CN103984947A - High-resolution remote sensing image house extraction method based on morphological house indexes - Google Patents
High-resolution remote sensing image house extraction method based on morphological house indexes Download PDFInfo
- Publication number
- CN103984947A CN103984947A CN201410238937.2A CN201410238937A CN103984947A CN 103984947 A CN103984947 A CN 103984947A CN 201410238937 A CN201410238937 A CN 201410238937A CN 103984947 A CN103984947 A CN 103984947A
- Authority
- CN
- China
- Prior art keywords
- house
- index
- morphology
- pixel
- remote sensing
- 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.)
- Pending
Links
Landscapes
- Image Processing (AREA)
Abstract
The invention discloses a high-resolution remote sensing image house extraction method based on morphological house indexes. The morphological house indexes are constructed on the basis of the morphological algorithm according to the characteristics of high brightness, isotropy and similar rectangle degree of a house on a high-resolution remote sensing image, and a remote sensing image house is automatically extracted according to the morphological house indexes. On this basis, morphological shadow indexes are derived from the morphological house indexes according to similar spatial characteristics and opposite optical characteristics of a shadow and the house, house extraction is restrained through the morphological shadow indexes, and consequently the accuracy degree of house extraction is further optimized. According to the high-resolution remote sensing image house extraction method based on the morphological house indexes, manual segmentation and artificial training are not needed, full-automatic extraction of the remote sensing image house can be achieved; the morphological shadow indexes are added for restraint, the accuracy degree and the similar rectangle degree of house extraction can be obviously improved.
Description
Technical field
The invention belongs to Remote Sensing Image Processing Technology field, relate in particular to a kind of high-resolution remote sensing image house extracting method based on morphology house index.
Background technology
The automatic extraction of buildings is mainly essential characteristic and the priori based on image, yet the disappearance of information while expressing three-dimensional scenic due to bidimensional image, makes full-automatic extractive technique immature, is still in the research and probe stage.
At present, building construction extracting method mainly comprises: (1) extracting method based on Region Segmentation, (2) extracting method based on multisource data fusion and (3) are based on OO extracting method.
(1) extracting method based on Region Segmentation
[1]
The method is exactly that associated pixel on space is gathered into the object of homogeneous, and guarantees that object is homogeneous on spectrum and space.First piece image is divided into many zonules, these prime areas may be very little connected region or single pixels; In each region, with one, can reflect that certain Properties of Objects distinguishes pixel in different objects; Then the All Ranges of adjacent boundary is calculated, determine affiliated area and merge, such iterative process merges the set of pixels with similarity to form a region, and the pixel region of this similar quality constantly increases.Region growing is a kind of typical serial region segmentation method, and the processing of its subsequent step will be according to determining after the result of preceding step is judged.Liu etc.
[2]image is carried out to multi-scale division, considered the information such as texture, context, after cutting apart, utilize fuzzy decision Tree Classifier to classify, then by random Hough transformation, realize the extraction at edge, house.Fua and Hanso
[3]image is divided into a series of regions, according to region, obtains closed edge, finally utilize geometrical rule that edge is connected to complicated structure.
(2) extracting method based on multisource data fusion
[4~7]
Data fusion is intended to make full use of the information resources in different time and space, adopt computer technology under certain criterion, to carry out analysis and synthesis to the multisensor observation information obtaining chronologically, acquisition is explained the consistance of measurand and is described, to complete required decision-making and estimation task, make system obtain the performance more superior than its each ingredient.Multisource data fusion is that the different information that a plurality of sensors are obtained merge.Modal multisource data fusion is to merge the data with elevation information, wherein take LIDAR data as representative.Cheng Liang, Gong Jianya
[8]proposing a kind of LIDAR utilizes ultrahigh resolution image to carry out contour extraction method under auxiliary.Zhang keqi, Yan jianhua etc.
[9]the information providing by LIDAR data, has realized the automatic extraction of house point set by series of algorithms.First the method is left ground and non-ground distributor by a morphologic filtering, then by the region growing algorithm based on plane fitting technology, non-ground house is identified, last fillet point is again cut apart and is obtained house extraction result the point set in house.
