CN102930269B - Method for modulating multiscale semanteme of remote-sensing image - Google Patents
Method for modulating multiscale semanteme of remote-sensing image Download PDFInfo
- Publication number
- CN102930269B CN102930269B CN201210344453.7A CN201210344453A CN102930269B CN 102930269 B CN102930269 B CN 102930269B CN 201210344453 A CN201210344453 A CN 201210344453A CN 102930269 B CN102930269 B CN 102930269B
- Authority
- CN
- China
- Prior art keywords
- sensing image
- remote sensing
- pixel
- handling object
- scale
- 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.)
- Expired - Fee Related
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method for modulating multiscale semanteme of a remote-sensing image. The method comprises the following steps of: selecting local feature types of processing objects of the remote-sensing image; determining the size range of observation scales which are used by the processing objects of the remote-sensing image; partitioning the levels of the observation scales of the processing objects of the remote-sensing image; generating single-pixel feature value sequences; generating multiscale feature values of pixels; and generating a scale manifesting image. According to the method, the internal consistency of the processing objects of the remote-sensing image is improved and the relative property between the processing objects of the remote-sensing image and a background is enhanced by utilizing the acquired scale manifesting image modulated by the principle. Test results prove that the multiscale feature-based scale manifesting image can reflect the actual condition of surface features well, so the method has high using value.
Description
Technical field
The present invention relates to a kind of modulator approach of remote sensing image multiscale semanteme, belong to the remote sensing image preconditioning technique towards the identification of remote sensing image ground object target, can be applicable to the extraction of the ground object target such as road, buildings based on remote sensing image.
Background technology
In remote sensing image information extractive technique field, yardstick has different implications, first refers to the geometric scale of the true atural object described by remote sensing image, and this geometric scale was ignorant before can obtaining terrestrial object information by remote sensing image; Next observes yardstick, this observation yardstick shows as the resolution of remote sensing image on original remote sensing image, the remote sensing image observing yardstick little (high resolving power) can generate according to image-forming principle the remote sensing image observing yardstick large (low resolution), and thus high-resolution remote sensing image is that multiple dimensioned observational study provides the foundation; It is finally the yardstick in Remote Sensing Image Segmentation research, yardstick refers generally to the remote sensing image object size (or controlling the parameter of cutting unit size) split, the parameter controlling cutting unit size is generally inner consistent degree parameter, although there is close relationship between consistent degree and the cell size of segmentation, but be difficult to clearly describe, this is also the aspect being difficult to most in eCognition software manipulate, and must obtain suitable scale parameter value by repetition test.
All characteristics of remote sensing images are all the features under certain yardstick, i.e. yardstick characteristics of remote sensing image, general referred to as characteristics of remote sensing image or feature, when studying multiple yardstick remote sensing image, be referred to as Analysis On Multi-scale Features, but carry out the feature generation net result that single scale research finally considers each yardstick more under study for action often respectively.Fractal object has feature not with the characteristic of dimensional variation; More general situation is that certain feature only can remain unchanged within the scope of some scale, and SIFT method employs this feature; On remote sensing image, atural object relation is complicated, and feature changes along with the change of yardstick, and closely related with its Environmental effect.
In the multiscale morphology image identification patent documentation related to, its disposal route is generally identify under each single scale image, is finally merged by the recognition result under different scale condition; In the multiscale morphology Image Segmentation related to, the general order adopting yardstick to increase progressively produces the segmentation result of different scale successively, is finally merged by the segmentation result under different scale condition, can see 200710304466.0,200810032390.5,201110137557.6, as shown in Figure 1.In above-mentioned patent, when using yardstick to analyze, scale feature lies in remote sensing image context, and uses single scale feature in the application, and does not consider multiple dimensioned variation characteristic.And as high-resolution remote sensing image, the high detail characteristic that will reflect remote sensing image of precision on the one hand, the region of remote sensing image is all larger on the other hand, thus remote sensing image has multiple dimensioned characteristic, multiple dimensioned characteristic is the important evidence of remote sensing target identification, if the yardstick characteristic image clear generated can not describe this scale feature by explicitly, then can affect the image enhaucament ability of remote sensing image.
