CN107273868A - A kind of method that the dump and coal gangue area of coal field are distinguished in remote sensing images - Google Patents
A kind of method that the dump and coal gangue area of coal field are distinguished in remote sensing images Download PDFInfo
- Publication number
- CN107273868A CN107273868A CN201710505726.4A CN201710505726A CN107273868A CN 107273868 A CN107273868 A CN 107273868A CN 201710505726 A CN201710505726 A CN 201710505726A CN 107273868 A CN107273868 A CN 107273868A
- Authority
- CN
- China
- Prior art keywords
- image block
- area
- water body
- coal
- coal field
- 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
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
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of method that coal field dump and coal gangue area are distinguished in remote sensing images.Specially:Region segmentation is first carried out to high-resolution remote sensing image first, irregular object is obtained;Decision tree is recycled, water body area/non-water body area, vegetation region/nonvegetated area, residential block/non-residential areas, bare area/non-bare area, the classification of coal field/non-coal field is carried out successively to irregular object;Then dump, coal gangue area are distinguished from the irregular object for being categorized as coal field using LBP operator extractions textural characteristics.The present invention successively classifies on the basis of decision tree is combined, a preliminary decision region is provided for coal field, converted using K T, effectively coal field is extracted, and utilizes LBP texture operators, effective differentiation is carried out to the dump in coal field and coal gangue area.
Description
Technical field
The present invention relates to field of remote sensing image processing, more particularly to a kind of Classifying Method in Remote Sensing Image of feature extraction
Background technology
Since twentieth century end, the information extraction technology of object-oriented analysis is continued to develop, and the classification of object-oriented is gradually
The main flow analyzed as high-resolution remote sensing image.The basic ideas of object-oriented classification method are that remote sensing images are carried out first
Segmentation, is secondly carried out feature extraction to the image-region that segmentation is obtained, remote sensing images is classified using the feature extracted.
Characteristics of objects refers in the case where being added without any knowledge, certain gone out based on the information extraction that object has in itself
A little attributes, including figure layer value, shape, texture, level and thematic attribute.Wherein figure layer value, shape, the category feature of texture three and on
Context information is relatively common in image procossing and analysis.
K-T convert (i.e. K-T Transformation), be according to the information such as soil, vegetation in multispectral remote sensing in multidimensional spectral space
The empirical linear orthogonal transformation that information distributed architecture is done to image, equivalent to the coordinate space of spin data, refers to reference axis
Particularly relevant with growing process and soil to the direction closely related with atural object, K-T conversion can both help to realize letter
Breath compression, can help to interpret atural object, with important practical significance again.
Abundant texture information is there is in the resolution ratio more and more higher of high-resolution remote sensing image, image, texture is special
Take over for use and have great importance in High spatial resolution remote sensing classification.LBP operators, local binary patterns (Local binary
Patterns, LBP), be the operator for being used in field of machine vision describe image local textural characteristics, with rotational invariance and
The significant advantage such as gray scale consistency.Its basic thought is:Using the gray value of center pixel as threshold value, around center pixel
Pixel in neighborhood constituted after thresholding processing the binary system chain code of description local grain pattern, obtains each pixel in image
Corresponding LBP chains code value, then can obtain a histogram on chain code value, and for describing the textural characteristics of image.
In remote sensing images, due to dump and coal gangue area in spectral signature it is similar, there is no rule on shape facility,
It is high using only the spectral signature mistake point rate of image, and shape facility availability is less, so currently there are no a kind of preferable side
The method that method can distinguish dump and coal gangue area in remote sensing images.
The content of the invention
The goal of the invention of the present invention is:For above-mentioned problem high-resolution remote sensing image is distinguished there is provided one kind
On dump and coal gangue area method.