(3) based on OO extracting method
[10~11]
OO basic thought is the elementary cell using imaged object as image analysing computer, according to space or spectral signature, Remote Sensing Image Segmentation is become to discrete region or set, and imaged object just refers to that these cut apart some " homogeneity " pixel set of rear generation.OO method is by image is cut apart, and extracts homogeneous region, thereby then regional is carried out to signature analysis, extracts building construction.Compare with the method based on pixel, OO method can produce higher information extraction precision conventionally.Baatz and Schpe
[12]proposition, by multi-scale division and decision tree classification combination, is used for spectrum, texture, contextual information combination the identification of target.
Traditional house extracting method can not be obtained degree of precision in actual classification, but also need to add many manual interventions, as sample training or yardstick, cuts apart etc., and automaticity is not high.
Along with the development of mathematical morphology, it has more and more significantly advantage at image processing method face.Maurya, R etc.
[13]proposed to use vegetation index NDVI and morphological operator based on cutting apart to carry out house extraction to high-resolution remote sensing image.The method has been considered spectral signature and the space characteristics in house in remote sensing image.Spectral signature is the NDVI based on cutting apart, and space characteristics is relevant to morphological operator.By NDVI, by Region Segmentation, be two parts, wherein a part is for having comprised the similar objects of spectral signature such as house and road, then uses morphological operator with space characteristics, house is separated with road.This method has obtained extraordinary effect.Pertinent literature:
House shade
[14]as the constraint on a kind of spectrum, be used to improve the precision that house extracts.The shade extracting from difference morphology can be for determining that position and the shape in house provides reliable contextual information.Shade and solar azimuth, sun altitude, sensor orientation angle are relevant.
The list of references relating in literary composition:
[1] Song Xiaoyu, single newly-built. the Preliminary Applications [J] of high-resolution satellite image in City Building identification. sensor information, 2000, (1): 26-30.
[2]Liu,Z.J.,Wang,J.,Liu,W.P.,2005.Building extraction from high resolution imagery based on multi-scale object oriented classification and probabilistic Hough transform.In:Proc.IGARSS2005Symposium,Seoul,Korea,25–29July2005.pp.2250–2253.
[3]P.Fua and A.J.Hanson.Resegmentation Using generic shape:locating general cultural object,Technical Report,Artificial Intelligence Center,SRI international.May1986.
[4]Cartographica Sinica,2008,37(3):391-393.
[5]T.Schenk,B.Csatho.Fusion of LiDAR data and aerial imagery for a more complete surface description International Archives of Photogrammetry,Remote Sensing and Spatial Information Sciences,34(Part3)(2002),pp.310–317;
[6]Wang,Z.and Schenk,T.2000.Building extraction and reconstruction from lidar data.In:International Archives of Photogrammetry and Remote Sensing,33(B3):958-964.
[7]CHENG Liang,GONG Jianya.Building Boundary Extraction Using Very High Resolution Images and LiDAR.[J].Acta Geodaetica et
[8] Cheng Liang, utilizes ultrahigh resolution Extraction of Image contour of building method [J] under the refined .LiDAR of Gong Jian is auxiliary. mapping journal, 2008,37 (3): 391-393.
[9]Zhang K.,Yan J.,Chen S.Automatic Construction of Building Footprints From Airborne LIDAR Data.IEEE Transactions on Geoscience and Remote Sensing,Vol.44,pp.2523-2533,2006;
[10] Qiao Cheng, Luo Jiancheng, Wu Quanyuan etc. OO high resolution image City Building extracts [J]. geography and Geographical Information Sciences, 2008,24 (5): 36-39;
[11] Liu Zhengjun, Zhang Jixian, Meng Yabin etc. based on the comprehensive high resolution image buildings Study on Extraction Method [J] of classification and form. mapping science, 2007,32 (3): 38-46
[12]Baatz M,Schpe A.Object-Oriented Multi-Scale Image Analysis in Semantic Networks[C].In:Proceeding of the2nd International Symposium on perationalization of Remote Sensing.Enschede,IT(C).1999.16-20;
[13]D.Singh,R.Maurya,A.S.Shukla,M.K.Sharma and P.R Gupta,"Building Extraction from Very High Resolution Multispectral Images using NDVI based Segmentation and Morphological Operators,"Engineering and Systems(SCES),2012.