Summary of the invention
The technology of the present invention is dealt with problems: overcome the deficiencies in the prior art, a kind of modulator approach of remote sensing image multiscale semanteme is provided, by a kind of multiscale semanteme modulator approach by clear for remote sensing image scale feature in the pixel of remote sensing image, generate yardstick image clear, the yardstick image explicitly clear generated describes remote sensing image scale feature, the dimensional variation feature of remote sensing image can be reflected, belong to high-order characteristics of remote sensing image, for the identification of high-level semantic remote sensing image object.
The technical solution adopted for the present invention to solve the technical problems is as follows: a kind of modulator approach of remote sensing image multiscale semanteme, and implementation step is as follows:
(1) the local feature type of selected remote sensing image handling object and selected remote sensing image handling object; Described remote sensing image handling object refers in remote sensing image, an interested pending region; The local feature of remote sensing image handling object is reflected in the relative characteristic of the interior remote sensing image handling object adjacent with other of subrange of this remote sensing image handling object, relative characteristic between the pixel that the local feature of remote sensing image handling object is also reflected between set of pixels that this remote sensing image handling object comprises and remote sensing image handling object adjacent with other comprises, utilizes the local feature information of remote sensing image handling object as the similarity foundation judged between remote sensing image handling object;
(2) on the basis of selected remote sensing image handling object, the size of the observation yardstick of used remote sensing image handling object is determined according to the geometric scale size of selected remote sensing image handling object;
(3) employing scale feature curve expresses the local feature situation of change of pixel in observation range scale that remote sensing image handling object comprises in a continuous manner; Adopt discrete form to carry out sampling and encoding for scale feature curve, by step 2 determine the observation scale size scope of remote sensing image handling object according to m times of progression be divided into n observe yardstick;
(4) according to the remote sensing image handling object local feature type of step 1, calculate the local feature under n divided in step 3 observation yardstick of each pixel in remote sensing image, in remote sensing image, each pixel becomes single pixel characteristic value sequence of this pixel at the local feature that n obtains under observing yardstick; In remote sensing image, each pixel has single pixel characteristic value sequence of this pixel;
(5) need to encode to be summarised as an eigenwert of this pixel to single pixel characteristic value sequence of pixel each in the remote sensing image obtained in step 4, as the Analysis On Multi-scale Features value of this pixel in remote sensing image, to describe the Analysis On Multi-scale Features on remote sensing image pixel, in remote sensing image, a pixel Analysis On Multi-scale Features value can reflect that pixel observes the local feature of yardstick at certain, also can reflect the local feature of this pixel on each yardstick observing range scale simultaneously;
(6) in step 5, in remote sensing image, each pixel generates an Analysis On Multi-scale Features value, using the Analysis On Multi-scale Features value of this pixel as the pixel value of this pixel, form a width identical with original remote sensing image size, the remote sensing image of the Analysis On Multi-scale Features value that this pixel value shows in original remote sensing image for this pixel, i.e. yardstick image clear.
The scope of the observation scale size of the employing remote sensing image handling object in described step (1) is 2 times of the scope of remote sensing image handling object geometric scale size.
M=2 in described step (3).
The present invention's advantage is compared with prior art:
(1), in the process of traditional multiscale transform remote sensing image, be the remote sensing image process carrying out towards final goal to multiple yardstick, and by each result simple combination, dimensional variation feature cannot be used; The yardstick characteristic image explicitly clear that the present invention generates describes this feature, can reflect the dimensional variation feature of remote sensing image, belong to high-order characteristics of remote sensing image, may be used for the identification of high-level semantic object.
(2) yardstick of the present invention image clear is using yardstick as modulation framework, under remote sensing image handling object characteristic type elects bright dark characteristic type as, carry out the method for the invention carry out modulating obtained yardstick image clear, the spectral characteristic of raw video can be retained.
(3), in the process of traditional multiscale transform remote sensing image, the feature under large scale often departs from real location point because resolution reduces; As described in step 6, and the feature on yardstick of the present invention image clear is all generate Analysis On Multi-scale Features value based on pixel each in original remote sensing image, forms yardstick image clear, thus maintains the positional precision of original remote sensing image.
(4) as described in step 4 and step 5, in yardstick of the present invention characteristic image clear, each pixel value is this pixel Analysis On Multi-scale Features value, summarized by the single scale characteristic sequence of this pixel to form, can part or specific several yardstick be selected very easily to carry out being summarised as the Analysis On Multi-scale Features value of this pixel when summarizing single scale characteristic sequence, thus the reinforcement realized yardstick or suppression, with the spatial contrast degree of certain yardstick object outstanding, this is difficult to accomplish in conventional methods where.
(5) what the pixel value in yardstick of the present invention image clear reflected is the multiple dimensioned performance of this pixel in raw video, thus the similarity in yardstick image clear between neighbor reflects these pixels in original remote sensing image and whether belongs to same remote sensing image handling object, Analysis On Multi-scale Features in image clear for yardstick of the present invention can be applied in classification or segmentation research accordingly, this is a kind of new similarity definition, can make up in classic method that to lack dimensional variation similarity not enough, make segmentation or nicety of grading higher.
Accompanying drawing explanation
Fig. 1 is usual scale processing process schematic;
Fig. 2 is for using scale processing process schematic of the present invention;
Fig. 3 is yardstick of the present invention image modulated process clear process flow diagram;
Fig. 4 is the spectrum section of object of the present invention in contiguous range, and wherein W is pixel or remote sensing image handling object, and A, B, C are its different Size of Neighborhood;
Fig. 5 is the relative brightness-scale feature change curve of object of the present invention;
Fig. 6 modulates lamppost model in the present invention;
Fig. 7 is yardstick of the present invention image effect figure clear, and the picture left above is original remote sensing image, and top right plot is the yardstick image clear that the picture left above is corresponding, and lower-left figure is another width original remote sensing image, and bottom-right graph is the yardstick image clear that lower-left figure is corresponding.Therefrom can find out, yardstick image energy clear better expresses remote sensing image handling object.
Embodiment
The invention belongs to the remote sensing image preconditioning technique extracted towards remote sensing image ground object target, for the space scale meaning of one's words of remote sensing image handling object clear, it by a kind of multiscale semanteme modulator approach by clear for remote sensing image scale feature in the pixel of remote sensing image, generate yardstick image clear, the yardstick image explicitly clear generated describes remote sensing image scale feature, the dimensional variation feature of remote sensing image can be reflected, belong to high-order characteristics of remote sensing image, can be used for the identification of high-level semantic remote sensing image object.
Yardstick image modulated process clear process flow diagram is as Fig. 3.Comprising selecting the local feature type expressing remote sensing image handling object, determine observe scale size scope and observe scale level for concrete remote sensing image handling object, local feature value is extracted under difference observes yardstick, form the single scale characteristic sequence of pixel, and carry out single scale characteristic value sequence summarizing the scale feature value forming pixel, finally form yardstick image clear.
As shown in Figure 2,3, concrete steps of the present invention are as follows.
1. select the local feature type of remote sensing image handling object
The local feature type of selected remote sensing image handling object and selected remote sensing image handling object; The local feature of remote sensing image handling object is reflected in the relative characteristic of the interior remote sensing image handling object adjacent with other of subrange of this remote sensing image handling object, relative characteristic between the pixel that the local feature of remote sensing image handling object is also reflected between set of pixels that this remote sensing image handling object comprises and remote sensing image handling object adjacent with other comprises, utilizes the local feature information of remote sensing image handling object as the similarity foundation judged between remote sensing image handling object.Local feature divides into consistance characteristics and comparison feature, consistance feature is the common trait had between remote sensing image handling object in the certain area in remote sensing image, such as average, deviation, texture etc., contrast characteristic is that other remote sensing image handling object carries out contrasting and the feature presented in remote sensing image handling object and surrounding neighbors, such as light and shade, number, foreground-background, linear direction etc.
