CN108154138B - A method of the automatic identification taxus chinensis in northeast in high-resolution remote sensing image - Google Patents
A method of the automatic identification taxus chinensis in northeast in high-resolution remote sensing image Download PDFInfo
- Publication number
- CN108154138B CN108154138B CN201810051779.8A CN201810051779A CN108154138B CN 108154138 B CN108154138 B CN 108154138B CN 201810051779 A CN201810051779 A CN 201810051779A CN 108154138 B CN108154138 B CN 108154138B
- Authority
- CN
- China
- Prior art keywords
- pixel
- image
- northeast
- value
- taxus chinensis
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of method identifying taxus chinensis in northeast in high-resolution remote sensing image, only need a small amount of ground sample that can realize the identification taxus chinensis in northeast in high-resolution remote sensing image degree of precision, the advantage for making full use of remote sensing image low cost area coverage larger, the characteristic feature that taxus chinensis in northeast can be directed to is identified from remote sensing image taxus chinensis in northeast based on less sample, and preferable identification quality is obtained, there is higher social value and economic value.
Description
Technical field
The invention belongs to remote sensing images analysis and processing technology field, and in particular to one kind is in high-resolution remote sensing image
The method of automatic identification taxus chinensis in northeast.
Background technology
Taxus chinensis in northeast is a kind of species of the endangered plants of first-grade state protection.It is that the precious of Tertiary Period few survivors is set at it
Kind, have 2,500,000 years history so far, there is higher researching value;A variety of anticancers can be extracted from taxus chinensis in northeast
Substance, most to there is one of the anticancer drug of exploitation future in the world.Such seeds majority is distributed in Jilin Province master ridge, Zhang Guangcai
Ridge and Changbaishan area position and identify the position of these trees, help to protect and make full use of this endemic species, have
The very high ecological value, social value, economic value.
In current application and research, taxus chinensis in northeast is found in an area, by two kinds by the way of:First, passing through
The seeds are found in on-the-spot investigation, and since taxus chinensis in northeast is distributed in 500-1000 meters of mountain areas of height above sea level, such area is vast and most
It is difficult to arrive at, by on-the-spot investigation mode, not only cost is higher and less efficient;Second is that using supervised classification algorithm in high-resolution
Rate remote sensing image identifies that taxus chinensis in northeast, the content of the remote sensing image of satellite shooting can cover larger area, pass through
The supervised classifications algorithms such as SVM, convolutional Neural net can be to the knowledge of fast and low-cost on the basis of learning large number of ground target sample
Do not go out the target in remote sensing image, however since taxus chinensis in northeast is less and often mixes life with other conifer trees, so on the one hand
Carry out training algorithm it is difficult to obtain and meet the taxus chinensis in northeast ground sample of quantity, on the other hand since Parameter identification algorithm carries
The characteristic of tolerable scale or color change causes to supervise so more difficult distinguish taxus chinensis in northeast and other bushes
Sorting algorithm is superintended and directed to be difficult to play a role.
Invention content:
In view of the problems of the existing technology, the present invention provides one kind and identifying northeast red bean in high-resolution remote sensing image
The method of China fir only needs a small amount of ground sample that can realize the knowledge in high-resolution remote sensing image degree of precision by this method
Other taxus chinensis in northeast.