[14]X.Jin and C.H.Davis,“Automated building extraction from high resolution satellite imagery in urban areas using structural,contextual,and spectral information,”EURASIP J.Appl.Signal Process.,vol.14,pp.2196–2206,Jan.2005.
Summary of the invention
The deficiency existing for prior art, the present invention considers house feature on high-resolution remote sensing image, and a kind of high-resolution remote sensing image house extracting method based on morphology house index is provided.
For solving the problems of the technologies described above, the present invention adopts following technical scheme:
A high-resolution remote sensing image house extracting method based on morphology house index, comprises step:
Step 1, obtains the brightness image of remote sensing image;
Step 2, arranges a series of linear structure element yardsticks according to house size in remote sensing image and remote sensing image resolution;
Step 3, under each linear structure element yardstick, carry out respectively following operation:
With linear structure element, brightness image is carried out to the cap conversion of the white top of different directions, obtain the white top cap value of all directions, institute's directive white top cap value is averaging, obtain multi-direction white top cap average;
Step 4, asks the absolute value of the difference of the multi-direction white top cap average under all adjacent linear structural element yardsticks, obtains the white top cap value based on difference morphology attribute, and the mean value of the white top cap value based on difference morphology attribute is morphology house index;
Step 5, extracts remote sensing image house based on morphology house index, that is, the pixel that morphology house index in remote sensing image is greater than to default house index threshold is judged to be house pixel.
The brightness image that obtains remote sensing image described in step 1 is specially:
Obtain the maximal value of each pixel spectral value in visible light wave range on remote sensing image as the brightness value of this pixel, thereby obtain brightness image.
As preferably, the white top cap described in step 3 is transformed to the white top cap conversion based on rebuilding.
Different directions described in step 3 can be expressed as
d represents direction, and n represents direction number, k=1, and 2,3 ..., n.
As preferably, the present invention also comprises that take house geometric index carries out aftertreatment as the house object that constraint obtains step 5, is specially:
The house object that house geometric index is greater than to house geometric index threshold value is left house, and the house object that house geometric index is not more than to house geometric index threshold value is judged to be non-house object;
Described house geometric index=adjustment factor * rectangular degree/length breadth ratio, rectangular degree is a rectangle equating with house object area, length breadth ratio refers to the length breadth ratio of the rectangle that equates with house object area; Adjustment factor is used for regulating house geometric index size.
As preferably, the present invention also comprises that take area carries out aftertreatment as the house object that constraint obtains step 5, is specially:
The house object that area is greater than to preset area threshold value is left house, and the house object that area is not more than to preset area threshold value is judged to be noise.
As preferably, the present invention also comprises step:
Step 6, under each linear structure element yardstick, carry out respectively following operation:
With linear structure element, brightness image is carried out to the cap conversion of the black top of different directions, obtain the black top cap value of all directions, institute's directive black top cap value is averaging, obtain multi-direction black top cap average;
Step 7, asks the absolute value of the difference of the multi-direction black top cap average under all adjacent linear structural element yardsticks, obtains the black top cap value based on difference morphology attribute, and the mean value of the black top cap value based on difference morphology attribute is morphology shadow index;
Step 8, the morphology shadow index pixel in the house object that constraint obtains step 5 of take is further judged:
8.1 rule of thumb set shadow index threshold value, and the pixel that morphology shadow index is greater than to shadow index threshold value is judged to be shade pixel, and the pixel that morphology shadow index is not more than shadow index threshold value is judged to be non-shade pixel, and shade pixel forms shadow object;
8.2 according to actual conditions default the first morphology house index threshold T_high, the second morphology house index threshold T_low, the first distance threshold D_high and second distance threshold value D_low, T_high>T_low, T_high>T_low;
If the distance that the morphology house index of pixel is greater than between T_high and pixel place house object and adjacent shadow object is less than default the first distance threshold D_high, this pixel is judged to be house pixel; If the morphology house index of pixel is less than default second distance threshold value D_low in the distance between the second morphology house index threshold T_low and the first morphology house index threshold T_high and between pixel place house object and adjacent shade pixel, this pixel is also judged to be house pixel; Other pixels are judged to be non-house pixel.
As preferably, the black top cap described in step 6 is transformed to the black top cap conversion based on rebuilding.
