective image of the primary election;
By calculating the distance relation of the effective image of the second primary election of one or more intersected with the effective image of the first primary election,
It determines and the corresponding optimal effective image of second primary election of the effective image of the first primary election;
Using the optimal effective image of second primary election as the pending effective image of the first primary election, cycle, which performs, determines institute
The step of stating optimal second primary election effective image to remove the redundancy image in the effective image of the primary election, obtains preferably effective
Image.
Optionally, the step of extracting boundary line of the preferably effective image in the initial target image includes:
Determine the preferably effective fit line of the image in the initial target image;
Pixel band corresponding to according to the pixel on the fit line and with the pixel, according to the pixel
With on the pixel band pixel apart from aberration, extract the preferably effective boundary line of image.
Optionally it is determined that the step of fit line, includes:
According to the boundary point of preferably effective image, determine the preferably effectively image in the initial target image
Initial fit line;
It, will be described first according to splicing direction of the preferably effective image when splicing in the initial target image
Beginning fit line is translated, to obtain the preferably effective fit line of the image in the initial target image.
Optionally, described the step of preferably effectively image is iterated even color processing, is included:
According to the boundary line, the subregion preferably effectively corresponding to image is determined;
According to the subregion corresponding to the preferably effectively image, even color processing is iterated to adjacent subregion successively, with
Obtain the target image corresponding to the target splicing regions.
Optionally, the step of being iterated even color processing to adjacent subregion successively includes:
According to the subregion corresponding to the preferably effectively image, at least one set of adjacent subregion is determined;
From the adjacent sectors, iteration chooses the even colour band with different length, and is iterated even color processing, to obtain
The target image corresponding to the target splicing regions is obtained, wherein, the length of the even colour band is directly proportional to gradual change parameter.
According to another embodiment of the invention, a kind of processing equipment for remote sensing image splicing is additionally provided, wherein,
The processing equipment includes:
First device, for obtaining multiple remote sensing images;
Second device, for extracting effective image that the remote sensing image is included;
3rd device, for according to multiple effective images, determining corresponding preferably effectively with target splicing regions
Image simultaneously generates and the corresponding initial target image of target splicing regions;
4th device, for extracting the preferably effective boundary line of the image in the initial target image;
5th device, for according to the boundary line, being iterated even color to preferably effective image and handling, to obtain
Target image corresponding to the target splicing regions.
Optionally, the second device is used for:
The remote sensing image is progressively scanned, when scanning is at least one Pixel Information, is believed according to the pixel
The position of breath determines and corresponding at least one edge line of the remote sensing image;
According to the edge line, effective image that the remote sensing image is included is extracted.
Optionally, the 3rd device includes:
Unit 31, for according to multiple effective images, filtering out to have with the corresponding primary election of target splicing regions
Image is imitated, and determines the stitching position corresponding to each effective image of primary election;
Unit three or two, for the stitching position according to corresponding to each effective image of primary election, removing the primary election has
The redundancy image in image is imitated, to obtain preferably effective image;
Unit three or three are opposite with the target splicing regions to generate for the preferably effectively image to be spliced
The initial target image answered.
Optionally, the 3rd device includes:
Unit 31, for according to multiple effective images, filtering out to have with the corresponding primary election of target splicing regions
Image is imitated, and determines the stitching position corresponding to each effective image of primary election;
Unit three or four are opposite with the target splicing regions to generate for the effective image of the primary election to be spliced
The original target image answered;
Unit three or five, for the stitching position according to each effective image of primary election in the original target image,
Remove the redundancy image in the effective image of the primary election, with obtain preferably effectively image and with it is described preferably effectively image institute it is right
The initial target image answered.
Optionally, Unit 31 is used for:
According to effective image and the coordinate position relationship of target splicing regions, screened from image effective multiple described
Go out with the effective image of the corresponding primary election of target splicing regions, and determine each effective image of primary election corresponding to splice bits
It puts.
Optionally, Unit three or two or Unit three or five are used for:
According to longitude station, Position Latitude and the intersection location corresponding to each effective image of primary election, determine each
Stitching position corresponding to a effective image of the primary election;
By calculating the distance relation of the effective image of the second primary election of one or more intersected with the effective image of the first primary election,
It determines and the corresponding optimal effective image of second primary election of the effective image of the first primary election;
Using the optimal effective image of second primary election as the pending effective image of the first primary election, cycle, which performs, determines institute
The step of stating optimal second primary election effective image to remove the redundancy image in the effective image of the primary election, obtains preferably effective
Image.
Optionally, the 4th device includes:
Unit 41, for determining the preferably effective fit line of the image in the initial target image;
Unit four or two, for the pixel band corresponding to according to the pixel on the fit line and with the pixel,
According on the pixel and the pixel band pixel apart from aberration, extract the preferably effective boundary line of image.
Optionally, Unit 41 is used for:
According to the boundary point of preferably effective image, determine the preferably effectively image in the initial target image
Initial fit line;
It, will be described first according to splicing direction of the preferably effective image when splicing in the initial target image
Beginning fit line is translated, to obtain the preferably effective fit line of the image in the initial target image.
Optionally, the 5th device includes:
May Day unit, for according to the boundary line, determining the subregion preferably effectively corresponding to image;
Unit five or two for the subregion according to corresponding to the preferably effectively image, successively change to adjacent subregion
For the processing of even color, to obtain the target image corresponding to the target splicing regions.
Optionally, Unit five or two is used for:
According to the subregion corresponding to the preferably effectively image, at least one set of adjacent subregion is determined;
From the adjacent sectors, iteration chooses the even colour band with different length, and is iterated even color processing, to obtain
The target image corresponding to the target splicing regions is obtained, wherein, the length of the even colour band is directly proportional to gradual change parameter.
Compared with prior art, effective image included in multiple remote sensing images that present invention extraction is obtained, according to
Multiple described effective images determine preferably effective image corresponding with target splicing regions and generate and the target splice region
Then the corresponding initial target image in domain extracts the preferably effective boundary line of the image in the initial target image,
And according to the boundary line, even color is iterated to preferably effective image and is handled, to obtain the target splicing regions institute
Corresponding target image.So as to which the present invention is realized fast by extracting effective image and determining the modes such as preferably effective image
Speed efficiently handles remote sensing image, improves the efficiency of remote sensing image splicing;And pass through and extract boundary line and the even color of iteration etc.
Processing makes two images that the gray scale under different image-forming conditions has differences mutually coordinated consistent on hue and luminance, improves
The quality of the even color of image joint.
Description of the drawings
By reading the detailed description made to non-limiting example made with reference to the following drawings, of the invention is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 shows a kind of processing equipment schematic diagram spliced for remote sensing image according to one aspect of the invention;
Fig. 2 shows a kind of method flow diagrams for remote sensing image splicing according to a further aspect of the present invention;
One kind that Fig. 3 shows according to a preferred embodiment of the present invention filters out corresponding just with target splicing regions
Select the method schematic diagram of effective image;
Fig. 4 (1) to Fig. 4 (4) shows the preferably effective image of a kind of acquisition according to a preferred embodiment of the present invention
Method schematic diagram;
Fig. 5 (1) to Fig. 5 (3) shows a kind of extraction preferably effective shadow according to a preferred embodiment of the present invention
As the method schematic diagram in the vertical edges boundary line in the initial target image;
Fig. 6 (1) to Fig. 6 (3) shows a kind of extraction preferably effective shadow according to a preferred embodiment of the present invention
As the method schematic diagram of the horizontal border line in the initial target image.
The same or similar reference numeral represents the same or similar component in attached drawing.
Specific embodiment
It should be mentioned that some exemplary embodiments are described as before exemplary embodiment is discussed in greater detail
The processing described as flow chart or method.Although operations are described as the processing of sequence by flow chart, therein to be permitted
Multioperation can be implemented concurrently, concomitantly or simultaneously.In addition, the sequence of operations can be rearranged.When it
The processing can be terminated when operation is completed, it is also possible to have the additional step being not included in attached drawing.The processing
It can correspond to method, function, regulation, subroutine, subprogram etc..
So-called within a context " processing equipment ", as " computer equipment ", also referred to as " computer " refer to pass through
Preset program or instruction are run to perform the intelligent electronic device of the predetermined process process such as numerical computations and/or logical calculated,
It can include processor and memory, the survival to be prestored in memory by processor execution instructs to perform predetermined process mistake
Journey or predetermined process process is performed by hardware such as ASIC, FPGA, DSP or is realized by said two devices combination.
The computer equipment includes user equipment and/or the network equipment.Wherein, the user equipment includes but not limited to
Computer, smart mobile phone, PDA etc.;The network equipment includes but not limited to single network server, multiple network servers composition
Server group or the cloud being made of a large amount of computers or network server based on cloud computing (Cloud Computing),
In, cloud computing is one kind of Distributed Calculation, a super virtual computer being made of the computer collection of a group loose couplings.
Wherein, the computer equipment can isolated operation realize the present invention, also can access network and by with other meters in network
The interactive operation of machine equipment is calculated to realize the present invention.Wherein, the network residing for the computer equipment includes but not limited to interconnect
Net, wide area network, Metropolitan Area Network (MAN), LAN, VPN network etc..
Those skilled in the art will be understood that under normal circumstances heretofore described " processing equipment " can only be net
Network equipment is performed corresponding operation by the network equipment;Under special circumstances or by user equipment and the network equipment
Or server is integrated to form, i.e., matches with the network equipment to perform corresponding operation by user equipment, for example, by user
Equipment sends to the network equipment and instructs, to indicate that the network equipment starts the corresponding operating of execution " remote sensing image splicing ".
It should be noted that the user equipment, the network equipment and network etc. are only for example, other are existing or from now on may be used
The computer equipment or network that can occur such as are applicable to the present invention, should also be included within the scope of the present invention, and to draw
It is incorporated herein with mode.
Specific structure and function details disclosed herein are only representative, and are for describing showing for the present invention
The purpose of example property embodiment.But the present invention can be implemented, and be not interpreted as by many alternative forms
It is limited only by the embodiments set forth herein.