2. determine the observation scale size scope that remote sensing image handling object adopts
On the basis of the remote sensing image handling object of selected observation, determine the size of the observation yardstick of used remote sensing image handling object according to the geometric scale size of selected remote sensing image handling object; The present invention adopt the scope of the observation scale size of remote sensing image handling object to be 2 times of the scope of remote sensing image handling object geometric scale size; This step is mainly for the observation scale size scope of remote sensing image handling object geometric scale size determination remote sensing image handling object.Observe a remote sensing image handling object, so the observation scale size of remote sensing image handling object is at least greater than the geometric scale of this remote sensing image handling object itself.Observe range scale too little, the geometric scale size of remote sensing image handling object can not be reflected, observe scale size scope too large, then be not enough to the local feature in the subrange of this remote sensing image handling object of accurate expression, thus under the present invention considers energy accurate expression remote sensing image handling object local feature prerequisite, consider the problem of implementation efficiency simultaneously, determine that the observation scale size scope of remote sensing image handling object is about 2 times of its geometric scale magnitude range, the minimum value of namely observing yardstick is the half of remote sensing image handling object detail characteristic yardstick, maximal value is 2 times of overall remote sensing image handling object geometric scale, remote sensing image handling object is extended for unlimited, such as road, using its width as geometric scale foundation.
3. divide remote sensing image handling object and observe mid-scale level
Employing scale feature curve expresses the local feature situation of change of pixel in observation range scale that remote sensing image handling object comprises in a continuous manner; A pixel of the pixel comprised with remote sensing image handling object illustrates as shown in Figure 4, and wherein P, Q, W are pixel or remote sensing image handling object, and A, B, C are its different Size of Neighborhood.A bit bright or dark for showing as its neighborhood, and scope is different, and it is bright dark also different, in figure, the position of W shows as secretly in A neighborhood, then show as bright in B neighborhood, then show as dark in wider C neighborhood, the combination of this " dark-light-dark " just intactly have expressed the space scale change to attributes of this point, compares have expressed more semantic relation with simple pixel value, and can be understood as this point is the relatively bright point of in blackening in clear zone one.Fig. 5 display pixel point is its relative change curve of brightness in its neighborhood under different scale.A scale feature sequence will be obtained in discrete scale processing.
Scale feature curve can express pixel very well in the consecutive variations situation of observing the observation of the difference in range scale yardstick lower eigenvalue, but only can adopt discrete form in a computer, thus need to carry out sampling and encoding, namely divide in observation range scale and observe mid-scale level.The present invention is determining that remote sensing image handling object is observed on scale size scope, institute in step 2 is determined the observation scale size scope of remote sensing image handling object is divided into n observation yardstick according to 2 times of progression.Be convenient to expression and the post-processed of remote sensing image like this.Be about to observe metric space and be divided into n rank, higher leveled observation scale size is 2 times of the observation scale size of adjacent lower one-level, such division is operationally convenient to realize, and in theory, observe yardstick be after 2 times of sizes of previous observation yardstick, more can accurately and don't redundancy reflect remote sensing image handling object geometric scale information.
4. generate single pixel characteristic value sequence
According to the remote sensing image handling object local feature type of step 1, calculate the local feature under n divided in step 3 observation yardstick of each pixel in remote sensing image, in remote sensing image, each pixel becomes single pixel characteristic value sequence of this pixel at the local feature that n obtains under observing yardstick; In remote sensing image, each pixel has single pixel characteristic value sequence of this pixel.