The brightness that the present invention is directed to house on high-resolution remote sensing image is large, the feature of isotropy, class square degree, based on morphology operations, builds morphology house index, adopts morphology house index automatically to extract remote sensing image house.On this basis, the present invention also utilizes the shade spatial character similar with house and contrary optical characteristics, from morphology house index, derive morphology shadow index, and adopt morphology shadow index that house is extracted and retrained, thereby further optimize house extraction accuracy.
Compared to the prior art, the present invention has following beneficial effect:
1,, without manually cutting apart, without artificial training, can realize the full-automation in remote sensing image house and extract;
2, add morphology shadow index to retrain, can significantly improve house extraction accuracy.
Accompanying drawing explanation
Fig. 1 is morphology house index construction process flow diagram;
Fig. 2 is that morphology shadow index builds process flow diagram.
Embodiment
On high-resolution remote sensing image, house spy has following basic characteristics:
(1) contrast and brightness:
Roof directly receives solar light irradiation, generally has higher brightness; Meanwhile, consider the house shade that sunlight projection causes, thus roof and contrast around also higher.
(2) shape:
On remote sensing image, house is usually expressed as the structure of certain space scope, similar rectangle.
(3) size:
House has a plurality of yardsticks, but size is generally in a certain range scale.
(4) direction:
With respect to road, house has isotropy feature, so usable direction is distinguished house and road.
The present invention builds morphology house index according to above house basic characteristics, thereby realizes the extraction in house in remote sensing image, and detailed process is as follows:
Step 1, obtains the brightness image of remote sensing image.
Obtain the maximal value of each pixel spectral value in visible light wave range on remote sensing image, as the brightness value of this pixel, thereby obtain brightness image:
In formula (1), M
k(x) be the spectral value of pixel x in wave band k, b (x) represents the brightness value of pixel x, and K is total wave band number.
Step 2, the white top cap conversion of brightness image.
This step further comprises sub-step:
2.1 adopt linear structure element s to carry out opening operation to brightness image b, and first corrosion is expanded again:
γ
s(b)=δ
s(ε
s(b)) (2)
In formula (2), ε represents erosion operation, and δ represents dilation operation, and s represents linear structure element.
2.2 pairs of brightness images carry out the cap computing of white top.
The brightness image b that step 1 is obtained deducts the brightness image γ after opening operation
s(b), obtain the white top cap value of brightness image.
In order better to keep each atural object shape in remote sensing image, this step is carried out the white top cap computing based on rebuilding to brightness image, adopt linear structure element s to carry out the opening operation based on rebuilding to brightness image b, brightness image b is deducted to the brightness image after the opening operation based on rebuilding
obtain the white top cap value based on rebuilding of brightness image
The cap conversion of white top can detect the bright object that is less than or equal to linear structure length of element, and removes other darker pixels simultaneously, and white top cap value can reflect the luminance difference of the interior pixel of linear structure elemental areas and its contiguous pixel.
Step 3, the white top cap conversion of multidirectional.
Traditional form is learned the general disc-shaped structure element that adopts in house extracting method, and disc-shaped structure element is not considered directivity, and it is all identical to all directions.House has isotropy feature with respect to road, and feature can make a distinction road and house whereby.Linear structure element can effectively be considered multi-direction, think that the directivity of describing image structure, the present invention adopt linear structure element to replace disc-shaped structure element.
The direction set D of linear structure element can be expressed as:
In formula (3), n represents direction number.
The white top cap conversion of this step multidirectional is specially:
3.1 at direction set
under middle all directions, adopt linear structure element respectively brightness image b to be carried out to the cap conversion of white top, obtain the white top of the brightness image cap value THR of all directions
s.dir(b), dir represents the direction of linear structure element.
The 3.2 couples of white top of directive brightness image cap value THR
s.dir(b) be averaging and obtain multi-direction white top cap average
Because house has isotropy feature, therefore in all directions, all there is larger top cap value, its eigenwert is larger than all the other objects.
Step 4, multiple dimensioned white top cap conversion.
On remote sensing image, house is of different sizes size.Therefore, this step is carried out multiple dimensioned white top cap conversion to brightness image, obtains the white top of the brightness image cap value on each yardstick.