The method that the dump and coal gangue area of coal field are distinguished in a kind of remote sensing images of the present invention, including following step
Suddenly:
S1:Determine the coal field in pending remote sensing images:
S101:Image segmentation is carried out to pending remote sensing images by watershed method, initial segmentation region is obtained;Lead to again
Cross fractal net work evolution method to merge initial segmentation region, obtain the image block of the remote sensing images;
S102:The image block obtained to step S101 carries out the classification processing in water body area, non-water body area:
Extract the water body feature of image block;
Using the water body feature of image block as the input of water body area SVM classifier, each image block is obtained on water body area, non-
The classification results in water body area;
The training data of wherein water body area SVM classifier is:The image block of remote sensing images with first category label
Water body feature, the first category label includes water body area, non-water body area;
The water body of image block is characterized as:The normalization water body index of each pixel of image block is calculated, is owned by image block
The water body feature for being worth to image block of the normalization water body index of pixel, wherein normalizing water body indexGREEN represents green band of light spectrum, and NIR represents near infrared band spectral value;
S103:Classification results are carried out with the classification processing of vegetation region, nonvegetated area for the image block in non-water body area:
Extract the vegetation characteristics of image block;
Using classification results for non-water body area image block vegetation characteristics as the input of vegetation region SVM classifier, obtain
Each image block is on vegetation region, the classification results of nonvegetated area;
The training data of wherein vegetation region SVM classifier is:The image block of remote sensing images with second category label
Vegetation characteristics, the second category label includes vegetation region, nonvegetated area;
The vegetation characteristics of image block are:The normalized differential vegetation index of each pixel of image block is calculated, is owned by image block
The vegetation characteristics for being worth to image block of the normalized differential vegetation index of pixel, wherein normalized differential vegetation indexNIR represents near infrared band spectral value, and R represents red band of light spectrum;
S104:Classification results are carried out with the classification processing of residential block, non-residential areas for the image block of nonvegetated area:
Extract the residential block feature of image block;
Using classification results for nonvegetated area image block residential block feature as the input of residential block SVM classifier, obtain
To classification results of each image block on residential block, non-residential areas;
Wherein SVM classifier training data in residential block is:The plant of the image block of remote sensing images with the 3rd class label
By feature, the 3rd class label includes residential block, non-residential areas;
The residential block of image block is characterized as:Calculate the spectral value of three wave bands of red, green, blue of each pixel of image block
Variance δ, by the variance δ of the image block all pixels point residential block feature for being worth to image block;
S105:Classification results are carried out with the classification processing of bare area, non-bare area for the image block of non-residential areas:
Extract the bare area feature of image block;
Using classification results for the image block of non-residential areas bare area feature as the input of bare area SVM classifier, obtain each
Classification results of the image block on bare area, non-bare area;
The training data of wherein bare area SVM classifier is:The plant of the image block of remote sensing images with the 4th class label
By feature, the 4th class label includes bare area, non-bare area;
The bare area of image block is characterized as:The iron oxygen index (OI) of each pixel of image block is calculated, by image block all pixels point
Iron oxygen index (OI) the bare area feature for being worth to image block, wherein iron oxygen index (OI) is:R represents the red of remote sensing images
Band of light spectrum, GREEN represents the green band of light spectrum of remote sensing images;
S106:Classification results are carried out with the classification processing of coal field, non-coal field for the image block of non-bare area:
Extract the coal field feature of image block;
Using classification results for non-bare area image block coal field feature as the input of coal field SVM classifier, obtain
Each image block is on coal field, the classification results of non-coal field;
The training data of the coal field SVM classifier is:The image block of remote sensing images with the 5th class label
Vegetation characteristics, the 5th class label includes coal field, non-coal field, and wherein coal field includes coal gangue area, dump area
Domain;
The coal field of image block is characterized as:The spectral signature A of image block is calculated, K-T conversion is carried out to spectral signature A, obtained
Spectral signature KT=A*K after to conversion, using KT as image block coal field feature;
The wherein spectral signature A of image block includes the blue wave band spectral value, green band spectral value, red ripple of image block
Section spectral value, near infrared band spectral value, the spectral value of each image block is the corresponding spectral value of all pixels point of image block
Average;Transformation matrix
S2:Coal gangue area is carried out for the image block of coal field to classification results, the classification in dump region is handled:
Extract the textural characteristics of image block;
Using classification results for coal field image block textural characteristics as the input of the second coal field SVM classifier, obtain
To each image block on coal gangue area, the classification results in dump region;
The training data of the second coal field SVM classifier is:The image of remote sensing images with the 6th class label
The textural characteristics of block, the 6th class label includes coal gangue area, dump region.
Further, in step S2, the textural characteristics for extracting image block are specially:Use the rotation that radius is 8 for 1, neighborhood
Turn constant LBP patternsCalculate the LBP values of each pixel of image block;It is right10 LBP values ascending orders arrangement, will
3rd LBP value is designated as s*;LBP values are equal to s*Pixel account for image block total pixel ratio as image block line
Manage feature.The beneficial effects of the present invention are:A kind of method classified to coal field with feasibility is first proposed,
Coal field is searched in pending remote sensing images, textural characteristics are then based on, dump and gangue are carried out to coal field
Region disconnecting, realizes and indicates that atural object automatically extracts processing to mine on high-resolution remote sensing image.