A kind of method identifying taxus chinensis in northeast in high-resolution remote sensing image of the present invention, including following step
Suddenly:
S1, input high-resolution remote sensing image is (referred to as:Image), the resolution ratio (abbreviation of remote sensing image:
Resolution), the scale Size of image blocks to be analyzed is calculated, it includes taxus chinensis in northeast that one is manually chosen on Image
Position (row coordinate HHD, row coordinate LHD), the position (H of non-aciculignosaFZY, LFZY), one is confierophyte but is not east
Position (the H of Taxus cuspidataFHD, LFHD):The lowest resolution of high-resolution remote sensing image Image is that each pixel corresponds on image
20 cm section of ground;
Image is made of pixel, including R red bands, G green bands, B blue wave bands;Image is for Image
In a pixel Pixel include with properties:
The value of Pixel.R red bands;
The value of Pixel.G green bands;
The value of Pixel.B blue wave bands;
Pixel.Vfeature vegetation characteristics values, default value 0;
Pixel.ZYSelected needles select variable, default value 0;
Pixel.Zfeature needle characteristic values, default value 0;
Resolution be unit be centimetre remote sensing image resolution ratio;
The calculation formula of the scale Size of image blocks to be analyzed is as follows:
Wherein, round is the function to round up to the numerical value of input;Suguan is tree crown breadth index, value
Ranging from 4-8, default value 4.5;
HHD、HFZY、HFHDFor position coordinate line number residing on high resolution image;
LHD、LFZY、LFHDFor position coordinate columns residing on high resolution image;
S2 calculates its vegetation characteristics value for all pixels in Image according to the scale Size of image blocks to be analyzed
(Vfeature):
S201 takes out a pixel Pixel in Image;
S202, with the positions Pixel (HPixel, LPixel) centered on, it is 2 × size+1 that width is taken out in Image
Pixel, highly the image blocks Patch for 2 × size+1 pixel;
Include N number of pixel, wherein N=(2 × size+1) × (2 × size+1), these pixels difference in image blocks Patch
Labeled as P (1), P (2) ..., P (N);
S203 calculates Patch vegetation characteristics values Vfeature using following formula:
Wherein, P (i) is i-th of pixel in Patch, P (i) .R, P (i) .G, the red, green, blue wave that P (i) .B are the pixel
The value of section;Pixel.R, Pixel.G, Pixel.B are the value of the red, green, blue wave band of Pixel;Log is natural logrithm;
S204, in the storage to Pixel of Patch vegetation characteristics values;Pixel.Vfeature=Vfeature;
S205, if in Image all pixels calculated completion so go to S206, otherwise go to S201;
The processing procedure of S206, the part terminate;
S3, according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), non-needle plant
Position (the H of quiltFZY, LFZY) needles of all pixels of setting selects variable;
S301 takes out the position (H of taxus chinensis in northeastHD, LHD) corresponding pixel PHD, obtain its vegetation characteristics value V1=
PHD.Vfeature;
S302 takes out the position (H of non-aciculignosaFZY, LFZY) corresponding pixel PFZY, obtain its vegetation characteristics value V2=
PFZY.Vfeature;
S303 calculates vegetation characteristics mean distance Dmean, calculation formula is as follows:
Wherein abs is the function for calculating absolute value;
S304 takes out a pixel Pixel in Image;
S305 obtains the vegetation characteristics VT=Pixel.Vfeature of the pixel;
S306 calculates vegetation characteristics range formula DZB, formula is as follows:
Wherein abs is to calculate ABS function;
S307, if DZB>Dmean, then Pixel.ZYSelected=0, otherwise Pixel.ZYSelected=1;
S308, if in Image all pixels calculated completion so go to S309, it is no to go to S304;
The processing procedure of S309, the part terminate;
S4 calculates its needle feature for all pixels in all Image according to the scale Size of image blocks to be analyzed
It is worth (Pixel.Zfeature):
S401 takes out a pixel Pixel in Image;
Otherwise S402 goes to S412 if Pixel.ZYSelected=1 so goes to S403;
S403, with the positions Pixel (HPixel, LPixel) centered on, it is 2 × size+1 that width is taken out in Image
Pixel, highly the image blocks Patch for 2 × size+1 pixel;
Include N number of pixel, wherein N=(2 × size+1) × (2 × size+1), these pixels difference in image blocks Patch
Labeled as P (1), P (2) ..., P (N);
S404, pixel counter Counter=1, summation statistical variable sum=0;
S405 takes out the Counter pixel P (counter) in Patch;
S406, take out the adjacent left side in the position P (counter), upper left, upper, upper right, the right side, bottom right, under, lower-left totally 8
Neighborhood pixel is labeled as PN [1], PN [2] ..., PN [8];
S407 calculates the heterogeneous value DS of pixel:
S408 DS is added in sum, sum=sum+DS;
S409, Counter=Counter+1;
S410, if Counter>N so goes to S411, otherwise goes to S405;
S411, sets the needle characteristic value of Pixel, and Pixel.Zfeature=tanh (sum) goes to S413;
Wherein, tanh is hyperbolic tangent function;
S412 sets Pixel.Zfeature=-1;
S413, if in Image all pixels calculated completion so go to S414, it is no to go to S401;
The processing procedure of S414, the part terminate;
S5. according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), one be needle
Leaf plant but be not taxus chinensis in northeast position (HFHD, LFHD) obtain the selection result image ResultImage:S501 establishes one
A the selection result image ResultImage identical with Image sizes, all pixels of the image are filled with black;
S502 takes out the position (H of taxus chinensis in northeastHD, LHD) corresponding pixel PHD, obtain its needle characteristic value Z1=
PHD.Zfeature;
S503, taking-up be confierophyte but be not taxus chinensis in northeast position (HFHD, LFHD)PFHD, obtain its vegetation characteristics
Value Z2=PFHD.Zfeature;
S504 calculates needle characteristic mean distance Zmean, calculation formula it is as follows:
Wherein abs is to calculate absolute value;
S505 takes out a pixel Pixel in Image;
S506 obtains the vegetation characteristics ZT=Pixel.Zfeature of the pixel;
S507, if ZT<0 so goes to S512, otherwise goes to S508;
S508 calculates needle characteristic distance formula ZJL, formula is as follows:
Wherein abs is to calculate absolute value;
S509, if ZJL>Zmean, then going to S512, otherwise go to S510;
S510 obtains the position (H of Pixelp,Lp);
S511 the, by (H of ResultImagep,Lp) position pixel labeled as white;
S512, if in Image all pixels calculated completion so go to S513, it is no to go to S505;
S513 exports the selection result image ResultImage, and black portions are non-taxus chinensis in northeast in the image, and
White portion is the position for the taxus chinensis in northeast that this method identifies.