Multiple dimensioned top cap conversion is set up based on difference morphology attribute, and operational method is as follows:
In formula (5),
the white top cap value based on difference morphology attribute that represents different linear structure element yardsticks; s
minrepresent minimum linear structure element, s
maxrepresent maximum linear structure element, by visual, choose house in remote sensing image, by measuring house yardstick in remote sensing image, determine s
minand s
max, in remote sensing image, minimum house yardstick is defined as s
min, in remote sensing image, maximum house yardstick is defined as s
max, adopt default yardstick interval delta s to divide Scaling interval [s
min, s
max], obtain a series of linear structure element yardsticks, through the white top cap conversion based on difference morphology attribute, obtain the white top cap value based on difference morphology attribute that each yardstick is corresponding, yardstick interval delta s is according to the empirical value according to house size and remote sensing image resolution arrange in remote sensing image.
White top cap value based on difference morphology attribute
multi-direction white top cap average for a rear linear structure element yardstick
multi-direction white top cap average with last linear structure element yardstick
the absolute value of difference.
Step 5, builds morphology house index.
In step 1~4, brightness, contrast, directivity, the scale feature for house processed respectively, based on above result definition morphology house index M BI, is about to the white top cap value THR based on difference morphology attribute of each yardstick of all directions
dMPbe averaging:
In formula (6), MBI is morphology house index.MBI value is larger, and the possibility that its corresponding pixel is house is also just larger.
Step 6, builds morphology shadow index.
White top cap conversion all in step 2~4 is replaced with to the cap conversion of black top, obtain the black top cap value B-THR based on difference morphology attribute of each yardstick of linear structure element
dMP, build morphology shadow index:
Wherein, MSI is morphology shadow index, B-THR
dMPthe black top cap value based on difference morphology attribute for each yardstick of all directions.
The cap conversion of black top is that brightness image b is deducted to the brightness image after closed operation.Adopt the cap conversion of black top can show the dark colored structures in remote sensing image, detect dash area.Same morphology shadow index is larger, and the possibility that its respective pixel is shade is also larger.
Step 7, the remote sensing image house based on morphology house exponential sum morphology shadow index extracts.
Based on morphology house index binaryzation remote sensing image, obtain house object, take house geometric index and/or floor space carries out aftertreatment as constraint condition to house object, further removes noise and false-alarm, to optimize remote sensing image house, extracts result.
Embodiment based on morphology house index binaryzation remote sensing image is:
Rule of thumb set house index threshold, the pixel that morphology house index is greater than house index threshold is judged to be house pixel, and the pixel that is less than this empirical value is judged to be non-house pixel, and all house pixels form house object.
Area refers to the pixel count that partial structurtes body or object comprise, the embodiment that the floor space of take is constraint condition as:
Manually choose house object, and obtain its area, the house object that area is greater than preset area threshold value is left house, and the house object that area is not more than preset area threshold value is judged to be noise.
House geometric index is defined as:
House geometric index=adjustment factor * rectangular degree/length breadth ratio (8)
In formula (8), rectangular degree is defined as a rectangle equating with house object area; Length breadth ratio refers to the length breadth ratio of the rectangle that equates with house object area; Adjustment factor is used for regulating house geometric index size, is convenient to setting threshold, and in this concrete enforcement, setting adjustment factor is 10.
The embodiment that the house geometric index of take is constraint condition as:
The house object that house geometric index is greater than to house geometric index threshold value is still left house object, and the house object that house geometric index is not more than house geometric index threshold value is the non-house object such as road.
In this step, adopt house geometric index constraint to make a return journey to slide down the larger and object in irregular shape of the length and width such as road, adopt the area-constrained tiny noise of removing, thus raising house extraction accuracy.
For further improving house extraction accuracy, the present invention adopts morphology shadow index to retrain the house object obtaining based on morphology house index, so that pixel in house object is further judged, is specifically implemented as follows:
Rule of thumb set shadow index threshold value, the pixel that morphology shadow index is greater than shadow index threshold value is judged to be shade pixel, and the pixel that morphology shadow index is not more than shadow index threshold value is judged to be non-shade pixel, and shade pixel forms shadow object.