Brief description of the drawings
Fig. 1 is the specific implementation flow chart of the present invention.
Fig. 2 is decision tree structure figure.
Embodiment
To make the object, technical solutions and advantages of the present invention clearer, with reference to embodiment and accompanying drawing, to this hair
It is bright to be described in further detail.
Referring to Fig. 1, the method that coal field dump and coal gangue area are distinguished in a kind of remote sensing images of the invention, including under
Row step:
S1:Determine the coal field in pending remote sensing images:
Image segmentation is carried out to pending remote sensing images by watershed method first, initial segmentation region is obtained, then
Merge the image block that initial segmentation region obtains pending remote sensing images using FNEA methods (fractal net work evolution method);Again to each
Image block utilizes the judgement mode of decision tree, the classification judgement in water body area/non-water body area is carried out first, then to belonging to non-water body
Image block carry out the classification judgement of vegetation region/nonvegetated area, residential block/non-then is carried out to the image block for belonging to nonvegetated area
Resident's region class is adjudicated, and the classification for carrying out bare area/non-bare area to the image block for belonging to non-residential areas is adjudicated, then non-to belonging to
The image block of bare area carries out the classification judgement of coal field/non-coal field, as shown in Figure 2.When classifying judgement, based on what is trained
SVM (SVMs) graders carry out classification decision process.Each svm classifier is then that the training sample based on known class enters
Row training is obtained, and trains the characteristic information extracted during with Classification and Identification consistent.
In the classification judgement in water body area/non-water body area, the characteristic information of input SVM classifier is:Normalization water body refers to
Number
In the classification judgement of vegetation region/nonvegetated area, the characteristic information of input SVM classifier is:Normalization vegetation refers to
Number
In residential block/non-residential areas classification judgement, the characteristic information of input SVM classifier is:Remote sensing images are red, green,
The variance δ of the spectral value of blue three wave bands;
In the classification judgement of bare area/non-bare area, the characteristic information of input SVM classifier is:
The GREEN occurred during above-mentioned classification judgement represents green band of light spectrum, and NIR represents near infrared band spectral value, R tables
Show red band of light spectrum.
In the classification judgement of coal field/non-coal field, the characteristic information of input SVM classifier is:To the light of image block
Spectrum signature A obtains four features after carrying out K-T conversion, i.e., obtain the spectral signature KT after K-T is converted by KT=A*K, then by KT
Four features be used as the characteristic information for being input to corresponding SVM classifier.
Wherein A=[b;g;r;Nir], b, g, r, nir represent the blue wave band spectral value of image block, green band light respectively
Spectrum, red band spectral value, near infrared band spectral value;Transformation matrix
When extracting the character pair information of each image block, extract first each pixel characteristic information (NDWI, NDVI,
δ, iron oxygen index (OI), spectral value), then by image block all pixels point character pair information the feature for being worth to image block
Information.
S2:Coal gangue area, the differentiation processing in dump region are carried out to coal field:
Textural characteristics of the classification results for the image block of coal field are extracted, and textural characteristics are inputted for distinguishing bastard coal
In stone, the SVM classifier of dump, the classification results of each image block are obtained;Then coal is obtained by the image block of same classification results
Coal gangue area, dump region in mining area;The training data of the wherein SVM classifier is:Coal gangue area, dump are distinguished
The textural characteristics of the image block in region.
In present embodiment, when extracting the texture feature extraction of image block, using with uniformity invariable rotary
The LBP patterns of propertyI.e.:
Wherein,
In above-mentioned expression formula, P represents field element number, and R represents the distance of radius, i.e. distance center pixel, function
Sign (x) is sign function, if x is more than 0, sign (x)=1, otherwise sign (x)=0;gcRepresent center pixel value, gp-1、
gp、g0Represent and pixel identifier is designated as under the pixel value around center pixel, i.e. pixel value, wherein c represents center pixel
Point, other then represent central pixel point c field pixel.It is preferred thatForI.e. radius is 1, and field is 8
The form of individual element, is 10 LBP values by P+2 is produced, and by the LBP values of this 10 different values, number consecutively is 0 from small to large
~9, using LBP values label for 2 pixel number and the total pixel number of image block ratio as image block textural characteristics.