Although it should be understood that may have been used term " first ", " second " etc. herein to describe each unit,
But these units should not be limited by these terms.The use of these items is only for by a unit and another unit
It distinguishes.For example, in the case of the range without departing substantially from exemplary embodiment, it is single that first unit can be referred to as second
Member, and similarly second unit can be referred to as first unit.Term "and/or" used herein above include one of them or
The arbitrary and all combination of more listed associated items.
Term used herein above is not intended to limit exemplary embodiment just for the sake of description specific embodiment.Unless
Context clearly refers else, otherwise singulative used herein above "one", " one " also attempt to include plural number.Should also
When understanding, term " comprising " and/or "comprising" used herein above provide stated feature, integer, step, operation,
The presence of unit and/or component, and do not preclude the presence or addition of other one or more features, integer, step, operation, unit,
Component and/or a combination thereof.
It should further be mentioned that in some replaces realization modes, the function/action being previously mentioned can be according to different from attached
The sequence indicated in figure occurs.For example, depending on involved function/action, the two width figures shown in succession actually may be used
Substantially simultaneously to perform or can perform in a reverse order sometimes.
The present invention is described in further detail below in conjunction with the accompanying drawings.
Fig. 1 shows a kind of processing equipment schematic diagram spliced for remote sensing image according to one aspect of the invention;Wherein,
The processing equipment includes first device 1, second device 2,3rd device 3, the 4th device 4, the 5th device 5.
Specifically, the first device 1 obtains multiple remote sensing images;The second device 2 extracts the remote sensing image institute
Comprising effective image;The 3rd device 3 determines corresponding excellent with target splicing regions according to multiple effective images
It selects effective image and generates and the corresponding initial target image of target splicing regions;4th device 4 extracts described
It is preferred that boundary line of effective image in the initial target image;5th device 5 is according to the boundary line, to described excellent
Effective image is selected to be iterated even color processing, to obtain the target image corresponding to the target splicing regions.
The first device 1 obtains multiple remote sensing images.
Specifically, the first device 1 by with one or more capture apparatus (such as satellite, Aerial photography equipment)
And/or other equipment for being capable of providing remote sensing image interact, to obtain multiple remote sensing images.
Those skilled in the art, which will be understood that, includes corresponding geographical coordinate in the remote sensing image, in remote sensing image
Each pixel can obtain its unique geographical coordinate position.Preferably, every remote sensing image includes the remote sensing image
Orbital number, the track as shoots the satellite of this remote sensing image or Aerial photography equipment is shooting this remote sensing image
When, the running track that is run.One satellite can be different to cover by running one or more different running track
Region, and the remote sensing image captured by every, may both correspond to the number of a track.
The second device 2 extracts effective image that the remote sensing image is included.
Specifically, wherein, effective image is extracted, i.e., is removed the inactive area in the remote sensing image, to obtain
The method of slice containing effective information in the remote sensing image.
The second device 2 extracts effective image that the remote sensing image is included by one or more of mode:
1. the remote sensing image is cut with fixed size or fixed shape, to determine its effective coverage:Such as interception
The image of the fixed pixel size in the remote sensing image centre position, using the image cut as effective image.
2. different capture apparatus (satellite of such as different model), different shooting areas according to corresponding to the remote sensing image
The acquisition parameters such as domain, different styles of shooting, carry out the remote sensing image cutting of different shapes and sizes, by what is cut
Image is as effective image:Cutting corresponding to different, the different shooting areas of cutting profile as corresponding to the satellite of different model
It cuts of different sizes etc..
3. according to the absence information of the target area to be inlayed, according to the locational requirement of vacancy, from the remote sensing image
The corresponding appropriate image of upper interception, using as effective image:If one 100 pixel X100 pixel of target area vacancy
Image, then from the remote sensing image of corresponding (can judging whether to correspond to according to coordinate information), 110 pixel X110 pixels of interception
Image-region, using as effective image.
4. by being progressively scanned to the remote sensing image, to determine the corresponding effective image of remote sensing image.
I.e., it is preferable that the second device 2 can progressively scan the remote sensing image, when scanning at least one
During a Pixel Information, according to the position of the Pixel Information, determine and corresponding at least one edge line of the remote sensing image;
According to the edge line, effective image that the remote sensing image is included is extracted.
Specifically, the second device 2 can progressively scan the remote sensing image, such as from image periphery line by line to
Picture centre region scans;When the Pixel Information of scanning at least one non-blank-white, then show to have existed significant figure herein
According to rather than white space, therefore, according to the position of the Pixel Information, determine and the remote sensing image corresponding at least one
Edge line.
If here, those skilled in the art will be understood that some region of a remote sensing image is inactive area, it is somebody's turn to do
Without effective pixel in region, not comprising any initial data.What this parts of images was visually presented is black.And if
The effective pixel points comprising initial data are scanned, then illustrate to have existed valid data herein, then can stop scanning.
Here, if scanning is closed to a Pixel Information according to the Pixel Information and the shape of effective image to be obtained
System, setting-out (edge of the position and effective image to be intercepted such as is come drawing parallel lines) is to obtain edge line;And then
According to a plurality of edge line positioned at different directions, the image in edge line is extracted as effective image.If multiple pixels are arrived in scanning
Information, then by multiple Pixel Information lines, to obtain corresponding one or more edge line, using the image in edge line as having
Imitate image.
For example, pixel spot scan is carried out respectively in the upper left of the remote sensing image, lower-left, upper right, bottom right four direction, with
It determines edge line, and then determines effective image.
The 3rd device 3 determines corresponding preferably effectively with target splicing regions according to multiple effective images
Image simultaneously generates and the corresponding initial target image of target splicing regions.
Specifically, the 3rd device 3 can be first according to multiple effective images, and screening determines and target splicing regions
Corresponding preferably effective image then according to preferably effective image, generates corresponding with the target splicing regions
Initial target image;Alternatively, the 3rd device 3 can be first according to multiple effective images, generation is spliced with the target
After the corresponding initial target image in region, then further screened, it is corresponding preferably with target splicing regions to determine
Effective image.
Concrete analysis is as follows:
The 3rd device 3 can include 31 unit (not shown), three or two unit (not shown) and Unit three or three (not
It shows).Wherein, Unit 31 filters out and the corresponding primary election of target splicing regions according to multiple effective images
Effective image, and determine the stitching position corresponding to each effective image of primary election;Unit three or two is according to each described
Stitching position corresponding to the effective image of primary election removes the redundancy image in the effective image of the primary election, preferably effective to obtain
Image;Unit three or three splices the preferably effectively image, corresponding with the target splicing regions to generate
Initial target image.
Specifically, Unit 31 can be by one or more of mode, to screen the effective image of primary election:
1. according to effective image and the position relationship of the target splicing regions:Effective image as described in by judgement
Geographical coordinate whether be located in the coordinate range that is covered of the target splicing regions or judge effective image it is right
Whether the number answered is corresponding with the target splicing regions.It is covered if effective image is located at the target splicing regions
Coordinate range in or effective image corresponding to number it is corresponding with the target splicing regions, then have described
It is the effective image of the primary election to imitate image screening;Otherwise, it is screened out.
Wherein, the target splicing regions are the range for the remote sensing image for it is expected generation, such as global remote sensing image, Asia
Regional remote sensing image, regional remote sensing images or some longitude range and some latitude scope area defined remote sensing
Image etc..
2. it is screened according to the clarity of clear effective image:Using clarity be more than a certain threshold value as described in
The effective image of primary election, otherwise, is screened out.
3. other scheduled screening modes etc..
After Unit 31 filters out image effective with the corresponding primary election of target splicing regions, according to the primary election
Geographical coordinate corresponding to effective image according to the position relationship corresponding to the effective image of each primary election, determines each described first
Select the stitching position corresponding to effective image;It if, can direct basis alternatively, comprising name placement in the effective image of the primary election
Name placement corresponding to the effective image of primary election determines the stitching position corresponding to each effective image of the primary election.
Stitching position of the Unit three or two according to corresponding to each effective image of primary election, it is effective to remove the primary election
Redundancy image in image, to obtain preferably effective image.
Specifically, the effective image of each primary election is put into unified coordinate system by Unit three or two, according to each described
Stitching position corresponding to the effective image of primary election judges the coverage area corresponding to the effective image of each primary election, by described
Coverage area is detected, if the range that the effective images of certain Zhang Chuxuan are covered can completely be covered by the effective image of other primary election
Lid, then illustrate that the region corresponding to the effective images of the Zhang Chuxuan can be obtained according to the effective image of other primary election, therefore
The effective image of the primary election is redundancy image, and the effective image of the primary election is screened out, performs aforesaid operations repeatedly, will be owned
Redundancy image removes, final to obtain preferably effective image;Alternatively, by by the effective image of the primary election according to certain splicing side
To (such as from top to bottom, from left to right etc.) or splicing sequence, target splicing regions are spliced to, it is effective often to splice the primary election
During image, then one-time detection is performed:If certain effective image of primary election newly spliced is fully contemplated to the image range covered
It is interior, then the effective image of the primary election is screened out, if it has at least part pixel that can cover in new image range,
The effective image of the primary election is retained, so as to obtain preferably effective image.
Unit three or three splices the preferably effectively image, corresponding with the target splicing regions to generate
Initial target image.
Unit three or three is by acquired preferred effective image, according to the position relationship corresponding to it, according to certain
Splicing sequence, such as from top to bottom from left to right, spliced, with generation and the corresponding initial mesh of the target splicing regions
Logo image.
Alternatively, the 3rd device 3 can include 31 unit (not shown), three or four unit (not shown) and three or five lists
First (not shown).Wherein, Unit 31 filters out corresponding with target splicing regions according to multiple effective images
The effective image of primary election, and determine the stitching position corresponding to each effective image of primary election;Unit three or four will be described first
Effective image is selected to be spliced, with generation and the corresponding original target image of target splicing regions;Unit three or five
According to stitching position of each effective image of primary election in the original target image, remove in the effective image of the primary election
Redundancy image, with obtain preferably effectively image and with the initial target image preferably effectively corresponding to image.