Beneficial effects of the present invention are as follows:A kind of side identifying taxus chinensis in northeast in high-resolution remote sensing image is provided
Method, it is only necessary to which a small amount of ground sample can realize the identification taxus chinensis in northeast in high-resolution remote sensing image degree of precision, fill
Divide using the larger advantage of remote sensing image low cost area coverage, the characteristic feature of taxus chinensis in northeast can be directed to based on less
Sample identifies taxus chinensis in northeast from remote sensing image, and obtains preferable identification quality, has higher social value
And economic value.
Description of the drawings
Fig. 1 is overview flow chart of the present invention;
The corresponding flow chart of Fig. 2 steps 2 of the present invention;
The corresponding flow chart of Fig. 3 steps 3 of the present invention;
The corresponding flow chart of Fig. 4 steps 4 of the present invention;
The corresponding flow chart of Fig. 5 steps 5 of the present invention.
Specific implementation mode
By following embodiment further illustrate description the present invention, do not limit the invention in any way, without departing substantially from
Under the premise of technical solution of the invention, easy to implement any of those of ordinary skill in the art made for the present invention changes
Dynamic or change is fallen within scope of the presently claimed invention.
Embodiment 1
By taking a scape high-resolution remote sensing image in master ridge area as an example, it is 1 which, which is 2000*2000 resolution ratio,
Rice is compared by using the method that this patent describes with traditional nerve net, support vector machines, random forest method, red
The accuracy comparison of beans China fir identification is following (referring to Fig. 1~Fig. 5):
S1 inputs master ridge area high-resolution remote sensing image image, calculates the scale Size of image blocks to be analyzed,
A coordinate position (H for including taxus chinensis in northeast is chosen on ImageHD, LHD), the coordinate position (H of non-aciculignosaFZY,
LFZY), the confierophyte coordinate position (H of a non-taxus chinensis in northeastFHD, LFHD):
Wherein HHD=384, LHD=187
Wherein HFZY=1064, LHD=571
Wherein HFHD=426, LFHD=151
Size=450
S2 calculates its vegetation characteristics value for all pixels in Image according to the scale Size of image blocks to be analyzed
(Vfeature):
S3, according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), non-needle plant
Position (the H of quiltFZY, LFZY) needles of all pixels of setting selects variable;
S4 calculates its needle feature for all pixels in all Image according to the scale Size of image blocks to be analyzed
It is worth (Pixel.Zfeature):
S5. according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), one be needle
Leaf plant but be not taxus chinensis in northeast position (HFHD, LFHD) the selection result image ResultImage is obtained,
The pixel in ResultImage being white is exactly taxus chinensis in northeast.
It is as follows by using the pixel ability and existing method comparing result of the taxus chinensis in northeast of the method for the present invention label:
It can be seen that fewer pixel is marked relative to other three kinds of methods in the method for the present invention, this is manually examined on the spot
It looks into and brings larger facility.Simultaneously in marked pixel, the method for the present invention has reached very high accuracy rate, absolutely mostly
Number is confirmed to be by the pixel that present invention label is.It is less efficient compared to other methods, the labeled picture of majority
Member is not taxus chinensis in northeast.Above implementation shows that the present invention has higher application value.