To sum up, the side of a kind of dump distinguished on high-resolution remote sensing image and coal gangue area has been incorporated herein in the present invention
Method, using decision tree hierarchical classification, is extracted, effect is good using K-T conversion to coal field;Recycle LBP operators, it was found that
A kind of suitable textural characteristics, are relatively adapted to separate dump and coal gangue area, to a certain extent there is provided one kind point coal field
The feasible program of upper dump and coal gangue area, and it is simple operation, quick.
The foregoing is only a specific embodiment of the invention, any feature disclosed in this specification, except non-specifically
Narration, can alternative features equivalent by other or with similar purpose replaced;Disclosed all features or all sides
Method or during the step of, in addition to mutually exclusive feature and/or step, can be combined in any way.
Claims (2)
1. the method for the dump and coal gangue area of coal field is distinguished in a kind of remote sensing images, it is characterised in that including following step
Suddenly:
S1:Determine the coal field in pending remote sensing images:
S101:Image segmentation is carried out to pending remote sensing images by watershed method, initial segmentation region is obtained;Again by dividing
L network evolution method is merged to initial segmentation region, obtains the image block of the remote sensing images;
S102:The image block obtained to step S101 carries out the classification processing in water body area, non-water body area:
Extract the water body feature of image block;
Using the water body feature of image block as the input of water body area SVM classifier, each image block is obtained on water body area, non-water body
The classification results in area;
The training data of wherein water body area SVM classifier is:The water body of the image block of remote sensing images with first category label
Feature, the first category label includes water body area, non-water body area;
The water body of image block is characterized as:The normalization water body index of each pixel of image block is calculated, by image block all pixels point
The water body feature for being worth to image block of water body index is normalized, wherein normalizing water body index
GREEN represents green band of light spectrum, and NIR represents near infrared band spectral value;
S103:Classification results are carried out with the classification processing of vegetation region, nonvegetated area for the image block in non-water body area:
Extract the vegetation characteristics of image block;
Using classification results for non-water body area image block vegetation characteristics as the input of vegetation region SVM classifier, obtain each figure
As block is on vegetation region, the classification results of nonvegetated area;
The training data of wherein vegetation region SVM classifier is:The vegetation of the image block of remote sensing images with second category label
Feature, the second category label includes vegetation region, nonvegetated area;
The vegetation characteristics of image block are:The normalized differential vegetation index of each pixel of image block is calculated, by image block all pixels point
Normalized differential vegetation index the vegetation characteristics for being worth to image block, wherein normalized differential vegetation index
NIR represents near infrared band spectral value, and R represents red band of light spectrum;
S104:Classification results are carried out with the classification processing of residential block, non-residential areas for the image block of nonvegetated area:
Extract the residential block feature of image block;
Using classification results for the image block of nonvegetated area residential block feature as the input of residential block SVM classifier, obtain each
Classification results of the image block on residential block, non-residential areas;
Wherein SVM classifier training data in residential block is:The vegetation of the image block of remote sensing images with the 3rd class label is special
Levy, the 3rd class label includes residential block, non-residential areas;
The residential block of image block is characterized as:Calculate the variance of the spectral value of three wave bands of red, green, blue of each pixel of image block
δ, by the variance δ of the image block all pixels point residential block feature for being worth to image block;
S105:Classification results are carried out with the classification processing of bare area, non-bare area for the image block of non-residential areas:
Extract the bare area feature of image block;
Using classification results for non-residential areas image block bare area feature as the input of bare area SVM classifier, obtain each image
Classification results of the block on bare area, non-bare area;
The training data of wherein bare area SVM classifier is:The vegetation of the image block of remote sensing images with the 4th class label is special
Levy, the 4th class label includes bare area, non-bare area;
The bare area of image block is characterized as:The iron oxygen index (OI) of each pixel of image block is calculated, by the iron of image block all pixels point
The bare area feature for being worth to image block of oxygen index (OI), wherein iron oxygen index (OI) is:R represents the red wave band of remote sensing images
Spectral value, GREEN represents the green band of light spectrum of remote sensing images;
S106:Classification results are carried out with the classification processing of coal field, non-coal field for the image block of non-bare area:
Extract the coal field feature of image block;
Using classification results for non-bare area image block coal field feature as the input of coal field SVM classifier, obtain each figure
As block is on coal field, the classification results of non-coal field;
The training data of the coal field SVM classifier is:The vegetation of the image block of remote sensing images with the 5th class label
Feature, the 5th class label includes coal field, non-coal field, and wherein coal field includes coal gangue area, dump region;
The coal field of image block is characterized as:The spectral signature A of image block is calculated, K-T conversion is carried out to spectral signature A, become
Spectral signature KT=A*K after changing, using KT as image block coal field feature;
The wherein spectral signature A of image block includes blue wave band spectral value, green band spectral value, the red band light of image block
Spectrum, near infrared band spectral value, the spectral value of each image block are the equal of the corresponding spectral value of all pixels point of image block
Value;Transformation matrix
S2:Coal gangue area is carried out for the image block of coal field to classification results, the classification in dump region is handled:
Extract the textural characteristics of image block;
Using classification results for the image block of coal field textural characteristics as the input of the second coal field SVM classifier, obtain each
Image block is on coal gangue area, the classification results in dump region;
The training data of the second coal field SVM classifier is:The image block of remote sensing images with the 6th class label
Textural characteristics, the 6th class label includes coal gangue area, dump region.