5. generate pixel Analysis On Multi-scale Features value
Need to encode to be summarised as an eigenwert of this pixel to single pixel characteristic value sequence of pixel each in the remote sensing image obtained in step 4, as the Analysis On Multi-scale Features value of this pixel in remote sensing image, to describe the Analysis On Multi-scale Features on remote sensing image pixel, in remote sensing image, a pixel Analysis On Multi-scale Features value can reflect that pixel observes the local feature of yardstick at certain, also can reflect the local feature of this pixel on each yardstick observing range scale simultaneously.The coded representation that the present invention expresses Analysis On Multi-scale Features value is:
wherein, S is the Analysis On Multi-scale Features value of the pixel formed, i.e. global feature value; K is for observing mid-scale level k=1 ... n; t
kfor kth observes the switch of yardstick; w
kfor kth observes yardstick weight, observe yardstick weight and reflect the lower scale feature showed of different observation to the impact of global feature value; f
kfor kth observes the scale feature value under yardstick, the observation feature of reflection pixel under kth observes yardstick.When the weight of adjacent scale feature is 2 times of relations, w
k=2
k, a bit position can be used to describe the feature under single observation yardstick, and achieve the explicit description of feature under single observation yardstick, global feature value then describes multiple dimensioned comprehensive characteristics.
6. generate yardstick image clear
In steps of 5, in remote sensing image, each pixel generates an Analysis On Multi-scale Features value, using the Analysis On Multi-scale Features value of this pixel as the pixel value of this pixel, form a width identical with original remote sensing image size, but the remote sensing image of the Analysis On Multi-scale Features value that pixel value shows in original remote sensing image for this pixel, i.e. yardstick image clear.Why being referred to as image clear is because the pixel value of each pixel has not been simple radiance value or relative brightness value, but a kind of eigenwert of reacting under the multiple dimensioned observation of this point.
To the process of as above 6 steps, be referred to as modulated process, by the pixel value of primitive reaction terrestrial radiation intensity, from gray scale codomain, the observations (single scale feature) of being observed under yardstick by difference is modulated to Analysis On Multi-scale Features codomain.In realization, observe between yardstick and adopt equal proportion relation, the weight ratio of adjacent observation scale feature value selects 2 times of relations, and the modulated process of such yardstick characteristic image clear can pass through lamppost model realization.Modulator approach based on bright dark feature is: as shown in Figure 6, imagine the pillar that a frame stands in pixel center, the binary number of original pixel value is arranged in order from top to bottom, each binary digit arranges a lamp, the scope that illuminates of every lamp is 2 times that below it, adjacent lamps must illuminate scope, if the binary digit of correspondence is 1, then this position etc. are opened, otherwise this lamp is closed, then total brightness represents primary radiation gray-scale value.During modulation, opening and closing based on following judgment principle lamp in each binary digit, if namely the gray-scale value of this point belongs to bright pixel in the ranges of exposures such as this position, then turns on this lamp, otherwise closes.Binary bit sequence reflection after such modulation be the Analysis On Multi-scale Features based on bright dark feature on this aspect.The modulator approach of further feature is identical.
In lamppost model, binary pixel values on yardstick image clear illustrates the image feature of each yardstick level, only turn on one deck lamp (only being retained this value by masking operation) then yardstick image display clear be the characteristics of remote sensing image observed under this observation yardstick.When the lamp a high position extinguishes successively, the observation yardstick of the characteristics of remote sensing image that yardstick image clear describes will reduce gradually, finally only see micro-texture, and lose whole object; Otherwise little texture disappears, and large objects highlights.
Yardstick image clear enhances difference between ground object target object by considering Analysis On Multi-scale Features, can be used for more accurately carrying out Remote Sensing Image Segmentation, similarity measurement in cutting procedure can be comprehensive Analysis On Multi-scale Features value, also can be the dimensional variation form that Analysis On Multi-scale Features position describes.During using comprehensive Analysis On Multi-scale Features value as measuring similarity standard, the result of segmentation has the light and shade consistance of comprehensive observing; During using multiple dimensioned change shape as measuring similarity standard, pixel consistent for scale feature metamorphosis can be classified as a class, realize the unification of the similar remote sensing image handling object of different scale.