Embodiment 2
By taking a scape high-resolution remote sensing image in the ridges Zhang Guangcai as an example, it is 0.5 which, which is 5000*5000 resolution ratio,
Rice (referring to Fig. 1~Fig. 5);
The high-resolution remote sensing image image in the ridge S1, input Zhang Guangcai, calculates the scale Size of image blocks to be analyzed,
A coordinate position (H for including taxus chinensis in northeast is chosen on ImageHD, LHD), the coordinate position (H of non-aciculignosaFZY,
LFZY), the confierophyte coordinate position (H of a non-taxus chinensis in northeastFHD, LFHD):
Wherein HHD=384, LHD=187;
Wherein HFZY=1064, LHD=571;
Wherein HFHD=426, LFHD=151;
Size=900;
S2 calculates its vegetation characteristics value for all pixels in Image according to the scale Size of image blocks to be analyzed
(Vfeature):
S3, according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), non-needle plant
Position (the H of quiltFZY, LFZY) needles of all pixels of setting selects variable;
S4 calculates its needle feature for all pixels in all Image according to the scale Size of image blocks to be analyzed
It is worth (Pixel.Zfeature):
S5. according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), one be needle
Leaf plant but be not taxus chinensis in northeast position (HFHD, LFHD) the selection result image ResultImage is obtained,
The pixel in ResultImage being white is exactly taxus chinensis in northeast.
For the image of this area, the method that describes by using the present invention and traditional nerve net, support vector machines, with
Machine forest method is compared, and the accuracy comparison of Chinese yew identification is as follows:
It can be seen that fewer pixel is marked relative to other three kinds of methods in the method for the present invention, this is manually examined on the spot
It looks into and brings larger facility.Simultaneously in marked pixel, the method for the present invention has reached very high accuracy rate, absolutely mostly
Number is confirmed to be by the pixel that present invention label is.It is less efficient compared to other methods, the labeled picture of majority
Member is not taxus chinensis in northeast.Above implementation shows that the present invention has higher application value.
Claims (1)
1. a kind of method identifying taxus chinensis in northeast in high-resolution remote sensing image, includes the following steps:
S1, input high-resolution remote sensing image is (referred to as:Image), the resolution ratio (abbreviation of remote sensing image:Resolution), count
The scale Size for calculating image blocks to be analyzed chooses a coordinate position (H for including taxus chinensis in northeast on ImageHD, LHD), one
Coordinate position (the H of a non-aciculignosaFZY, LFZY), the confierophyte coordinate position (H of a non-taxus chinensis in northeastFHD, LFHD):
Image is made of pixel, including 3 wave bands:R red bands, G green bands, B blue wave bands;One in Image
A pixel Pixel includes with properties:
The value of the red wave bands of Pixel.R;
The value of the green wave bands of Pixel.G;
The value of Pixel.B indigo plant wave bands;
Pixel.Vfeature vegetation characteristics values, default value 0;
Pixel.ZYSelected needles select variable, default value 0;
Pixel.Zfeature needle characteristic values, default value 0;
Resolution be unit be centimetre remote sensing image resolution ratio;
The calculation formula of the scale Size of image blocks to be analyzed is as follows:
Wherein, round is the function to round up to the numerical value of input;Suguan is tree crown breadth index, value range
For 4-8, default value 4.5;
HHD、HFZY、HFHDFor position coordinate line number residing on high resolution image;
LHD、LFZY、LFHDFor position coordinate columns residing on high resolution image;
S2 calculates its vegetation characteristics value for all pixels in Image according to the scale Size of image blocks to be analyzed
(Vfeature):
S201 takes out a pixel Pixel in Image;
S202, with the positions Pixel (HPixel, LPixel) centered on, it is 2 × size+1 pixel that width is taken out in Image,
Height is the image blocks Patch of 2 × size+1 pixel;
Include N number of pixel in image blocks Patch, wherein N=(2 × size+1) × (2 × size+1), these pixels mark respectively
For P (1), P (2) ..., P (N);
S203 calculates Patch vegetation characteristics values Vfeature using following formula:
Wherein, P (i) is i-th of pixel in Patch, P (i) .R, P (i) .G, the red, green, blue wave band that P (i) .B are the pixel
Value;Pixel.R, Pixel.G, Pixel.B are the value of the red, green, blue wave band of Pixel;Log is natural logrithm;
S204, in the storage to Pixel of Patch vegetation characteristics values:Pixel.Vfeature=Vfeature;