According to default the first morphology house index threshold T_high of actual conditions and the second morphology house index threshold T_low, T_high>T_low.If the distance that the morphology house index of pixel is greater than between T_high and pixel place house object and adjacent shadow object is less than default the first distance threshold D_high, this pixel is judged to be house pixel; If the morphology house index of pixel is less than default second distance threshold value D_low in the distance between the second morphology house index threshold T_low and the first morphology house index threshold T_high and between pixel place house object and adjacent shade pixel, this pixel is judged to be house pixel, T_high>T_low, other pixels are judged to be non-house pixel.Above-mentioned all threshold values are empirical value.
Claims (8)
1. the high-resolution remote sensing image house extracting method based on morphology house index, is characterized in that:
Step 1, obtains the brightness image of remote sensing image;
Step 2, arranges a series of linear structure element yardsticks according to house size in remote sensing image and remote sensing image resolution;
Step 3, under each linear structure element yardstick, carry out respectively following operation:
With linear structure element, brightness image is carried out to the cap conversion of the white top of different directions, obtain the white top cap value of all directions, institute's directive white top cap value is averaging, obtain multi-direction white top cap average;
Step 4, asks the absolute value of the difference of the multi-direction white top cap average under all adjacent linear structural element yardsticks, obtains the white top cap value based on difference morphology attribute, and the mean value of the white top cap value based on difference morphology attribute is morphology house index;
Step 5, extracts remote sensing image house based on morphology house index, that is, the pixel that morphology house index in remote sensing image is greater than to default house index threshold is judged to be house pixel.
2. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 1, is characterized in that:
The brightness image that obtains remote sensing image described in step 1 is specially:
Obtain the maximal value of each pixel spectral value in visible light wave range on remote sensing image as the brightness value of this pixel, thereby obtain brightness image.
3. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 1, is characterized in that:
White top cap described in step 3 is transformed to the white top cap conversion based on rebuilding.
4. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 1, is characterized in that:
Different directions described in step 3 can be expressed as
d represents direction, and n represents direction number, k=1, and 2,3 ..., n.
5. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 1, is characterized in that:
Also comprise that take house geometric index carries out aftertreatment as the house object that constraint obtains step 5, is specially:
The house object that house geometric index is greater than to house geometric index threshold value is left house, and the house object that house geometric index is not more than to house geometric index threshold value is judged to be non-house object;
Described house geometric index=adjustment factor * rectangular degree/length breadth ratio, rectangular degree is a rectangle equating with house object area, length breadth ratio refers to the length breadth ratio of the rectangle that equates with house object area; Adjustment factor is used for regulating house geometric index size.
6. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 1, is characterized in that:
Also comprise that take area carries out aftertreatment as the house object that constraint obtains step 5, is specially:
The house object that area is greater than to preset area threshold value is left house, and the house object that area is not more than to preset area threshold value is judged to be noise.
7. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 1, is characterized in that:
Also comprise step:
Step 6, under each linear structure element yardstick, carry out respectively following operation:
With linear structure element, brightness image is carried out to the cap conversion of the black top of different directions, obtain the black top cap value of all directions, institute's directive black top cap value is averaging, obtain multi-direction black top cap average;
Step 7, asks the absolute value of the difference of the multi-direction black top cap average under all adjacent linear structural element yardsticks, obtains the black top cap value based on difference morphology attribute, and the mean value of the black top cap value based on difference morphology attribute is morphology shadow index;
Step 8, the morphology shadow index pixel in the house object that constraint obtains step 5 of take is further judged:
8.1 rule of thumb set shadow index threshold value, and the pixel that morphology shadow index is greater than to shadow index threshold value is judged to be shade pixel, and the pixel that morphology shadow index is not more than shadow index threshold value is judged to be non-shade pixel, and shade pixel forms shadow object;
8.2 according to actual conditions default the first morphology house index threshold T_high, the second morphology house index threshold T_low, the first distance threshold D_high and second distance threshold value D_low, T_high>T_low, T_high>T_low;
If the distance that the morphology house index of pixel is greater than between T_high and pixel place house object and adjacent shadow object is less than default the first distance threshold D_high, this pixel is judged to be house pixel; If the morphology house index of pixel is less than default second distance threshold value D_low in the distance between the second morphology house index threshold T_low and the first morphology house index threshold T_high and between pixel place house object and adjacent shade pixel, this pixel is also judged to be house pixel; Other pixels are judged to be non-house pixel.