The principle of the invention is utilized to modulate shown in the yardstick obtained image effect Fig. 7 clear, the picture left above is original remote sensing image, top right plot is the yardstick image clear that the picture left above is corresponding, and lower-left figure is another width original remote sensing image, and bottom-right graph is the yardstick image clear that lower-left figure is corresponding.Therefrom can find out, yardstick image energy clear better expresses remote sensing image handling object, therefrom can find out, yardstick image energy clear better improves the internal consistency of remote sensing image handling object, strengthens the comparative target of remote sensing image handling object and background.Experiment results proved more can the actual conditions of reasonable reflection atural object based on the yardstick image clear of Analysis On Multi-scale Features, show that the present invention has very strong use value.
The yardstick image effect clear utilizing the principle of the invention to modulate to obtain better can improve the internal consistency of remote sensing image handling object, strengthen the comparative target of remote sensing image handling object and background.Experiment results proved more can the actual conditions of reasonable reflection atural object based on the yardstick image clear of Analysis On Multi-scale Features, show that the present invention has very strong use value.
The content be not described in detail in instructions of the present invention belongs to the known prior art of professional and technical personnel in the field.
The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, under the premise without departing from the principles of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (3)
1. a modulator approach for remote sensing image multiscale semanteme, is characterized in that performing step is as follows:
(1) the local feature type of selected remote sensing image handling object and selected remote sensing image handling object; Described remote sensing image handling object refers in remote sensing image, an interested pending region; The local feature of remote sensing image handling object is reflected in the relative characteristic of the interior remote sensing image handling object adjacent with other of subrange of this remote sensing image handling object, relative characteristic between the pixel that the local feature of remote sensing image handling object is also reflected between set of pixels that this remote sensing image handling object comprises and remote sensing image handling object adjacent with other comprises, utilizes the local feature information of remote sensing image handling object as the similarity foundation judged between remote sensing image handling object;
(2) on the basis of selected remote sensing image handling object, the size of the observation yardstick of used remote sensing image handling object is determined according to the geometric scale size of selected remote sensing image handling object;
(3) employing scale feature curve expresses the local feature situation of change of pixel in observation range scale that remote sensing image handling object comprises in a continuous manner; Adopt discrete form to carry out sampling and encoding for scale feature curve, by step 2 determine the observation scale size scope of remote sensing image handling object according to m times of progression be divided into n observe yardstick;
(4) according to the remote sensing image handling object local feature type of step 1, calculate the local feature under n divided in step 3 observation yardstick of each pixel in remote sensing image, in remote sensing image, each pixel becomes single pixel characteristic value sequence of this pixel at the local feature that n obtains under observing yardstick; In remote sensing image, each pixel has single pixel characteristic value sequence of this pixel;
(5) need to encode to be summarised as an eigenwert of this pixel to single pixel characteristic value sequence of pixel each in the remote sensing image obtained in step 4, as the Analysis On Multi-scale Features value of this pixel in remote sensing image, to describe the Analysis On Multi-scale Features on remote sensing image pixel, in remote sensing image, a pixel Analysis On Multi-scale Features value can reflect that pixel observes the local feature of yardstick at certain, also can reflect the local feature of this pixel on each yardstick observing range scale simultaneously;
(6) in step 5, in remote sensing image, each pixel generates an Analysis On Multi-scale Features value, using the Analysis On Multi-scale Features value of this pixel as the pixel value of this pixel, form a width identical with original remote sensing image size, the remote sensing image of the Analysis On Multi-scale Features value that this pixel value shows in original remote sensing image for this pixel, i.e. yardstick image clear.
2. the modulator approach of remote sensing image multiscale semanteme according to claim 1, is characterized in that: the observation scale size scope of the employing remote sensing image handling object in described step (3) is 2 times of the scope of remote sensing image handling object geometric scale size.