Here, the operation and aforementioned Unit 31 performed by Unit 31 are same or similar, therefore details are not described herein.
The effective image of each primary election is put into unified coordinate system by Unit three or four, and the effective image of the primary election is pressed
According to the position relationship corresponding to it, according to certain splicing sequence, such as from top to bottom from left to right, spliced, with generation
With the corresponding original target image of target splicing regions.Here, each primary election in the original target image is effective
Image can be identified and perform the relevant operations such as removal.
For example, input parameter can be passed through:UTM corresponding to the serial number list of the effective image of primary election, target splicing regions
Area information, then output parameter be:Original target image.
Here, performed operation is:The regional number corresponding to tif files is retrieved, is then arranged according to the regional number
Sequence, the orderly list of file names of generation;It is ranked up and splices according to UTM area informations.
Wherein, UTM is:UNIVERSAL TRANSVERSE MERCARTOR GRID SYSTEM, Universal Trans Meridian lattice
Net system.
Stitching position of the Unit three or five according to each effective image of primary election in the original target image, goes
Except the redundancy image in the effective image of the primary election, to obtain preferably effective image and with described preferably effectively corresponding to image
Initial target image.
Specifically, Unit three or five is according to corresponding to each effective image of primary election in the original target image
Stitching position, judge the coverage area corresponding to the effective image of each primary election, by being detected to the coverage area, if
The range that the effective images of certain Zhang Chuxuan are covered can be completely covered by the effective image of other primary election, then illustrates that the Zhang Chuxuan has
Region corresponding to effect image can be obtained, therefore the effective image of the primary election is as superfluous according to the effective image of other primary election
The effective image of the primary election is screened out, performs aforesaid operations repeatedly by remaining image, and all redundancy images are removed, are finally obtained
Obtain preferably effective image.
Then, the original target image of the redundancy image is eliminated, as with described preferably effectively corresponding to image
Initial target image.
Preferably, Unit 31 can according to the coordinate position relationship of effective image and target splicing regions,
Filtered out from image effective multiple described with the effective image of the corresponding primary election of target splicing regions, and determine it is each it is described just
Select the stitching position corresponding to effective image.
Specifically, Fig. 3 shows that one kind according to a preferred embodiment of the present invention filters out and target splicing regions
The method schematic diagram of the corresponding effective image of primary election.
Unit 31 can determine the regional extent of the target splicing regions, then, for each effective image,
Obtain the central point latitude and longitude information corresponding to effective image.Longitude (longitude as shown in Figure 3 corresponding to effective image
X) there are two corresponding latitude point, Y1 points as shown in Figure 3 and Y2 points with the target splicing regions.If effective image
Latitude between Y1 and Y2, then illustrate effective image in target splicing regions, effective image for primary election it is effective
Image.
According to the position relationship corresponding to the effective image of the primary election, the splicing corresponding to the effective image of the primary election is determined
Position.
Preferably, Unit three or two or Unit three or five can be according to corresponding to each effective images of primary election
Longitude station, Position Latitude and intersection location determine the stitching position corresponding to each effective image of the primary election;Pass through meter
The distance relation of the effective image of the second primary election of one or more intersected with the effective image of the first primary election is calculated, is determined and described first
The corresponding optimal effective image of second primary election of the effective image of primary election;Using the optimal effective image of second primary election as pending
The effective image of the first primary election, cycle performs the step of determining optimal second primary election effective image, to remove the primary election
Redundancy image in effective image obtains preferably effective image.
Specifically, Unit three or two or Unit three or five when obtaining preferably effective image, may be used with lower section
Method so that irredundant minimum image covers the target splicing regions, reduces calculation amount, improves image processing efficiency:
First, Unit three or two or Unit three or five determine the longitude corresponding to each effective image of the primary election
Position, Position Latitude and intersection location, then, based on above-mentioned position relationship, to determine each effective image institute of primary election
Corresponding stitching position.Here, using longitude station, Position Latitude and intersection location as three constraintss, to splicing
Method for determining position illustrates.
Constraints 1 (longitude):The effective image of primary election of the condition of satisfaction is filtered out in longitude range.Firstly, for original
The effective image of each primary election in target image has choosing a fritter from certain side (such as certain row leftmost side) as initial primary election
Image is imitated, finds out and meets each primary election effective image of the midpoint longitude more than the midpoint longitude of the initial effective image of primary election, to protect
It is (effective including overlapping and nonoverlapping primary election on the right of the effective image of initial primary election to demonstrate,prove the effective image of each primary election filtered out
Image);Then the center position coordinates information of all effective images of primary election in the right is found out from tif files.
Constraints 2 (latitude):Fig. 4 (1) shows a kind of undesirable image joint combination.In order to avoid the feelings of Fig. 4 (1)
Condition, in order to prevent the effective image of each primary election that search obtains larger in vertical direction deviation, so added in latitude scope again
Constraints, the constraints are the ranges that the effective image of initial primary election (influence i.e. shown in dotted line) is made to be searched in vertical direction
It is from the distance of M, N point on midpoint P to up-and-down boundary line, i.e. the y direction in latitude and longitude coordinates system selects affiliated latitude
The slice of range, so that shown in final search splicing effect such as Fig. 4 (2).
In Fig. 4 (2), A is the effective image of initial primary election, and B is the effective image of primary election of redundancy, and C searches for obtain
It is preferred that the effective image of primary election.
Those skilled in the art will be understood that next can using image C as initial image, then in the manner described above after
Continue to the right by described, it is known that search the effective image of primary election of the rightmost in the original target image.
Constraints 3 (intersecting):Fig. 4 (3) shows another undesirable image joint combination.More than pair warp and weft degree adds
After constraints, in order to avoid there is the disjoint situation of each corresponding effective image of primary election in Fig. 4 (3), should also it add intersecting
Condition.
As shown in Fig. 4 (4), if the coordinate vector on the effective image AB directions boundary of original primary election be a, BD directions boundary to
Measure as b, the vector on DC directions boundary is c, and the vector on CA directions boundary is d, then in order to meet intersecting principle, calculate with it is initial just
Select midpoint P on the boundary line of the effective image of the primary election that effective image intersects, border vertices F, Q respectively with the effective shadow of original primary election
As the Outer Product of Vectors on each boundary.
Wherein, the Outer Product of Vectors on midpoint P and each boundary of the effective image of original primary election (AB, BD, DC, CA) calculates as follows respectively
It is shown:
PA×a
PB×b
PC×d
PD×c
Wherein, PA refers to P points to the vector of A points, and a is A points to the vector of B points, and PA × a is vector of the P points to A points
With the apposition of the vector of A points to B points.The computational methods of the apposition are:| PA × a |=| PA | | a | sin<PA,a>, also
It is to say when the Outer Product of Vectors is negative, illustrates that two vectorial angles are negative, it is meant that two polygons intersect.
Similarly, the apposition between PB × b, as vector PB and b, the apposition between PC × d, as vector PC and d, PD
Apposition between × c, as vector PD and d.
When the result of aforementioned four apposition is less than zero, illustrate that P points are fallen in the initial effective image ABCD of primary election,
Then the effective image of two primary election intersects.
Wherein, the Outer Product of Vectors on border vertices F and each boundary of the effective image of original primary election (AB, BD, DC, CA), which calculates, distinguishes
It is as follows:
FA×a
FB×b
FC×d
FD×c
It is similar with the calculating of the Outer Product of Vectors on each boundary of the effective image of original primary election (AB, BD, DC, CA) with midpoint P, work as FA
When this four appositions of × a, FB × b, FC × d, FD × c are respectively less than zero, illustrate that F points are fallen in the initial effective image ABCD of primary election
In, then the effective image of two primary election intersects.
Wherein, the Outer Product of Vectors on border vertices Q and each boundary of the effective image of original primary election (AB, BD, DC, CA), which calculates, distinguishes
It is as follows:
QA×a
QB×b
QC×d
QD×c
It is similar with the calculating of the Outer Product of Vectors on each boundary of the effective image of original primary election (AB, BD, DC, CA) with midpoint P, work as QA
When this four appositions of × a, QB × b, QC × d, QD × c are respectively less than zero, illustrate that Q points are fallen in the initial effective image ABCD of primary election
In, then the effective image of two primary election intersects.
In other words, four appositions that above-mentioned every group of apposition (corresponding respectively to P points, F points and Q points) is calculated (represent
The apposition of the point and vector a, b, c, d) when being respectively less than zero, then prove that any point in 3 points of point P, F, Q is fallen in initial primary election
In effective image, two effective images of primary election intersect;And any one apposition in arbitrary a line is more than zero, then proves at the beginning of two
Select effective image non-intersect.
After the primary election effective image for meeting above three constraints is obtained, then by the effective image of one of primary election
As the effective image of the first primary election, then, each effective image of second primary election intersected with the effective image of the first primary election is calculated
Distance relation, such as pass through the distance relation of the midpoint and the effective image of the first primary election that calculate the effective image of the second primary election.It will tool
Have the effective image of the second primary election of maximum distance relationship as with the effective image of the first primary election corresponding optimal second at the beginning of
Select effective image.
And so on, using the optimal effective image of second primary election as the pending effective image of the first primary election, cycle
The step of determining optimal second primary election effective image is performed, to remove the redundancy image in the effective image of the primary election, is obtained
Obtain preferably effective image.
Those skilled in the art will be understood that is performed both by aforesaid operations for the effective image of the primary election of every row or each column, with
It obtains and all preferred effective image corresponding to the target splicing regions.Those skilled in the art will be understood that open from row
Begin, from left to right performs detection or since row, perform detection from top to bottom and can complete obtaining for preferred effective image
It takes, i.e., detection direction is not fixed.As long as ensure that circulate operation covers entire target area.
4th device 4 extracts the preferably effective boundary line of the image in the initial target image.
Specifically, the 4th device 4 can determine and extract described according to the stitching position of preferably effective image
It is preferred that boundary line of effective image in the initial target image;Alternatively, the 4th device 4 can determine described preferably have
Imitate fit line of the image in the initial target image, then, according to the pixel on the fit line and with the picture
Pixel band corresponding to vegetarian refreshments, according on the pixel and the pixel band pixel apart from aberration, extract it is described preferably
The boundary line of effective image.
Preferably, the 4th device includes 41 unit (not shown) and four or two unit (not shown);Wherein, described four
Unit one determines the preferably effective fit line of the image in the initial target image;Unit four or two is according to the plan
Pixel in zygonema and with the pixel band corresponding to the pixel, according to pixel on the pixel and the pixel band
Point apart from aberration, extract the preferably effective boundary line of image.
Unit 41 determines according to the stitching position of preferably effective image and extracts preferably effective image
Fit line in the initial target image, wherein, the fit line, that is, approximate bounds line.
Alternatively, it is highly preferred that Unit 41 can determine described excellent according to the boundary point of preferably effective image
Select initial fit line of effective image in the initial target image;According to the preferably effectively image in the initial target
Splicing direction when splicing on image, the initial fit line is translated, and is existed with obtaining the preferably effectively image
Fit line in the initial target image.
Specifically, Unit 41 detects attached in the stitching position according to the stitching position of preferably effective image
The value of chromatism of near each pixel extracts the point of value of chromatism maximum;Vector on boundary line per adjacent point-to-point transmission has big
Small and directional similarity gets rid of the point of size or direction dissmilarity by threshold value;By pixel as one or more,
It can determine the preferably effective initial fit line of the image in the initial target image;If Fig. 5 (2) is with regard in Fig. 6 (2)
AB, CD two lines, as described initial fit line (or line of prediction).
Then, it is taken from the initial fit line a little, takes several pixel bands from the location of pixels where the point, find out picture
The RGB of each pixel of the element with range and the difference of previous pixel RGB color (calculate two pixels in rgb space apart from color
Difference), i.e., several adjacent pixels are made with RGB is poor two-by-two, selects the point of difference maximum as boundary point.In order to make boundary point institute structure
Into boundary line preferably close to fit line, look for take boundary point when, the pixel band of selection cannot take too long (such as 20 pixels), because
It is more than pixel band intermediate region RGB with the difference of adjacent pixel RGB color that boundary position occurs for the pixel in order to avoid selection
The difference of color.Then the splicing direction according to preferably effective image when splicing in the initial target image, often
Every several step-lengths, every row is continued search in the above manner.Finally the coordinate where the point on the boundary line found out is carried out
Duplicate removal according to the identification point without repetition, therefore obtains unique borderline identification point, so that it is determined that going out fit line.
It is highly preferred that Unit 41 can also be corrected operation to the fitting boundary line extracted, such as vertical
On direction translating step and/or step-length is translated up in the horizontal direction to the right.
Specifically, 4 points of each boundary preferably effectively where image correspond to a range substantially in big figure.
During being combined into big figure in small figure, resolution ratio can continuously decrease, and according to the preferably effectively image described initial
The deviation of splicing direction when splicing on target image, initial fit line and fit line can be increasing.If for example, arranging
Direction be from right to left, the direction of row from the top down, fit line is initial to hang down during vertical direction generates from right to left
The deviation of straight fit line and vertical fit line is increasing;Similarly, in the horizontal direction from the top down during, initial level intend
Zygonema and the deviation of horizontal fit line are increasing.In order to reduce the error between initial fit line and fit line, such as Fig. 5 (3)
And as shown in Fig. 6 (3), in the vertical direction, several step-lengths that each initial fit line should be stepped up respectively to right translation;
Similarly, in the horizontal direction, each initial fit line should be translated up to several step-lengths gradually reduced, so that it is determined that fitting respectively
Line.
5th device 5 is iterated even color to preferably effective image and handles, to obtain according to the boundary line
Target image corresponding to the target splicing regions.
Specifically, the 5th device 5 such as utilizes the mode of Histogram Matching, to described according to the boundary line
The adjacent region in boundary line is iterated even color processing, to obtain the target image corresponding to the target splicing regions.It is described
The even color of iteration, i.e., even color selects certain even colour band length, and gradually adjust even colour band length every time.
Preferably, the 5th device includes May Day unit (not shown) and five or two unit (not shown), wherein, it is described
May Day unit determines the subregion preferably effectively corresponding to image according to the boundary line;Unit five or two is according to
It is preferred that the subregion corresponding to effective image, is successively iterated adjacent subregion even color processing, is spliced with obtaining the target
Target image corresponding to region.
Specifically, May Day unit chooses the preferred effective image of the boundary line both sides according to the boundary line
Subregion to be treated;Then, Unit five or two is according to two adjacent subregions, the mode as utilized Histogram Matching, according to
It is secondary to be iterated even color processing, to obtain the target image corresponding to the target splicing regions.
It is highly preferred that Unit five or two can determine at least one according to the subregion corresponding to the preferably effectively image
The adjacent subregion of group;From the adjacent sectors, iteration chooses the even colour band with different length, and is iterated at even color
Reason, to obtain the target image corresponding to the target splicing regions, wherein, the length and gradual change parameter of the even colour band are into just
Than.
Specifically, Unit five or two chooses the preferred effective image of the boundary line both sides according to the boundary line
Subregion to be treated;If the area code template belonging to different subregions should be used to carry out even comprising multiple subregions in the range of some
Color.
In order to make the comparison of the image transition where every piece of splicing part preferably effectively where image naturally, according to
The subregion that effectively image is included each preferably is simultaneously iterated even color, but to ensure the even color chosen successively every certain step-length
The length Length of band and the relative smooth of the parameter beta variation of gradual change, not there is mutation.
Then, Unit five or two determines that at least one set of adjacent subregion is handled.Here, the side with Histogram Matching
Formula carries out the explanation of even color mode:
In order to reduce the splicing aberration of image, luminance difference, using Histogram Matching principle, carry out between each adjacent sectors
Even color.So-called Histogram Matching, i.e., using the histogram of standard video as standard, by the histogram of another width image to standard Nogata
Figure is converted, and makes the histograms of two images close to similar.The tone and contrast of two images can be adjusted by Histogram Matching,
So as to which the Luminance Distribution for making two images approximately reaches unanimity, in order to keep the population equilibrium of picture tone, completely one is generated
The visual effect of body.Using subregion T as template, subregion B is obtained after carrying out Histogram Matching to its adjacent subregion A, then use with
Lower weighting parameters mode carries out gradual change to A, so as to the A' after being changed.
A'=γ A+ (1- γ) B
Wherein, γ is fade factor weight (i.e. gradual change parameter), and w ∈ (0,1).From left image to the weight of right image transition
In folded region process, weight w fades to 0 from 1, therefore just forms the smooth registration of two images.In addition, in order to preferably eliminate plan
Zygonema is iterated in the above manner, the scheme gradually reduced to the intersecting area of each contiguous slices into row pixel value step-length
Even color is carried out, so as to fulfill the consistency of result in tone and geometrically is inlayed.
Fig. 2 shows a kind of method flow diagrams for remote sensing image splicing according to a further aspect of the present invention.
Specifically, in step sl, the processing equipment obtains multiple remote sensing images;In step s 2, the processing is set
The standby effective image for extracting the remote sensing image and being included;In step s3, the processing equipment is according to multiple effective shadows
Picture determines preferably effective image corresponding with target splicing regions and generates corresponding initial with the target splicing regions
Target image;In step s 4, the processing equipment extraction preferably effective side of the image in the initial target image
Boundary line;In step s 5, the processing equipment is iterated at even color the preferably effectively image according to the boundary line
Reason, to obtain the target image corresponding to the target splicing regions.
In step sl, the processing equipment obtains multiple remote sensing images.
Specifically, in step sl, the processing equipment is by the way that (such as satellite, aviation are clapped with one or more capture apparatus
Take the photograph equipment etc.) and/or other equipment for being capable of providing remote sensing image interact, to obtain multiple remote sensing images.
Those skilled in the art, which will be understood that, includes corresponding geographical coordinate in the remote sensing image, in remote sensing image
Each pixel can obtain its unique geographical coordinate position.Preferably, every remote sensing image includes the remote sensing image
Orbital number, the track as shoots the satellite of this remote sensing image or Aerial photography equipment is shooting this remote sensing image
When, the running track that is run.One satellite can be different to cover by running one or more different running track
Region, and the remote sensing image captured by every, may both correspond to the number of a track.
In step s 2, the processing equipment extracts effective image that the remote sensing image is included.
Specifically, wherein, effective image is extracted, i.e., is removed the inactive area in the remote sensing image, to obtain
The method of slice containing effective information in the remote sensing image.
In step s 2, the processing equipment is extracted the remote sensing image and is included by one or more of mode
Effective image:
1. the remote sensing image is cut with fixed size or fixed shape, to determine its effective coverage:Such as interception
The image of the fixed pixel size in the remote sensing image centre position, using the image cut as effective image.
2. different capture apparatus (satellite of such as different model), different shooting areas according to corresponding to the remote sensing image
The acquisition parameters such as domain, different styles of shooting, carry out the remote sensing image cutting of different shapes and sizes, by what is cut
Image is as effective image:Cutting corresponding to different, the different shooting areas of cutting profile as corresponding to the satellite of different model
It cuts of different sizes etc..
3. according to the absence information of the target area to be inlayed, according to the locational requirement of vacancy, from the remote sensing image
The corresponding appropriate image of upper interception, using as effective image:If one 100 pixel X100 pixel of target area vacancy
Image, then from the remote sensing image of corresponding (can judging whether to correspond to according to coordinate information), 110 pixel X110 pixels of interception
Image-region, using as effective image.
4. by being progressively scanned to the remote sensing image, to determine the corresponding effective image of remote sensing image.
I.e., it is preferable that in step s 2, the processing equipment can progressively scan the remote sensing image, when sweeping
When retouching at least one Pixel Information, according to the position of the Pixel Information, determine corresponding at least with the remote sensing image
One edge line;According to the edge line, effective image that the remote sensing image is included is extracted.
Specifically, in step s 2, the processing equipment can progressively scan the remote sensing image, such as from image
It is scanned line by line to picture centre region periphery;When the Pixel Information of scanning at least one non-blank-white, then show herein
There are valid data rather than white spaces, therefore, according to the position of the Pixel Information, determine opposite with the remote sensing image
At least one edge line answered.
If here, those skilled in the art will be understood that some region of a remote sensing image is inactive area, it is somebody's turn to do
Without effective pixel in region, not comprising any initial data.What this parts of images was visually presented is black.And if
The effective pixel points comprising initial data are scanned, then illustrate to have existed valid data herein, then can stop scanning.
Here, if scanning is closed to a Pixel Information according to the Pixel Information and the shape of effective image to be obtained
System, setting-out (edge of the position and effective image to be intercepted such as is come drawing parallel lines) is to obtain edge line;And then
According to a plurality of edge line positioned at different directions, the image in edge line is extracted as effective image.If multiple pixels are arrived in scanning
Information, then by multiple Pixel Information lines, to obtain corresponding one or more edge line, using the image in edge line as having
Imitate image.
For example, pixel spot scan is carried out respectively in the upper left of the remote sensing image, lower-left, upper right, bottom right four direction, with
It determines edge line, and then determines effective image.
In step s3, the processing equipment determines corresponding with target splicing regions according to multiple effective images
Preferred effective image and generate with the corresponding initial target image of target splicing regions.
Specifically, in step s3, the processing equipment can be first according to multiple effective images, and screening determines and mesh
The corresponding preferably effective image of splicing regions is marked, then according to preferably effective image, generation and the target splice region
The corresponding initial target image in domain;Alternatively, in step s3, the processing equipment can be first according to multiple effective shadows
Picture after generation and the corresponding initial target image of target splicing regions, then is further screened, with determining and target
The corresponding preferably effective image of splicing regions.
Concrete analysis is as follows:
The step S3 can include step S31 (not shown), step S32 (not shown) and step S33 (not shown).
Wherein, in step S31, the processing equipment filters out corresponding with target splicing regions according to multiple effective images
The effective image of primary election, and determine each effective image of primary election corresponding to stitching position;In step s 32, the place
Stitching position of the equipment according to corresponding to each effective image of primary election is managed, removes the redundancy shadow in the effective image of the primary election
Picture, to obtain preferably effective image;In step S33, the processing equipment splices the preferably effectively image, with life
Into with the corresponding initial target image of target splicing regions.
Specifically, in step S31, the processing equipment can be by one or more of mode, to screen primary election
Effective image:
1. according to effective image and the position relationship of the target splicing regions:Effective image as described in by judgement
Geographical coordinate whether be located in the coordinate range that is covered of the target splicing regions or judge effective image it is right
Whether the number answered is corresponding with the target splicing regions.It is covered if effective image is located at the target splicing regions
Coordinate range in or effective image corresponding to number it is corresponding with the target splicing regions, then have described
It is the effective image of the primary election to imitate image screening;Otherwise, it is screened out.
Wherein, the target splicing regions are the range for the remote sensing image for it is expected generation, such as global remote sensing image, Asia
Regional remote sensing image, regional remote sensing images or some longitude range and some latitude scope area defined remote sensing
Image etc..
2. it is screened according to the clarity of clear effective image:Using clarity be more than a certain threshold value as described in
The effective image of primary election, otherwise, is screened out.
3. other scheduled screening modes etc..
After the processing equipment filters out image effective with the corresponding primary election of target splicing regions, according to the primary election
Geographical coordinate corresponding to effective image according to the position relationship corresponding to the effective image of each primary election, determines each described first
Select the stitching position corresponding to effective image;It if, can direct basis alternatively, comprising name placement in the effective image of the primary election
Name placement corresponding to the effective image of primary election determines the stitching position corresponding to each effective image of the primary election.
In step s 32, stitching position of the processing equipment according to corresponding to each effective image of primary election, removal
Redundancy image in the effective image of primary election, to obtain preferably effective image.
Specifically, in step s 32, the effective image of each primary election is put into unified coordinate system by the processing equipment,
According to the stitching position corresponding to each effective image of primary election, the covering model corresponding to the effective image of each primary election is judged
It encloses, by being detected to the coverage area, if the range that the effective images of certain Zhang Chuxuan are covered can be had by other primary election
It is completely covered to imitate image, then illustrates that the region corresponding to the effective images of the Zhang Chuxuan can be according to the effective shadow of other primary election
As being obtained, therefore the effective image of the primary election is redundancy image, the effective image of the primary election is screened out, is performed repeatedly above-mentioned
Operation removes all redundancy images, final to obtain preferably effective image;Alternatively, by by the effective image of the primary election according to
According to certain stitching direction (such as from top to bottom, from left to right etc.) or splicing sequence, target splicing regions are spliced to, often splice one
During the effective image of Zhang Suoshu primary election, then one-time detection is performed:If certain effective image of primary election newly spliced is fully contemplated to
In the image range of covering, then the effective image of the primary election is screened out, if it has at least part pixel that can cover newly
Image range in, then the effective image of the primary election is retained, so as to obtain preferably effective image.
In step S33, the processing equipment splices the preferably effectively image, is spelled with generation and the target
Connect the corresponding initial target image in region.
In step S33, the processing equipment closes acquired preferred effective image according to the position corresponding to it
System according to certain splicing sequence, such as from top to bottom from left to right, is spliced, with generation and the target splicing regions
Corresponding initial target image.
Alternatively, the step S3 can include step S31 (not shown), step S34 (not shown) and step S35 (not
It shows).Wherein, in step S31, the processing equipment filters out and target splicing regions according to multiple effective images
The corresponding effective image of primary election, and determine the stitching position corresponding to each effective image of primary election;In step S34,
The processing equipment splices the effective image of the primary election, with generation and the corresponding original mesh of the target splicing regions
Logo image;In step s 35, the processing equipment according to each effective image of primary election in the original target image
Stitching position removes the redundancy image in the effective image of the primary election, to obtain preferably effective image and preferably have with described
Imitate the initial target image corresponding to image.
Here, in step S31, the operation and aforementioned Unit 31 performed by the processing equipment are same or similar, therefore
Details are not described herein.
In step S34, the effective image of each primary election is put into unified coordinate system by the processing equipment, will be described first
Effective image is selected according to certain splicing sequence, such as from top to bottom from left to right, to be carried out according to the position relationship corresponding to it
Splicing, with generation and the corresponding original target image of target splicing regions.It is here, every in the original target image
A effective image of primary election can be identified and perform the relevant operations such as removal.
For example, input parameter can be passed through:UTM corresponding to the serial number list of the effective image of primary election, target splicing regions
Area information, then output parameter be:Original target image.
Here, performed operation is:The regional number corresponding to tif files is retrieved, is then arranged according to the regional number
Sequence, the orderly list of file names of generation;It is ranked up and splices according to UTM area informations.
Wherein, UTM is:UNIVERSAL TRANSVERSE MERCARTOR GRID SYSTEM, Universal Trans Meridian lattice
Net system.
In step s 35, the processing equipment according to each effective image of primary election in the original target image
Stitching position, remove the redundancy image in the effective image of the primary election, with obtain preferably effectively image and with it is described preferably
Initial target image corresponding to effective image.
Specifically, in step s 35, the processing equipment according to each effective image of primary election in the original object
Corresponding stitching position, judges the coverage area corresponding to the effective image of each primary election, by the covering model in image
It encloses and is detected, if the range that the effective images of certain Zhang Chuxuan are covered can be completely covered by the effective image of other primary election,
Illustrate that the region corresponding to the effective images of the Zhang Chuxuan can be obtained, therefore the primary election according to the effective image of other primary election
Effective image is redundancy image, and the effective image of the primary election is screened out, performs aforesaid operations repeatedly, by all redundancy shadows
It is final to obtain preferably effective image as removal.
Then, the original target image of the redundancy image is eliminated, as with described preferably effectively corresponding to image
Initial target image.
Preferably, in step S31, the processing equipment can be according to effective image and the seat of target splicing regions
Cursor position relationship, filtered out from image effective multiple described with the effective image of the corresponding primary election of target splicing regions, and really
Stitching position corresponding to fixed each effective image of primary election.
Specifically, Fig. 3 shows that one kind according to a preferred embodiment of the present invention filters out and target splicing regions
The method schematic diagram of the corresponding effective image of primary election.
In step S31, the processing equipment can determine the regional extent of the target splicing regions, then, for
Each effective image, obtains the central point latitude and longitude information corresponding to effective image.Longitude corresponding to effective image is (such as
Longitude X shown in Fig. 3) there are two corresponding latitude point, Y1 points as shown in Figure 3 and Y2 points with the target splicing regions.
It is described effective if the latitude of effective image between Y1 and Y2, illustrates effective image in target splicing regions
Image is the effective image of primary election.
According to the position relationship corresponding to the effective image of the primary election, the splicing corresponding to the effective image of the primary election is determined
Position.
Preferably, in the step S32 or described steps S35, the processing equipment can be according to each primary election
Longitude station, Position Latitude and intersection location corresponding to effective image, determine corresponding to each effective image of the primary election
Stitching position;It is closed by the distance for calculating the effective image of the second primary election of one or more intersected with the effective image of the first primary election
System determines and the corresponding optimal effective image of second primary election of the effective image of the first primary election;By optimal second primary election
Effective image performs the step for determining the optimal effective image of second primary election as the pending effective image of the first primary election, cycle
Suddenly, to remove the redundancy image in the effective image of the primary election, preferably effective image is obtained.
Specifically, when obtaining preferably effective image, following methods may be used, so that irredundant in the processing equipment
Minimum image cover the target splicing regions, reduce calculation amount, improve image processing efficiency:
First, the processing equipment determine longitude station corresponding to each effective image of the primary election, Position Latitude with
And intersection location, then, based on above-mentioned position relationship, to determine the stitching position corresponding to each effective image of primary election.
Here, using longitude station, Position Latitude and intersection location as three constraintss, to the determining method of stitching position into
Row explanation.