8. the high-resolution remote sensing image house extracting method based on morphology house index as claimed in claim 7, is characterized in that:
Black top cap described in step 6 is transformed to the black top cap conversion based on rebuilding.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410238937.2A CN103984947A (en) | 2014-05-30 | 2014-05-30 | High-resolution remote sensing image house extraction method based on morphological house indexes |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410238937.2A CN103984947A (en) | 2014-05-30 | 2014-05-30 | High-resolution remote sensing image house extraction method based on morphological house indexes |
Publications (1)
Publication Number | Publication Date |
---|---|
CN103984947A true CN103984947A (en) | 2014-08-13 |
Family
ID=51276909
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410238937.2A Pending CN103984947A (en) | 2014-05-30 | 2014-05-30 | High-resolution remote sensing image house extraction method based on morphological house indexes |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103984947A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105608458A (en) * | 2015-10-20 | 2016-05-25 | 武汉大学 | High-resolution remote sensing image building extraction method |
CN105654495A (en) * | 2016-01-07 | 2016-06-08 | 太原工业学院 | Method for detecting internal defect of automobile brake disc |
WO2016106960A1 (en) * | 2014-12-30 | 2016-07-07 | 华中科技大学 | Denoising method for restraining spectrum characteristic of remote sensing image of ground building group |
CN107516100A (en) * | 2017-08-31 | 2017-12-26 | 北京航天绘景科技有限公司 | A kind of image building extracting method based on elevation morphology building index |
CN107657246A (en) * | 2017-10-19 | 2018-02-02 | 武汉大学 | A kind of remote sensing image building detection method based on multi-scale filtering building index |
CN109472785A (en) * | 2018-11-01 | 2019-03-15 | 中国石油大学(华东) | Based on Multi-structure elements earthquake damaged structure object method for quickly identifying |
CN110866493A (en) * | 2019-11-14 | 2020-03-06 | 生态环境部南京环境科学研究所 | Remote sensing image-based large-scale live pig farm identification method |
CN111144335A (en) * | 2019-12-30 | 2020-05-12 | 自然资源部国土卫星遥感应用中心 | Method and device for building deep learning model |
CN115909050A (en) * | 2022-10-26 | 2023-04-04 | 中国电子科技集团公司第五十四研究所 | Remote sensing image airport extraction method combining line segment direction and morphological difference |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1916935A (en) * | 2006-09-11 | 2007-02-21 | 武汉大学 | Hybrid sorting process of adjustable spectrum and space characteristics |
CN1932850A (en) * | 2006-10-09 | 2007-03-21 | 武汉大学 | Remoto sensing image space shape characteristics extracting and sorting method |
US7340099B2 (en) * | 2003-01-17 | 2008-03-04 | University Of New Brunswick | System and method for image fusion |
-
2014
- 2014-05-30 CN CN201410238937.2A patent/CN103984947A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7340099B2 (en) * | 2003-01-17 | 2008-03-04 | University Of New Brunswick | System and method for image fusion |
CN1916935A (en) * | 2006-09-11 | 2007-02-21 | 武汉大学 | Hybrid sorting process of adjustable spectrum and space characteristics |
CN1932850A (en) * | 2006-10-09 | 2007-03-21 | 武汉大学 | Remoto sensing image space shape characteristics extracting and sorting method |
Non-Patent Citations (2)
Title |
---|
XIN HUANG等: ""Morphological Building/Shadow Index for Building Extraction From High-Resolution Imagery Over Urban Areas"", 《IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING》 * |
黄昕: ""高分辨率遥感影像多尺度纹理、形状特征提取与面向对象分类研究"", 《中国博士学位论文全文数据库 信息科技辑 》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2016106960A1 (en) * | 2014-12-30 | 2016-07-07 | 华中科技大学 | Denoising method for restraining spectrum characteristic of remote sensing image of ground building group |
CN105608458B (en) * | 2015-10-20 | 2019-01-18 | 武汉大学 | A kind of high-resolution remote sensing image building extracting method |
CN105608458A (en) * | 2015-10-20 | 2016-05-25 | 武汉大学 | High-resolution remote sensing image building extraction method |
CN105654495A (en) * | 2016-01-07 | 2016-06-08 | 太原工业学院 | Method for detecting internal defect of automobile brake disc |