S205, if in Image all pixels calculated completion so go to S206, otherwise go to S201;
The processing procedure of S206, the part terminate;
S3, according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), the position of non-aciculignosa
Set (HFZY, LFZY) needles of all pixels of setting selects variable;
S301 takes out the position (H of taxus chinensis in northeastHD, LHD) corresponding pixel PHD, obtain its vegetation characteristics value V1=
PHD.Vfeature;
S302 takes out the position (H of non-aciculignosaFZY, LFZY) corresponding pixel PFZY, obtain its vegetation characteristics value V2=
PFZY.Vfeature;
S303 calculates vegetation characteristics mean distance Dmean, calculation formula is as follows:
Wherein, abs is the function for calculating absolute value;
S304 takes out a pixel Pixel in Image;
S305 obtains the vegetation characteristics VT=Pixel.Vfeature of the pixel;
S306 calculates vegetation characteristics range formula DZB, formula is as follows:
S307, if DZB>Dmean, then Pixel.ZYSelected=0, otherwise Pixel.ZYSelected=1;
S308, if in Image all pixels calculated completion so go to S309, it is no to go to S304;
The processing procedure of S309, the part terminate;
S4 calculates its needle characteristic value for all pixels in all Image according to the scale Size of image blocks to be analyzed
(Pixel.Zfeature):
S401 takes out a pixel Pixel in Image;
Otherwise S402 goes to S412 if Pixel.ZYSelected=1 so goes to S403;
S403, with the positions Pixel (HPixel, LPixel) centered on, it is 2 × size+1 pixel that width is taken out in Image,
Height is the image blocks Patch of 2 × size+1 pixel;
Include N number of pixel in image blocks Patch, wherein N=(2 × size+1) × (2 × size+1), these pixels mark respectively
For P (1), P (2) ..., P (N);
S404, pixel counter Counter=1, summation statistical variable sum=0;
S405 takes out the Counter pixel P (counter) in Patch;
S406, take out the adjacent left side in the position P (counter), upper left, upper, upper right, the right side, bottom right, under, lower-left totally 8 neighborhoods
Pixel is labeled as PN [1], PN [2] ..., PN [8];
S407 calculates the heterogeneous value DS of pixel:
S408 DS is added in sum, sum=sum+DS;
S409, Counter=Counter+1;
S410, if Counter>N so goes to S411, otherwise goes to S405;
S411, sets the needle characteristic value of Pixel, and Pixel.Zfeature=tanh (sum) goes to S413;
Wherein, tanh is hyperbolic tangent function;
S412 sets Pixel.Zfeature=-1;
S413, if in Image all pixels calculated completion so go to S414, it is no to go to S401;
The processing procedure of S414, the part terminate;
S5. according to the vegetation characteristics value of all pixels in Image, the position (H of taxus chinensis in northeastHD, LHD), one be confierophyte
But it is not the position (H of taxus chinensis in northeastFHD, LFHD) obtain the selection result image ResultImage:
S501 establishes a selection result image ResultImage identical with Image sizes, all pixel fillings of the image
For black;
S502 takes out the position (H of taxus chinensis in northeastHD, LHD) corresponding pixel PHD, obtain its needle characteristic value Z1=
PHD.Zfeature;
S503, taking-up be confierophyte but be not taxus chinensis in northeast position (HFHD, LFHD)PFHD, obtain its vegetation characteristics value Z2
=PFHD.Zfeature;
S504 calculates needle characteristic mean distance Zmean, calculation formula it is as follows:
S505 takes out a pixel Pixel in Image;
S506 obtains the vegetation characteristics ZT=Pixel.Zfeature of the pixel;
S507, if ZT<0 so goes to S512, otherwise goes to S508;
S508 calculates needle characteristic distance formula ZJL, formula is as follows:
S509, if ZJL>Zmean, then going to S512, otherwise go to S510;
S510 obtains the position (H of Pixelp,Lp);
S511 the, by (H of ResultImagep,Lp) position pixel labeled as white;
S512, if in Image all pixels calculated completion so go to S513, it is no to go to S505;
S513 exports the selection result image ResultImage, and black portions are non-taxus chinensis in northeast in the image, and white
Part is the position for the taxus chinensis in northeast that this method identifies.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810051779.8A CN108154138B (en) | 2018-01-19 | 2018-01-19 | A method of the automatic identification taxus chinensis in northeast in high-resolution remote sensing image |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810051779.8A CN108154138B (en) | 2018-01-19 | 2018-01-19 | A method of the automatic identification taxus chinensis in northeast in high-resolution remote sensing image |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108154138A CN108154138A (en) | 2018-06-12 |
CN108154138B true CN108154138B (en) | 2018-11-06 |
Family
ID=62461637