2. the method as described in claim 1, it is characterised in that in step S2, the textural characteristics for extracting image block are specially:
Use the invariable rotary LBP patterns that radius is 8 for 1, neighborhoodCalculate the LBP values of each pixel of image block;
It is right10 LBP values ascending orders arrangement, the 3rd LBP value is designated as s*;
LBP values are equal to s*Pixel account for image block total pixel ratio as image block textural characteristics.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710505726.4A CN107273868A (en) | 2017-06-28 | 2017-06-28 | A kind of method that the dump and coal gangue area of coal field are distinguished in remote sensing images |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710505726.4A CN107273868A (en) | 2017-06-28 | 2017-06-28 | A kind of method that the dump and coal gangue area of coal field are distinguished in remote sensing images |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107273868A true CN107273868A (en) | 2017-10-20 |
Family
ID=60071505
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710505726.4A Pending CN107273868A (en) | 2017-06-28 | 2017-06-28 | A kind of method that the dump and coal gangue area of coal field are distinguished in remote sensing images |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107273868A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108710862A (en) * | 2018-05-24 | 2018-10-26 | 河海大学 | A kind of high-resolution remote sensing image Clean water withdraw method |
CN109215034A (en) * | 2018-07-06 | 2019-01-15 | 成都图必优科技有限公司 | A kind of Weakly supervised image, semantic dividing method for covering pond based on spatial pyramid |
CN109558859A (en) * | 2018-12-13 | 2019-04-02 | 中国地质大学(武汉) | Mining area distributed intelligence extracting method and system based on DInSAR and DCNN |
CN112200813A (en) * | 2020-09-30 | 2021-01-08 | 中国矿业大学(北京) | Coal and gangue identification method and system considering illumination factor |
CN112215186A (en) * | 2020-10-21 | 2021-01-12 | 深圳市赛为智能股份有限公司 | Marsh wetland vegetation classification method, device, computer equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104143191A (en) * | 2014-07-29 | 2014-11-12 | 中国科学院遥感与数字地球研究所 | Remote sensing image change detection method based on texton |
CN105404753A (en) * | 2015-12-08 | 2016-03-16 | 中国科学院东北地理与农业生态研究所 | Marsh wetland mapping method based on object-oriented random forest classification method and medium-resolution remote sensing image |
CN106650689A (en) * | 2016-12-30 | 2017-05-10 | 厦门理工学院 | Coastal city time sequence land utilization information extracting method |
US20170140205A1 (en) * | 2013-12-03 | 2017-05-18 | Digitalglobe, Inc. | Automated compound structure characterization in overhead imagery |
WO2017099951A1 (en) * | 2015-12-07 | 2017-06-15 | The Climate Corporation | Cloud detection on remote sensing imagery |
-
2017
- 2017-06-28 CN CN201710505726.4A patent/CN107273868A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170140205A1 (en) * | 2013-12-03 | 2017-05-18 | Digitalglobe, Inc. | Automated compound structure characterization in overhead imagery |
CN104143191A (en) * | 2014-07-29 | 2014-11-12 | 中国科学院遥感与数字地球研究所 | Remote sensing image change detection method based on texton |
WO2017099951A1 (en) * | 2015-12-07 | 2017-06-15 | The Climate Corporation | Cloud detection on remote sensing imagery |
CN105404753A (en) * | 2015-12-08 | 2016-03-16 | 中国科学院东北地理与农业生态研究所 | Marsh wetland mapping method based on object-oriented random forest classification method and medium-resolution remote sensing image |
CN106650689A (en) * | 2016-12-30 | 2017-05-10 | 厦门理工学院 | Coastal city time sequence land utilization information extracting method |
Non-Patent Citations (1)
Title |
---|
魏雯婷: "基于图像纹理和SVR的视线估计系统研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108710862A (en) * | 2018-05-24 | 2018-10-26 | 河海大学 | A kind of high-resolution remote sensing image Clean water withdraw method |
CN108710862B (en) * | 2018-05-24 | 2022-06-28 | 河海大学 | High-resolution remote sensing image water body extraction method |
CN109215034A (en) * | 2018-07-06 | 2019-01-15 | 成都图必优科技有限公司 | A kind of Weakly supervised image, semantic dividing method for covering pond based on spatial pyramid |
CN109558859A (en) * | 2018-12-13 | 2019-04-02 | 中国地质大学(武汉) | Mining area distributed intelligence extracting method and system based on DInSAR and DCNN |
CN109558859B (en) * | 2018-12-13 | 2020-11-24 | 中国地质大学(武汉) | Mining area distribution information extraction method and system based on DInSAR and DCNN |
CN112200813A (en) * | 2020-09-30 | 2021-01-08 | 中国矿业大学(北京) | Coal and gangue identification method and system considering illumination factor |
CN112200813B (en) * | 2020-09-30 | 2024-02-06 | 中国矿业大学(北京) | Coal gangue identification method and system considering illumination factors |
CN112215186A (en) * | 2020-10-21 | 2021-01-12 | 深圳市赛为智能股份有限公司 | Marsh wetland vegetation classification method, device, computer equipment and storage medium |
CN112215186B (en) * | 2020-10-21 | 2024-04-05 | 深圳市赛为智能股份有限公司 | Classification method, device, computer equipment and storage medium for marsh wetland vegetation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107273868A (en) | A kind of method that the dump and coal gangue area of coal field are distinguished in remote sensing images | |
CN108573276B (en) | Change detection method based on high-resolution remote sensing image | |
Zhang et al. | A multi-level context-guided classification method with object-based convolutional neural network for land cover classification using very high resolution remote sensing images | |
CN104915636B (en) | Remote sensing image road recognition methods based on multistage frame significant characteristics | |
Erikson | Species classification of individually segmented tree crowns in high-resolution aerial images using radiometric and morphologic image measures | |
CN102194114B (en) | Method for recognizing iris based on edge gradient direction pyramid histogram | |
CN104182763B (en) | A kind of floristics identifying system based on flower feature | |
CN102496034B (en) | High-spatial resolution remote-sensing image bag-of-word classification method based on linear words | |
CN107657257A (en) | A kind of semantic image dividing method based on multichannel convolutive neutral net | |
CN105069466A (en) | Pedestrian clothing color identification method based on digital image processing | |
CN107392130A (en) | Classification of Multispectral Images method based on threshold adaptive and convolutional neural networks | |
CN111626092B (en) | Unmanned aerial vehicle image ground crack identification and extraction method based on machine learning | |
CN110598564B (en) | OpenStreetMap-based high-spatial-resolution remote sensing image transfer learning classification method | |
CN111191628B (en) | Remote sensing image earthquake damage building identification method based on decision tree and feature optimization | |
Timilsina et al. | Mapping urban trees within cadastral parcels using an object-based convolutional neural network | |
CN106971397B (en) | Based on the city high-resolution remote sensing image dividing method for improving JSEG algorithms | |
Rahman et al. | Segmentation of color image using adaptive thresholding and masking with watershed algorithm | |
CN104217440B (en) | A kind of method extracting built-up areas from remote sensing images | |
CN113221881B (en) | Multi-level smart phone screen defect detection method | |
CN107992856A (en) | High score remote sensing building effects detection method under City scenarios | |
CN104463091B (en) | A kind of facial image recognition method based on image LGBP feature subvectors | |
CN104392431B (en) | A kind of image matching method based on image Variable Length Code | |
Zheng et al. | Category-wise fusion and enhancement learning for multimodal remote sensing image semantic segmentation | |
O’Neil-Dunne et al. | Incorporating contextual information into object-based image analysis workflows | |
CN111882573A (en) | Cultivated land plot extraction method and system based on high-resolution image 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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171020 |
|
RJ01 | Rejection of invention patent application after publication |