CN107516100A (en) * | 2017-08-31 | 2017-12-26 | 北京航天绘景科技有限公司 | A kind of image building extracting method based on elevation morphology building index |
CN107657246B (en) * | 2017-10-19 | 2019-12-10 | 武汉大学 | Remote sensing image building detection method based on multi-scale filtering building index |
CN107657246A (en) * | 2017-10-19 | 2018-02-02 | 武汉大学 | A kind of remote sensing image building detection method based on multi-scale filtering building index |
CN109472785A (en) * | 2018-11-01 | 2019-03-15 | 中国石油大学(华东) | Based on Multi-structure elements earthquake damaged structure object method for quickly identifying |
CN109472785B (en) * | 2018-11-01 | 2019-12-20 | 中国石油大学(华东) | Quick identification method for damaged building caused by earthquake damage based on multiple structural elements |
CN110866493A (en) * | 2019-11-14 | 2020-03-06 | 生态环境部南京环境科学研究所 | Remote sensing image-based large-scale live pig farm identification method |
CN111144335A (en) * | 2019-12-30 | 2020-05-12 | 自然资源部国土卫星遥感应用中心 | Method and device for building deep learning model |
CN115909050A (en) * | 2022-10-26 | 2023-04-04 | 中国电子科技集团公司第五十四研究所 | Remote sensing image airport extraction method combining line segment direction and morphological difference |
CN115909050B (en) * | 2022-10-26 | 2023-06-23 | 中国电子科技集团公司第五十四研究所 | Remote sensing image airport extraction method combining line segment direction and morphological difference |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103984947A (en) | High-resolution remote sensing image house extraction method based on morphological house indexes | |
Bakhtiari et al. | Semi automatic road extraction from digital images | |
Willmes et al. | Pan-Arctic lead detection from MODIS thermal infrared imagery | |
Ghaffarian | Automatic building detection based on supervised classification using high resolution Google Earth images | |
Hormese et al. | Automated road extraction from high resolution satellite images | |
Miao et al. | Object-based detection of Arctic sea ice and melt ponds using high spatial resolution aerial photographs | |
Trinder et al. | Aerial images and LiDAR data fusion for disaster change detection | |
CN103839267A (en) | Building extracting method based on morphological building indexes | |
US11804025B2 (en) | Methods and systems for identifying topographic features | |
Demir et al. | Integrated shoreline extraction approach with use of Rasat MS and SENTINEL-1A SAR Images | |
Chen et al. | Automatic building extraction via adaptive iterative segmentation with LiDAR data and high spatial resolution imagery fusion | |
Im et al. | Optimum Scale in Object‐Based Image Analysis | |
Ichim et al. | Road detection and segmentation from aerial images using a CNN based system | |
Rashidi et al. | Ground filtering LiDAR data based on multi-scale analysis of height difference threshold | |
Rashidi et al. | Extraction of ground points from LiDAR data based on slope and progressive window thresholding (SPWT) | |
Manandhar et al. | Segmentation based building detection in high resolution satellite images | |
Jamali et al. | Automated extraction of buildings from aerial LiDAR point clouds and digital imaging datasets | |
CN115294183A (en) | Disc-shaped sub-lake water body time sequence extraction method based on multi-source remote sensing data | |
CN114792322A (en) | Method for detecting cloud and cloud shadow of mountain domestic high-resolution satellite image | |
Večeř et al. | Segmentation and Object-Based Land Cover Classification of Airborne Images in Kraliky County | |
CN109359583B (en) | Method for extracting significant planar specific target in remote sensing image | |
CN114140698A (en) | Water system information extraction algorithm based on FasterR-CNN | |
Zhao et al. | Building extraction from lidar point cloud data using marked point process | |
Ye et al. | Research on Automatic Highway Extraction Technology Based on Spectral Information of Remote Sensing Images. | |
Mohammadzadeh et al. | A self-organizing fuzzy segmentation (SOFS) method for road detection from high resolution satellite images |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20140813 |