Constraints 1 (longitude):The effective image of primary election of the condition of satisfaction is filtered out in longitude range.Firstly, for original
The effective image of each primary election in target image has choosing a fritter from certain side (such as certain row leftmost side) as initial primary election
Image is imitated, finds out and meets each primary election effective image of the midpoint longitude more than the midpoint longitude of the initial effective image of primary election, to protect
It is (effective including overlapping and nonoverlapping primary election on the right of the effective image of initial primary election to demonstrate,prove the effective image of each primary election filtered out
Image);Then the center position coordinates information of all effective images of primary election in the right is found out from tif files.
Constraints 2 (latitude):Fig. 4 (1) shows a kind of undesirable image joint combination.In order to avoid the feelings of Fig. 4 (1)
Condition, in order to prevent the effective image of each primary election that search obtains larger in vertical direction deviation, so added in latitude scope again
Constraints, the constraints are the ranges that the effective image of initial primary election (influence i.e. shown in dotted line) is made to be searched in vertical direction
It is from the distance of M, N point on midpoint P to up-and-down boundary line, i.e. the y direction in latitude and longitude coordinates system selects affiliated latitude
The slice of range, so that shown in final search splicing effect such as Fig. 4 (2).
In Fig. 4 (2), A is the effective image of initial primary election, and B is the effective image of primary election of redundancy, and C searches for obtain
It is preferred that the effective image of primary election.
Those skilled in the art will be understood that next can using image C as initial image, then in the manner described above after
Continue to the right by described, it is known that search the effective image of primary election of the rightmost in the original target image.
Constraints 3 (intersecting):Fig. 4 (3) shows another undesirable image joint combination.More than pair warp and weft degree adds
After constraints, in order to avoid there is the disjoint situation of each corresponding effective image of primary election in Fig. 4 (3), should also it add intersecting
Condition.
As shown in Fig. 4 (4), if the coordinate vector on the effective image AB directions boundary of original primary election be a, BD directions boundary to
Measure as b, the vector on DC directions boundary is c, and the vector on CA directions boundary is d, then in order to meet intersecting principle, calculate with it is initial just
Select midpoint P on the boundary line of the effective image of the primary election that effective image intersects, border vertices F, Q respectively with the effective shadow of original primary election
As the Outer Product of Vectors on each boundary.
Wherein, the Outer Product of Vectors on midpoint P and each boundary of the effective image of original primary election (AB, BD, DC, CA) calculates as follows respectively
It is shown:
PA×a
PB×b
PC×d
PD×c
Wherein, PA refers to P points to the vector of A points, and a is A points to the vector of B points, and PA × a is vector of the P points to A points
With the apposition of the vector of A points to B points.The computational methods of the apposition are:| PA × a |=| PA | | a | sin<PA,a>, also
It is to say when the Outer Product of Vectors is negative, illustrates that two vectorial angles are negative, it is meant that two polygons intersect.
Similarly, the apposition between PB × b, as vector PB and b, the apposition between PC × d, as vector PC and d, PD
Apposition between × c, as vector PD and d.
When the result of aforementioned four apposition is less than zero, illustrate that P points are fallen in the initial effective image ABCD of primary election,
Then the effective image of two primary election intersects.
Wherein, the Outer Product of Vectors on border vertices F and each boundary of the effective image of original primary election (AB, BD, DC, CA), which calculates, distinguishes
It is as follows:
FA×a
FB×b
FC×d
FD×c
It is similar with the calculating of the Outer Product of Vectors on each boundary of the effective image of original primary election (AB, BD, DC, CA) with midpoint P, work as FA
When this four appositions of × a, FB × b, FC × d, FD × c are respectively less than zero, illustrate that F points are fallen in the initial effective image ABCD of primary election
In, then the effective image of two primary election intersects.
Wherein, the Outer Product of Vectors on border vertices Q and each boundary of the effective image of original primary election (AB, BD, DC, CA), which calculates, distinguishes
It is as follows:
QA×a
QB×b
QC×d
QD×c
It is similar with the calculating of the Outer Product of Vectors on each boundary of the effective image of original primary election (AB, BD, DC, CA) with midpoint P, work as QA
When this four appositions of × a, QB × b, QC × d, QD × c are respectively less than zero, illustrate that Q points are fallen in the initial effective image ABCD of primary election
In, then the effective image of two primary election intersects.
In other words, four appositions that above-mentioned every group of apposition (corresponding respectively to P points, F points and Q points) is calculated (represent
The apposition of the point and vector a, b, c, d) when being respectively less than zero, then prove that any point in 3 points of point P, F, Q is fallen in initial primary election
In effective image, two effective images of primary election intersect;And any one apposition in arbitrary a line is more than zero, then proves at the beginning of two
Select effective image non-intersect.
After the primary election effective image for meeting above three constraints is obtained, then by the effective image of one of primary election
As the effective image of the first primary election, then, each effective image of second primary election intersected with the effective image of the first primary election is calculated
Distance relation, such as pass through the distance relation of the midpoint and the effective image of the first primary election that calculate the effective image of the second primary election.It will tool
Have the effective image of the second primary election of maximum distance relationship as with the effective image of the first primary election corresponding optimal second at the beginning of
Select effective image.
And so on, using the optimal effective image of second primary election as the pending effective image of the first primary election, cycle
The step of determining optimal second primary election effective image is performed, to remove the redundancy image in the effective image of the primary election, is obtained
Obtain preferably effective image.
Those skilled in the art will be understood that is performed both by aforesaid operations for the effective image of the primary election of every row or each column, with
It obtains and all preferred effective image corresponding to the target splicing regions.Those skilled in the art will be understood that open from row
Begin, from left to right performs detection or since row, perform detection from top to bottom and can complete obtaining for preferred effective image
It takes, i.e., detection direction is not fixed.As long as ensure that circulate operation covers entire target area.
In step s 4, the processing equipment extraction preferably effective boundary of the image in the initial target image
Line.
Specifically, in step s 4, the processing equipment can be determined according to the stitching position of preferably effective image
And extract the preferably effective boundary line of the image in the initial target image;Alternatively, in step s 4, the processing is set
It is standby to determine the preferably effective fit line of the image in the initial target image, then, according on the fit line
Pixel and with the pixel band corresponding to the pixel, according on the pixel and the pixel band pixel away from
From aberration, the boundary line of preferably effective image is extracted.
Preferably, the step S4 includes step S41 (not shown) and step S42 (not shown);Wherein, in step S41
In, the processing equipment determines the preferably effective fit line of the image in the initial target image;In step S42,
Pixel band of the processing equipment corresponding to according to the pixel on the fit line and with the pixel, according to the picture
On vegetarian refreshments and the pixel band pixel apart from aberration, extract the preferably effective boundary line of image.
In step S41, the processing equipment is determined and is extracted described according to the stitching position of preferably effective image
It is preferred that fit line of effective image in the initial target image, wherein, the fit line, that is, approximate bounds line.
Alternatively, it is highly preferred that in step S41, the processing equipment can be according to the boundary of preferably effective image
Point determines the preferably effective initial fit line of the image in the initial target image;According to preferably effective image
Splicing direction when splicing in the initial target image, the initial fit line is translated, with described in acquisition
It is preferred that fit line of effective image in the initial target image.
Specifically, in step S41, the processing equipment is detected according to the stitching position of preferably effective image
The value of chromatism of each pixel near the stitching position extracts the point of value of chromatism maximum;Per adjacent point-to-point transmission on boundary line
Vector there is size and Orientation similitude, the point of size or direction dissmilarity is got rid of by threshold value;Pass through one or more
Such pixel, it may be determined that the preferably effective initial fit line of the image in the initial target image;Such as Fig. 5
(2) with regard to AB, CD two lines in Fig. 6 (2), as described initial fit line (or line of prediction).
Then, it is taken from the initial fit line a little, takes several pixel bands from the location of pixels where the point, find out picture
The RGB of each pixel of the element with range and the difference of previous pixel RGB color (calculate two pixels in rgb space apart from color
Difference), i.e., several adjacent pixels are made with RGB is poor two-by-two, selects the point of difference maximum as boundary point.In order to make boundary point institute structure
Into boundary line preferably close to fit line, look for take boundary point when, the pixel band of selection cannot take too long (such as 20 pixels), because
It is more than pixel band intermediate region RGB with the difference of adjacent pixel RGB color that boundary position occurs for the pixel in order to avoid selection
The difference of color.Then the splicing direction according to preferably effective image when splicing in the initial target image, often
Every several step-lengths, every row is continued search in the above manner.Finally the coordinate where the point on the boundary line found out is carried out
Duplicate removal according to the identification point without repetition, therefore obtains unique borderline identification point, so that it is determined that going out fit line.
It is highly preferred that in step S41, the processing equipment can also be corrected behaviour to the fitting boundary line extracted
Make, such as translating step and/or translate up step-length in the horizontal direction to the right in vertical direction.
Specifically, 4 points of each boundary preferably effectively where image correspond to a range substantially in big figure.
During being combined into big figure in small figure, resolution ratio can continuously decrease, and according to the preferably effectively image described initial
The deviation of splicing direction when splicing on target image, initial fit line and fit line can be increasing.If for example, arranging
Direction be from right to left, the direction of row from the top down, fit line is initial to hang down during vertical direction generates from right to left
The deviation of straight fit line and vertical fit line is increasing;Similarly, in the horizontal direction from the top down during, initial level intend
Zygonema and the deviation of horizontal fit line are increasing.In order to reduce the error between initial fit line and fit line, such as Fig. 5 (3)
And as shown in Fig. 6 (3), in the vertical direction, several step-lengths that each initial fit line should be stepped up respectively to right translation;
Similarly, in the horizontal direction, each initial fit line should be translated up to several step-lengths gradually reduced, so that it is determined that fitting respectively
Line.
In step s 5, the processing equipment is iterated even color according to the boundary line to the preferably effectively image
Processing, to obtain the target image corresponding to the target splicing regions.
Specifically, in step s 5, the processing equipment such as utilizes the side of Histogram Matching according to the boundary line
Formula, the region adjacent to the boundary line is iterated even color processing, to obtain the target corresponding to the target splicing regions
Image.The even color of iteration, i.e., even color selects certain even colour band length, and gradually adjust even colour band length every time.
Preferably, the step S5 includes step S51 (not shown) and step S52 (not shown), wherein, in step
In S51, the processing equipment determines the subregion preferably effectively corresponding to image according to the boundary line;In step S52
In, subregion of the processing equipment according to corresponding to the preferably effectively image is iterated even color to adjacent subregion successively
Processing, to obtain the target image corresponding to the target splicing regions.
Specifically, in step s 51, the processing equipment chooses the excellent of the boundary line both sides according to the boundary line
Select the subregion to be treated of effective image;Then, in step S52, the processing equipment is according to two adjacent subregions, such as
Using the mode of Histogram Matching, it is iterated even color processing successively, to obtain the target corresponding to the target splicing regions
Image.
It is highly preferred that in step S52, the processing equipment can be according to point corresponding to the preferably effectively image
Area determines at least one set of adjacent subregion;From the adjacent sectors, iteration chooses the even colour band with different length, goes forward side by side
The processing of row iteration even color, to obtain the target image corresponding to the target splicing regions, wherein, the length of the even colour band with
Gradual change parameter is directly proportional.
Specifically, the processing equipment chooses the preferred effective image of the boundary line both sides according to the boundary line
Subregion to be treated;If the area code template belonging to different subregions should be used to carry out even comprising multiple subregions in the range of some
Color.
In order to make the comparison of the image transition where every piece of splicing part preferably effectively where image naturally, according to
The subregion that effectively image is included each preferably is simultaneously iterated even color, but to ensure the even color chosen successively every certain step-length
The length Length of band and the relative smooth of the parameter beta variation of gradual change, not there is mutation.
Then, the processing equipment determines that at least one set of adjacent subregion is handled.Here, the side with Histogram Matching
Formula carries out the explanation of even color mode:
In order to reduce the splicing aberration of image, luminance difference, using Histogram Matching principle, carry out between each adjacent sectors
Even color.So-called Histogram Matching, i.e., using the histogram of standard video as standard, by the histogram of another width image to standard Nogata
Figure is converted, and makes the histograms of two images close to similar.The tone and contrast of two images can be adjusted by Histogram Matching,
So as to which the Luminance Distribution for making two images approximately reaches unanimity, in order to keep the population equilibrium of picture tone, completely one is generated
The visual effect of body.Using subregion T as template, subregion B is obtained after carrying out Histogram Matching to its adjacent subregion A, then use with
Lower weighting parameters mode carries out gradual change to A, so as to the A' after being changed.
A'=γ A+ (1- γ) B
Wherein, γ is fade factor weight (i.e. gradual change parameter), and w ∈ (0,1).From left image to the weight of right image transition
In folded region process, weight w fades to 0 from 1, therefore just forms the smooth registration of two images.In addition, in order to preferably eliminate plan
Zygonema is iterated in the above manner, the scheme gradually reduced to the intersecting area of each contiguous slices into row pixel value step-length
Even color is carried out, so as to fulfill the consistency of result in tone and geometrically is inlayed.
It should be noted that the present invention can be carried out in the assembly of software and/or software and hardware, for example, can adopt
With application-specific integrated circuit (ASIC), general purpose computer or any other realized similar to hardware device.In one embodiment
In, software program of the invention can perform to realize steps described above or function by processor.Similarly, it is of the invention
Software program can be stored in computer readable recording medium storing program for performing (including relevant data structure), for example, RAM memory,
Magnetic or optical driver or floppy disc and similar devices.In addition, hardware can be used to realize in some steps or function of the present invention, example
Such as, as coordinating with processor so as to perform the circuit of each step or function.
In addition, the part of the present invention can be applied to computer program product, such as computer program instructions, when its quilt
When computer performs, by the operation of the computer, it can call or provide according to the method for the present invention and/or technical solution.
And the program instruction of the method for the present invention is called, it is possibly stored in fixed or moveable recording medium and/or passes through
Broadcast or the data flow in other signal loaded mediums and be transmitted and/or be stored according to described program instruction operation
In the working storage of computer equipment.Here, including a device according to one embodiment of present invention, which includes using
Memory in storage computer program instructions and processor for executing program instructions, wherein, when the computer program refers to
When order is performed by the processor, method and/or skill of the device operation based on aforementioned multiple embodiments according to the present invention are triggered
Art scheme.
It is obvious to a person skilled in the art that the present invention is not limited to the details of above-mentioned exemplary embodiment, Er Qie
In the case of without departing substantially from spirit or essential attributes of the invention, the present invention can be realized in other specific forms.Therefore, no matter
From the point of view of which point, the present embodiments are to be considered as illustrative and not restrictive, and the scope of the present invention is by appended power
Profit requirement rather than above description limit, it is intended that all by what is fallen within the meaning and scope of the equivalent requirements of the claims
Variation includes within the present invention.Any reference numeral in claim should not be considered as to the involved claim of limitation.This
Outside, it is clear that one word of " comprising " is not excluded for other units or step, and odd number is not excluded for plural number.That is stated in device claim is multiple
Unit or device can also be realized by a unit or device by software or hardware.The first, the second grade words are used for table
Show title, and do not represent any particular order.
Claims (20)
1. a kind of method for remote sensing image splicing, wherein, this method includes the following steps:
Obtain multiple remote sensing images;
Extract effective image that the remote sensing image is included;
According to multiple effective images, determine preferably effective image corresponding with target splicing regions and generate and the mesh
Mark the corresponding initial target image of splicing regions;
Extract the preferably effective boundary line of the image in the initial target image;
According to the boundary line, even color is iterated to preferably effective image and is handled, to obtain the target splicing regions
Corresponding target image.
2. according to the method described in claim 1, wherein, the step of effective image that every remote sensing image of extraction is included, wraps
It includes:
The remote sensing image is progressively scanned, when scanning is at least one Pixel Information, according to the Pixel Information
Position determines and corresponding at least one edge line of the remote sensing image;
According to the edge line, effective image that the remote sensing image is included is extracted.
3. according to the method described in claim 1, wherein it is determined that with corresponding preferably effectively image and the life of target splicing regions
Include into the step of initial target image corresponding with the target splicing regions:
According to multiple effective images, filter out with the effective image of the corresponding primary election of target splicing regions, and determine each
Stitching position corresponding to the effective image of primary election;
According to the stitching position corresponding to each effective image of primary election, the redundancy shadow in the effective image of the primary election is removed
Picture, to obtain preferably effective image;
The preferably effectively image is spliced, with generation and the corresponding initial target image of target splicing regions.
4. according to the method described in claim 1, wherein it is determined that with corresponding preferably effectively image and the life of target splicing regions
Include into the step of initial target image corresponding with the target splicing regions:
According to multiple effective images, filter out with the effective image of the corresponding primary election of target splicing regions, and determine each
Stitching position corresponding to the effective image of primary election;
The effective image of the primary election is spliced, with generation and the corresponding original target image of target splicing regions;
According to stitching position of each effective image of primary election in the original target image, the effective shadow of the primary election is removed
Redundancy image as in, the initial target image with the preferably effective image of acquisition and corresponding to the preferred effectively image.
5. method according to claim 3 or 4, wherein, the effective image of screening primary election simultaneously determines the primary election effective image
The step of corresponding stitching position, includes:
According to the coordinate position relationship of effective image and target splicing regions, filtered out from image effective multiple described with
The effective image of the corresponding primary election of target splicing regions, and determine the stitching position corresponding to each effective image of primary election.
6. method according to claim 3 or 4, wherein, the step of obtaining preferably effective image, includes:
According to longitude station, Position Latitude and the intersection location corresponding to each effective image of primary election, each institute is determined
State the stitching position corresponding to the effective image of primary election;
By calculating the distance relation of the effective image of the second primary election of one or more intersected with the effective image of the first primary election, determine
With the corresponding optimal effective image of second primary election of the effective image of the first primary election;
Using the optimal effective image of second primary election as the pending effective image of the first primary election, cycle perform determine it is described most
The step of excellent second primary election effective image, to remove the redundancy image in the effective image of the primary election, obtain preferably effective image.
7. method according to any one of claim 1 to 6, wherein, the preferably effectively image is extracted described initial
The step of boundary line in target image, includes:
Determine the preferably effective fit line of the image in the initial target image;
Pixel band corresponding to according to the pixel on the fit line and with the pixel, according to the pixel and institute
State pixel on pixel band apart from aberration, extract the preferably effective boundary line of image.
8. according to the method described in claim 7, wherein it is determined that the step of fit line include:
According to the boundary point of preferably effective image, determine that preferably effective image is first in the initial target image
Beginning fit line;
According to splicing direction of the preferably effective image when splicing in the initial target image, by the initial plan
Zygonema is translated, to obtain the preferably effective fit line of the image in the initial target image.
9. method according to any one of claim 1 to 8, wherein, even color is iterated to the preferably effectively image
The step of processing, includes:
According to the boundary line, the subregion preferably effectively corresponding to image is determined;
According to the subregion corresponding to the preferably effectively image, even color processing is iterated to adjacent subregion successively, to obtain
Target image corresponding to the target splicing regions.
10. according to the method described in claim 9, wherein, the step of even color is handled is iterated to adjacent subregion successively and is wrapped
It includes:
According to the subregion corresponding to the preferably effectively image, at least one set of adjacent subregion is determined;
From the adjacent sectors, iteration chooses the even colour band with different length, and is iterated even color processing, to obtain
The target image corresponding to target splicing regions is stated, wherein, the length of the even colour band is directly proportional to gradual change parameter.
11. a kind of processing equipment for remote sensing image splicing, wherein, the processing equipment includes:
First device, for obtaining multiple remote sensing images;
Second device, for extracting effective image that the remote sensing image is included;
3rd device, for according to multiple effective images, determining and the corresponding preferably effective image of target splicing regions
And it generates and the corresponding initial target image of target splicing regions;
4th device, for extracting the preferably effective boundary line of the image in the initial target image;
5th device, for according to the boundary line, being iterated even color to preferably effective image and handling, with described in acquisition
Target image corresponding to target splicing regions.
12. processing equipment according to claim 11, wherein, the second device is used for:
The remote sensing image is progressively scanned, when scanning is at least one Pixel Information, according to the Pixel Information
Position determines and corresponding at least one edge line of the remote sensing image;
According to the edge line, effective image that the remote sensing image is included is extracted.
13. processing equipment according to claim 11, wherein, the 3rd device includes:
Unit 31, for according to multiple effective images, filtering out and the effective shadow of the corresponding primary election of target splicing regions
Picture, and determine the stitching position corresponding to each effective image of primary election;
Unit three or two for the stitching position according to corresponding to each effective image of primary election, remove the effective shadow of the primary election
Redundancy image as in, to obtain preferably effective image;
Unit three or three, it is corresponding with the target splicing regions to generate for the preferably effectively image to be spliced
Initial target image.
14. processing equipment according to claim 11, wherein, the 3rd device includes:
Unit 31, for according to multiple effective images, filtering out and the effective shadow of the corresponding primary election of target splicing regions
Picture, and determine the stitching position corresponding to each effective image of primary election;
Unit three or four, it is corresponding with the target splicing regions to generate for the effective image of the primary election to be spliced
Original target image;
Unit three or five, for the stitching position according to each effective image of primary election in the original target image, removal
Redundancy image in the effective image of primary election, to obtain preferably effective image and with described preferably effectively corresponding to image
Initial target image.
15. the processing equipment according to claim 13 or 14, wherein, Unit 31 is used for:
According to the coordinate position relationship of effective image and target splicing regions, filtered out from image effective multiple described with
The effective image of the corresponding primary election of target splicing regions, and determine the stitching position corresponding to each effective image of primary election.
16. the processing equipment according to claim 13 or 14, wherein, Unit three or two or Unit three or five are used for:
According to longitude station, Position Latitude and the intersection location corresponding to each effective image of primary election, each institute is determined
State the stitching position corresponding to the effective image of primary election;
By calculating the distance relation of the effective image of the second primary election of one or more intersected with the effective image of the first primary election, determine
With the corresponding optimal effective image of second primary election of the effective image of the first primary election;
Using the optimal effective image of second primary election as the pending effective image of the first primary election, cycle perform determine it is described most
The step of excellent second primary election effective image, to remove the redundancy image in the effective image of the primary election, obtain preferably effective image.
17. the processing equipment according to any one of claim 11 to 16, wherein, the 4th device includes:
Unit 41, for determining the preferably effective fit line of the image in the initial target image;
Unit four or two, for the pixel band corresponding to according to the pixel on the fit line and with the pixel, according to
On the pixel and the pixel band pixel apart from aberration, extract the preferably effective boundary line of image.
18. processing equipment according to claim 17, wherein, Unit 41 is used for:
According to the boundary point of preferably effective image, determine that preferably effective image is first in the initial target image
Beginning fit line;
According to splicing direction of the preferably effective image when splicing in the initial target image, by the initial plan
Zygonema is translated, to obtain the preferably effective fit line of the image in the initial target image.
19. the processing equipment according to any one of claim 11 to 18, wherein, the 5th device includes:
May Day unit, for according to the boundary line, determining the subregion preferably effectively corresponding to image;
Unit five or two for the subregion according to corresponding to the preferably effectively image, are successively iterated adjacent subregion even
Color processing, to obtain the target image corresponding to the target splicing regions.
20. processing equipment according to claim 19, wherein, Unit five or two is used for:
According to the subregion corresponding to the preferably effectively image, at least one set of adjacent subregion is determined;
From the adjacent sectors, iteration chooses the even colour band with different length, and is iterated even color processing, to obtain
The target image corresponding to target splicing regions is stated, wherein, the length of the even colour band is directly proportional to gradual change parameter.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711131587.XA CN108154471A (en) | 2017-11-15 | 2017-11-15 | A kind of method and apparatus for remote sensing image splicing |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711131587.XA CN108154471A (en) | 2017-11-15 | 2017-11-15 | A kind of method and apparatus for remote sensing image splicing |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108154471A true CN108154471A (en) | 2018-06-12 |
Family
ID=62468734
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711131587.XA Withdrawn CN108154471A (en) | 2017-11-15 | 2017-11-15 | A kind of method and apparatus for remote sensing image splicing |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108154471A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109712071A (en) * | 2018-12-14 | 2019-05-03 | 电子科技大学 | Unmanned plane image mosaic and localization method based on track constraint |
CN110097498A (en) * | 2019-01-25 | 2019-08-06 | 电子科技大学 | More air strips image mosaics and localization method based on unmanned aerial vehicle flight path constraint |
CN111311622A (en) * | 2020-01-30 | 2020-06-19 | 桂林理工大学 | Ant colony algorithm-based optimization selection method for multiple image mosaic wire nets |
CN111754397A (en) * | 2019-07-09 | 2020-10-09 | 北京市商汤科技开发有限公司 | Remote sensing image embedding method and device, electronic equipment and storage medium |
CN116309057A (en) * | 2023-03-14 | 2023-06-23 | 中科星图数字地球合肥有限公司 | Remote sensing image splicing method, device, computer equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102663394A (en) * | 2012-03-02 | 2012-09-12 | 北京航空航天大学 | Method of identifying large and medium-sized objects based on multi-source remote sensing image fusion |
CN103218821A (en) * | 2013-04-27 | 2013-07-24 | 航天恒星科技有限公司 | Regional-coverage-oriented automatic image mosaic method |
CN104766328A (en) * | 2015-04-16 | 2015-07-08 | 武汉大学 | Remote-sensing image Bowtie effect correction method based on path tracing |
CN105279735A (en) * | 2015-11-20 | 2016-01-27 | 沈阳东软医疗系统有限公司 | Fusion method, fusion device and fusion equipment of image splicing |
-
2017
- 2017-11-15 CN CN201711131587.XA patent/CN108154471A/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102663394A (en) * | 2012-03-02 | 2012-09-12 | 北京航空航天大学 | Method of identifying large and medium-sized objects based on multi-source remote sensing image fusion |
CN103218821A (en) * | 2013-04-27 | 2013-07-24 | 航天恒星科技有限公司 | Regional-coverage-oriented automatic image mosaic method |
CN104766328A (en) * | 2015-04-16 | 2015-07-08 | 武汉大学 | Remote-sensing image Bowtie effect correction method based on path tracing |
CN105279735A (en) * | 2015-11-20 | 2016-01-27 | 沈阳东软医疗系统有限公司 | Fusion method, fusion device and fusion equipment of image splicing |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109712071A (en) * | 2018-12-14 | 2019-05-03 | 电子科技大学 | Unmanned plane image mosaic and localization method based on track constraint |
CN109712071B (en) * | 2018-12-14 | 2022-11-29 | 电子科技大学 | Unmanned aerial vehicle image splicing and positioning method based on track constraint |
CN110097498A (en) * | 2019-01-25 | 2019-08-06 | 电子科技大学 | More air strips image mosaics and localization method based on unmanned aerial vehicle flight path constraint |
CN111754397A (en) * | 2019-07-09 | 2020-10-09 | 北京市商汤科技开发有限公司 | Remote sensing image embedding method and device, electronic equipment and storage medium |
CN111311622A (en) * | 2020-01-30 | 2020-06-19 | 桂林理工大学 | Ant colony algorithm-based optimization selection method for multiple image mosaic wire nets |
CN111311622B (en) * | 2020-01-30 | 2023-06-23 | 桂林理工大学 | Optimization selection method for embedding wire nets of multiple images based on ant colony algorithm |
CN116309057A (en) * | 2023-03-14 | 2023-06-23 | 中科星图数字地球合肥有限公司 | Remote sensing image splicing method, device, computer equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108154471A (en) | A kind of method and apparatus for remote sensing image splicing | |
US9959653B2 (en) | Mosaic oblique images and methods of making and using same | |
USRE42881E1 (en) | Method and system for scanning images in a photo kiosk | |
RU2368006C1 (en) | Method and system for adaptive reformatting of digital images | |
US8248410B2 (en) | Synthesizing detailed depth maps from images | |
CN104680501B (en) | The method and device of image mosaic | |
US7889947B2 (en) | Image completion | |
Shih et al. | Exemplar-based video inpainting without ghost shadow artifacts by maintaining temporal continuity | |
JP2019053732A (en) | Dynamic generation of image of scene based on removal of unnecessary object existing in the scene | |
JP3690391B2 (en) | Image editing apparatus, image trimming method, and program | |
JP4482778B2 (en) | Image processing apparatus, image processing method, and recording medium | |
JP2002514359A (en) | Method and apparatus for creating a mosaic image | |
JP2001236497A (en) | Method for automatically forming cropped and zoomed deformation of photographic image | |
CN110443141A (en) | Data set processing method, data set processing unit and storage medium | |
CN105825478A (en) | Structure analysis method for recovering missing structures in an image after object removal | |
CN107122792B (en) | Indoor arrangement estimation method and system based on study prediction | |
Chen et al. | 3D texture mapping in multi-view reconstruction | |
Liu et al. | Image contrast enhancement based on intensity expansion-compression | |
JP2010244398A (en) | Image processing apparatus and image processing program | |
CN106504261B (en) | A kind of image partition method and device | |
CN114155150A (en) | Image stitching method and device applied to large parallax scene | |
CN105590307A (en) | Transparency-based matting method and apparatus | |
CN111105351B (en) | Video sequence image splicing method and device | |
US20170372134A1 (en) | Method and subsystem for identifying document subimages within digital images | |
Li et al. | Optimal seamline detection in dynamic scenes via graph cuts for image mosaicking |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20180605 Address after: 100085 Beijing Haidian District Xueyuan Road armor 5, 2 square B North -1081 Applicant after: Beijing star science and Technology Co., Ltd. Address before: 100085 Beijing Haidian District Xueyuan Road a 5, 2 cottage B-1011 Applicant before: Beijing to gather information technology Co. Ltd. |
|
TA01 | Transfer of patent application right | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20180612 |
|
WW01 | Invention patent application withdrawn after publication |