3. the modulator approach of remote sensing image multiscale semanteme according to claim 1, is characterized in that: the m=2 in described step (3).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210344453.7A CN102930269B (en) | 2012-09-17 | 2012-09-17 | Method for modulating multiscale semanteme of remote-sensing image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210344453.7A CN102930269B (en) | 2012-09-17 | 2012-09-17 | Method for modulating multiscale semanteme of remote-sensing image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102930269A CN102930269A (en) | 2013-02-13 |
| CN102930269B true CN102930269B (en) | 2015-05-13 |
Family
ID=47645066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210344453.7A Expired - Fee Related CN102930269B (en) | 2012-09-17 | 2012-09-17 | Method for modulating multiscale semanteme of remote-sensing image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102930269B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109376599A (en) * | 2018-09-19 | 2019-02-22 | 中国科学院东北地理与农业生态研究所 | A kind of remote sensing image processing method and system extracted towards wetland information |
| CN110096948B (en) * | 2019-03-15 | 2020-11-17 | 中国科学院西安光学精密机械研究所 | Remote sensing image identification method based on characteristic aggregation convolutional network |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102169545A (en) * | 2011-04-25 | 2011-08-31 | 中国科学院自动化研究所 | Detection method for changes of high-resolution remote sensing images |
| CN102496151A (en) * | 2011-12-08 | 2012-06-13 | 南京大学 | Method for multi-scale segmentation of high-resolution remote sensing images |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2398379A (en) * | 2003-02-11 | 2004-08-18 | Qinetiq Ltd | Automated digital image analysis |
-
2012
- 2012-09-17 CN CN201210344453.7A patent/CN102930269B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102169545A (en) * | 2011-04-25 | 2011-08-31 | 中国科学院自动化研究所 | Detection method for changes of high-resolution remote sensing images |
| CN102496151A (en) * | 2011-12-08 | 2012-06-13 | 南京大学 | Method for multi-scale segmentation of high-resolution remote sensing images |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102930269A (en) | 2013-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ling et al. | Land cover change mapping at the subpixel scale with different spatial-resolution remotely sensed imagery | |
| Hall et al. | A multiscale object-specific approach to digital change detection | |
| CN105956557B (en) | A kind of sequential remote sensing image cloud covered areas domain automatic testing method of object-oriented | |
| Zhou et al. | Multiscale water body extraction in urban environments from satellite images | |
| Sommer et al. | Learning-based mitotic cell detection in histopathological images | |
| CN107067405B (en) | Remote sensing image segmentation method based on scale optimization | |
| CN110428450B (en) | Scale-adaptive target tracking method applied to mine tunnel mobile inspection image | |
| JP5103665B2 (en) | Object tracking device and object tracking method | |
| Tang et al. | A multiple-point spatially weighted k-NN method for object-based classification | |
| JP2014534699A (en) | System and method for digital image signal compression using unique images | |
| Kalkan et al. | Comparison of support vector machine and object based classification methods for coastline detection | |
| Pathak | New Change Detection Techniques to monitor land cover dynamics in mine environment | |
| CN108492288B (en) | Random forest based multi-scale layered sampling high-resolution satellite image change detection method | |
| CN106529472A (en) | Target detection method and apparatus based on large-scale high-resolution and high-spectral image | |
| Laupheimer et al. | The importance of radiometric feature quality for semantic mesh segmentation | |
| CN114612315B (en) | High-resolution image missing region reconstruction method based on multitask learning | |
| CN102930269B (en) | Method for modulating multiscale semanteme of remote-sensing image | |
| Lizarazo et al. | Fuzzy image segmentation for urban land-cover classification | |
| Zhang et al. | An optimal approach for crack extraction from UAV sub-images after cutting | |
| CN114463379A (en) | Dynamic capturing method and device for video key points | |
| Samadzadegan | Data integration related to sensors, data and models | |
| Zulkifle et al. | Integrated NIR-HE based SPOT-5 image enhancement method for features preservation and edge detection | |
| Wang et al. | Hybrid textons: modeling surfaces with reflectance and geometry | |
| Esch et al. | Robust approach towards an automated detection of built-up areas from high resolution radar imagery | |
| Li et al. | A MTMCMC-Based Dark Spot Detection and Feature Extraction Under the Energy Function Framework |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150513 Termination date: 20190917 |