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810051779.8A Active CN108154138B (en) | 2018-01-19 | 2018-01-19 | A method of the automatic identification taxus chinensis in northeast in high-resolution remote sensing image |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108154138B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109166148B (en) * | 2018-08-08 | 2021-08-27 | 长春工程学院 | Method for extracting crown diameter in remote sensing image of regional gray and directional component operator |
CN112348951B (en) * | 2020-11-30 | 2022-08-26 | 长春工程学院 | Digital elevation data reconstruction method for heterogeneous remote sensing image content |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107330413A (en) * | 2017-07-06 | 2017-11-07 | 中国科学院遥感与数字地球研究所 | A kind of mother drug plants recognition methods based on remote sensing technology |
CN107392133A (en) * | 2017-07-14 | 2017-11-24 | 中国科学院新疆生态与地理研究所 | Utilize the ermophyte remote sensing recognition method of object-oriented Multi-source Information Fusion |
-
2018
- 2018-01-19 CN CN201810051779.8A patent/CN108154138B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107330413A (en) * | 2017-07-06 | 2017-11-07 | 中国科学院遥感与数字地球研究所 | A kind of mother drug plants recognition methods based on remote sensing technology |
CN107392133A (en) * | 2017-07-14 | 2017-11-24 | 中国科学院新疆生态与地理研究所 | Utilize the ermophyte remote sensing recognition method of object-oriented Multi-source Information Fusion |
Non-Patent Citations (2)
Title |
---|
《高分辨率遥感影像天然林与人工林植被覆盖信息提取》;王荣,江东,韩惠等;《资源科学》;20130430;第35卷(第4期);第868-874页 * |
《高分辨率遥感影像目标从分类与识别研究进展》;刘扬,付征叶,郑逢斌;《地球信息科学》;20150930;第17卷(第9期);第1081-1091页 * |
Also Published As
Publication number | Publication date |
---|---|
CN108154138A (en) | 2018-06-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108363949B (en) | Cotton remote sensing monitoring method based on phenological analysis | |
CN107609526A (en) | Rule-based fine dimension city impervious surface rapid extracting method | |
CN110163138B (en) | Method for measuring and calculating wheat tillering density based on multispectral remote sensing image of unmanned aerial vehicle | |
CN105404753A (en) | Marsh wetland mapping method based on object-oriented random forest classification method and medium-resolution remote sensing image | |
CN108154138B (en) | A method of the automatic identification taxus chinensis in northeast in high-resolution remote sensing image | |
CN112800973B (en) | Spartina alterniflora extraction method based on vegetation phenological feature decision | |
CN107504923B (en) | Kelp culture area monitoring method integrating remote sensing image and extension rope information | |
CN105844632B (en) | Rice strain identification based on machine vision and localization method | |
CN105758806B (en) | Remote sensing monitoring method for mulching film farmland based on spectral characteristics | |
CN102175626A (en) | Method for predicting nitrogen content of cucumber leaf based on spectral image analysis | |
CN114612896B (en) | Rice yield prediction method, device and equipment based on remote sensing image | |
CN105404873A (en) | Winter wheat recognition method based on NDVI time sequence coordinate conversion | |
CN102339378A (en) | Method and device for automatically extracting cotton seeds | |
Nayak et al. | Coastal geomorphological and land use and land cover study on some sites of Gulf of Kachchh, Gujarat, West Coast of India using multi-temporal remote sensing data | |
CN106683069A (en) | Method for recognizing inline crops and weeds in seedling stage of farmland | |
CN108537164A (en) | Attack a vital point sowing bud ratio monitoring method and device based on unmanned aerial vehicle remote sensing | |
CN115496939A (en) | Wetland information extraction method based on multi-source remote sensing data | |
Lelong et al. | Application of textural analysis on very high resolution panchromatic images to map coffee orchards in Uganda | |
CN106504258A (en) | A kind of leaf image extracting method and device | |
Finkl et al. | Interpretation of bottom types from IKONOS satellite images of the southern Key West National Wildlife Refuge, Florida, USA | |
Wiratmoko et al. | Using visible spectral-index as alternative methods for identifying levels of Ganoderma Boninese infection | |
Zhongyang et al. | Study on the Land Use and Cover Classification of Zhengzhou Based on Decision Tree | |
Leite et al. | Use of digital images for evaluating soil compaction in the culture of beans | |
Gorokhova et al. | Application of geoinformational technologies and aerospace photography materials for monitoring the land irrigated by the Svetlyi Yar irrigation system (Volgograd oblast) | |
CN114972838B (en) | Winter wheat identification method based on satellite data |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |