CN111860204A - Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology - Google Patents

Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology Download PDF

Info

Publication number
CN111860204A
CN111860204A CN202010602640.5A CN202010602640A CN111860204A CN 111860204 A CN111860204 A CN 111860204A CN 202010602640 A CN202010602640 A CN 202010602640A CN 111860204 A CN111860204 A CN 111860204A
Authority
CN
China
Prior art keywords
semantic segmentation
remote sensing
change
sensing image
segmentation model
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
Application number
CN202010602640.5A
Other languages
Chinese (zh)
Inventor
不公告发明人
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chengdu Shuzhilian Technology Co Ltd
Original Assignee
Chengdu Shuzhilian Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Chengdu Shuzhilian Technology Co Ltd filed Critical Chengdu Shuzhilian Technology Co Ltd
Priority to CN202010602640.5A priority Critical patent/CN111860204A/en
Publication of CN111860204A publication Critical patent/CN111860204A/en
Pending legal-status Critical Current

Links

Images

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10032—Satellite or aerial image; Remote sensing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20084—Artificial neural networks [ANN]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Astronomy & Astrophysics (AREA)
  • Remote Sensing (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a multi-temporal remote sensing image change detection method and medium based on a semantic segmentation technology, relates to the field of intelligent interpretation of remote sensing images, and solves the problems that the traditional multi-temporal remote sensing change detection is mostly manually searched and compared on an image region by region, and the method is time-consuming, labor-consuming, inefficient and the like. Comparing the multi-temporal remote sensing images, and exporting an annotation result graph after whether the annotation changes; and analyzing the pixel pseudo change of the geographical deviation in the labeling result graph, setting a threshold value, processing the labeling result graph based on the morphological operation of the image, and screening the threshold value of a communication area in the labeling result graph. According to the method, the change area is found out through a deep learning technology, so that the manual work is more focused on the area automatically detected by the model when the change is judged, and the effects of saving the manual work and improving the efficiency are achieved.

Description

Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology
Technical Field
The invention relates to the field of intelligent interpretation of remote sensing images, in particular to a multi-temporal remote sensing image change detection method and medium based on a semantic segmentation technology.
Background
In recent years, deep learning techniques have made significant progress in the fields of image, speech, text, and time series data analysis, and a Convolutional Neural Network (CNN) has been widely used in the field of remote sensing images and has been highly successful. The change detection based on the remote sensing images refers to a technology of extracting change information from a plurality of remote sensing images acquired from the same region at different times, analyzing and understanding the change information, and generating a change distribution map and other detection results.
The traditional multi-temporal remote sensing change detection is mostly to search and compare the images region by region in a manual mode, and the method has the problems of time consumption, labor consumption, low efficiency and the like.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the traditional multi-temporal remote sensing change detection is mostly to search and compare the images region by region in a manual mode, and the method has the problems of time consumption, labor consumption, low efficiency and the like.
A multi-temporal remote sensing image change detection method based on a semantic segmentation technology comprises the following steps:
s1, preprocessing stage: the method comprises the steps of data preparation, semantic segmentation model construction and training of the semantic segmentation model;
s2, prediction stage: a model prediction stage is carried out based on the trained semantic segmentation model in the S1;
Comparing the multi-temporal remote sensing images, and deriving an annotation result graph after whether the annotation changes;
and analyzing the pixel pseudo change of the geographical deviation in the labeling result graph, setting a threshold value, processing the labeling result graph based on the morphological operation of the image, and screening the threshold value of a communication area in the labeling result graph.
Further, the data are remote sensing images and ground class labels corresponding to the remote sensing images, and the data form a training data set.
Further, the semantic segmentation model construction process:
the method comprises the steps of building a semantic segmentation model based on a CNN technology, building a convolutional neural network of the semantic segmentation model, and predicting by adopting a pixel-level deep learning algorithm, wherein the method comprises the following steps: the semantic segmentation algorithm based on deep learning is a pixel-level prediction, namely: if the spatial dimension of the input image is wxh, then the spatial dimension of the prediction map is also wxh. Common semantic segmentation networks comprise FCN, Deeplab, SegNet and the like, and the semantic segmentation network is not specially specified in the patent;
and training the semantic segmentation model by using the data in an iterative optimization mode.
Further, the detailed steps of the prediction phase are as follows:
s21, preparing data to be predicted:
obtaining a remote sensing image of an area to be predicted on a time axis;
S22, data prediction:
using the semantic segmentation model to segment multiple temporal phases tmImage sum tnThe image is predicted to obtain the corresponding prediction map predmAnd predn；
S23, comparison of prediction results:
will predict the map predmAnd prednFor comparison, compare the graphs:
Figure BDA0002559597220000021
wherein (i, j) represents a position in the prediction map, y (i, j) being 0 represents that the position (i, j) is changed, and 1 represents no change;
s24, post-processing of comparison results:
presetting a threshold T, and performing post-processing comparison on the image y by using morphological operation of the imagemn；
The treatment process comprises the following steps: and (4) carrying out threshold value screening on the connected area, and reserving the area of which the connected area is larger than the threshold value T to obtain a final change detection result.
Further, the semantic segmentation model constructed in the S1 adopts a deep lab V3 algorithm, and adds an attention mechanism for image segmentation in the output process of a decoder in the semantic segmentation model.
Further, the remote sensing image in the data and the ground class label corresponding to the remote sensing image are cut into the same size and used for training the semantic segmentation model.
Further, a computer-readable storage medium, in which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method. The specific use of the method relies on a large number of calculations and it is therefore preferred that the above calculation is performed by a computer program, so any computer program and its storage medium containing the steps protected in the method also fall within the scope of the present application.
The invention has the following advantages and beneficial effects:
according to the method, the change area is found out through a deep learning technology, so that the manual work is more focused on the area automatically detected by the model when the change is judged, and the effects of saving the manual work and improving the efficiency are achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the embodiments of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the principles of the invention. In the drawings:
FIG. 1 is a schematic diagram of a segmentation model according to the present invention.
FIG. 2 is a flow chart of the change detection of the present invention.
FIG. 3 is a schematic view of the attention-adding mechanism of the present invention.
Detailed Description
Before any embodiments of the invention are explained in detail, it is to be understood that the invention is not limited in its application to the details of construction and the arrangements of components set forth in the following description or illustrated in the drawings. The invention is capable of other embodiments and of being practiced or of being carried out in various ways. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any inventive changes, are within the scope of the present invention.
Example 1:
as shown in fig. 1-3, the method for detecting changes of multi-temporal remote sensing images based on semantic segmentation technology includes the following steps:
S1, preprocessing stage: the method comprises the steps of data preparation, semantic segmentation model construction and training of the semantic segmentation model;
s2, prediction stage: a model prediction stage is carried out based on the trained semantic segmentation model in the S1;
comparing the multi-temporal remote sensing images, and deriving an annotation result graph after whether the annotation changes;
and analyzing the pixel pseudo change of the geographical deviation in the labeling result graph, setting a threshold value, processing the labeling result graph based on the morphological operation of the image, and screening the threshold value of a communication area in the labeling result graph.
Further, the data are remote sensing images and ground class labels corresponding to the remote sensing images, and the data form a training data set.
Further, the semantic segmentation model construction process:
the method comprises the steps of building a semantic segmentation model based on a CNN technology, building a convolutional neural network of the semantic segmentation model, and predicting by adopting a pixel-level deep learning algorithm, wherein the method comprises the following steps: the semantic segmentation algorithm based on deep learning is a pixel-level prediction, namely: if the spatial dimension of the input image is wxh, then the spatial dimension of the prediction map is also wxh. Common semantic segmentation networks comprise FCN, Deeplab, SegNet and the like, and the semantic segmentation network is not specially specified in the patent;
And training the semantic segmentation model by using the data in an iterative optimization mode.
Further, the detailed steps of the prediction phase are as follows:
s21, preparing data to be predicted:
obtaining a remote sensing image of an area to be predicted on a time axis;
s22, data prediction:
using the semantic segmentation model to segment multiple temporal phases tmImage sum tnThe image is predicted to obtain the corresponding prediction map predmAnd predn；
S23, comparison of prediction results:
will predict the map predmAnd prednFor comparison, compare the graphs:
Figure BDA0002559597220000041
wherein (i, j) represents a position in the prediction map, y (i, j) being 0 represents that the position (i, j) is changed, and 1 represents no change;
s24, post-processing of comparison results:
presetting a threshold T, and performing post-processing comparison on the image y by using morphological operation of the imagemn；
The treatment process comprises the following steps: and (4) carrying out threshold value screening on the connected area, and reserving the area of which the connected area is larger than the threshold value T to obtain a final change detection result.
Further, the semantic segmentation model constructed in the S1 adopts a deep lab V3 algorithm, and adds an attention mechanism for image segmentation in the output process of a decoder in the semantic segmentation model.
Further, the remote sensing image in the data and the ground class label corresponding to the remote sensing image are cut into the same size and used for training the semantic segmentation model.
Further, a computer-readable storage medium, in which a computer program is stored, which computer program, when being executed by a processor, carries out the steps of the method. The specific use of the method relies on a large number of calculations and it is therefore preferred that the above calculation is performed by a computer program, so any computer program and its storage medium containing the steps protected in the method also fall within the scope of the present application.
Example 2:
1) preparing data:
a) the samples of vegetation, buildings, crops and the like in the same city are distributed relatively uniformly. Administrative counties of Panzhihua city include: east, west, benevolence and district, mieyi county, salina county. In the embodiment, the Panzhihua city is selected as a demonstration point, remote sensing images of the west region, the kernel and region, the Miyi county and the Yanbian county are used as training samples, and remote sensing images of the east region are used as testing samples. In the training set and the test set, the resolution of the remote sensing images adopted by all counties is 1m level.
b) The corresponding relation between the labels and the categories in the training data is as follows:
tag and category comparison table
Figure BDA0002559597220000051
A non-building area:
consists of landform images related to cultivated land, garden land, forest land and grassland;
And (3) building areas:
the system consists of a building land, push filling soil, and landform images related to mining areas and parks.
Road:
consists of a road-related relief image.
And other land occupation:
it is composed of other landform images such as port and dock, water surface and mud flat, glacier and permanent snow.
2) Constructing a semantic segmentation model:
when the semantic segmentation model is constructed, the algorithm framework of deep lab V3 is adopted. Meanwhile, in order to make reasonable use of spatial resolution and channel resolution in the feature map, the reference DANet adds a mechanism of attention in deep lab V3, as shown in fig. 3.
3) Training a segmentation model:
and cutting the training data set and the corresponding labels into sizes, and training a semantic segmentation model.
Prediction phase of the model:
(1) preparing data:
and acquiring multi-temporal remote sensing images of the east region at t1 time and t2 time, wherein the resolutions of the two images are 1m level.
(2) And (3) data prediction:
and respectively inputting the test data of the east region into the trained model to obtain the prediction graphs corresponding to the t1 and t 2.
(3) And (3) comparison of predicted results:
and comparing the prediction image with the pixel by pixel to find out areas with different predicted values.
(4) And (4) post-processing comparison results:
processing comparison results after morphological operation, and performing threshold screening on a connected region, specifically: a threshold of 1500 is set, indicating that connected regions greater than 1500 in the image are retained.
Prediction results of the model:
item Number of changes Ratio of occupation of Area (, square meter) Area to area ratio
Changing pattern spot 294 100％ 3707657 100％
Actual change 215 73％ 2846054 77％
Pseudo variations 53 18％ 602986 16％
Cloud overlay induced changes 8 3％ 134098 4％
Image offset variation 18 6％ 124519 3％
In order to verify the effectiveness of change detection, 294 multi-temporal change pattern spots of t1 and t2 times are extracted manually; the number of change patterns detected by the algorithm is 215, the number of pseudo changes is 53, the number of changes due to cloud coverage is 8, and the number of changes due to multi-phase pixel shift is 18. The area ratio of the change pattern spots detected by the algorithm to the manually extracted change pattern spots was 77%.
Example 3:
in the technical effect, for the multi-temporal remote sensing image, 1 is a change area of change detection, 0 is an unchanged area of change detection, and the change area and the unchanged area are marked by two color coating marks.
The traditional multi-temporal remote sensing change detection is mostly to search and compare the images region by region in a manual mode, and the method has the problems of time consumption, labor consumption, low efficiency and the like. This patent aims at assisting artifical judgement, promptly: the change area is found out through the deep learning technology, so that the manual work is more focused on the area detected by the automatic model when the change is judged, and the effects of saving the manual work and improving the efficiency are achieved.
The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims (7)

1. The method for detecting the change of the multi-temporal remote sensing image based on the semantic segmentation technology is characterized by comprising the following steps of:
s1, preprocessing stage: the method comprises the steps of data preparation, semantic segmentation model construction and training of the semantic segmentation model;
s2, prediction stage: a model prediction stage is carried out based on the trained semantic segmentation model in the S1;
comparing the multi-temporal remote sensing images, and deriving an annotation result graph after whether the annotation changes;
and analyzing the pixel pseudo change of the geographical deviation in the labeling result graph, setting a threshold value, processing the labeling result graph based on the morphological operation of the image, and screening the threshold value of a communication area in the labeling result graph.
2. The method for detecting the change of the multi-temporal remote sensing image based on the semantic segmentation technology as claimed in claim 1, wherein the data are the remote sensing image and a ground class label corresponding to the remote sensing image, and the data form a training data set.
3. The method for detecting the change of the multi-temporal remote sensing image based on the semantic segmentation technology as claimed in claim 2, wherein the semantic segmentation model construction process comprises the following steps:
building a semantic segmentation model based on a CNN technology, building a convolutional neural network of the semantic segmentation model, and predicting by adopting a pixel-level deep learning algorithm;
the semantic segmentation network in the semantic segmentation model comprises FCN, Deeplab and SegNet;
and training the semantic segmentation model by using the data in an iterative optimization mode.
4. The method for detecting the change of the multi-temporal remote sensing image based on the semantic segmentation technology as claimed in claim 3, wherein the detailed steps of the prediction stage are as follows:
s21, preparing data to be predicted:
obtaining a remote sensing image of an area to be predicted on a time axis;
s22, data prediction:
using the semantic segmentation model to segment multiple temporal phases tmImage sum tnThe image is predicted to obtain the corresponding prediction map predmAnd predn；
S23, comparison of prediction results:
will predict the map predmAnd prednFor comparison, compare the graphs:
Figure FDA0002559597210000011
wherein (i, j) represents a position in the prediction map, y (i, j) being 0 represents that the position (i, j) is changed, and 1 represents no change;
s24, post-processing of comparison results:
Presetting a threshold T, and performing post-processing comparison on the image y by using morphological operation of the imagemn；
The treatment process comprises the following steps: and (4) carrying out threshold value screening on the connected area, and reserving the area of which the connected area is larger than the threshold value T to obtain a final change detection result.
5. The method for detecting the change of the multi-temporal remote sensing image based on the semantic segmentation technology as claimed in claim 4, wherein the semantic segmentation model constructed in the step S1 adopts a DeepLab V3 algorithm, and an attention mechanism of image segmentation is added in the output process of a decoder in the semantic segmentation model.
6. The method for detecting the change of the multi-temporal remote sensing image based on the semantic segmentation technology as claimed in claim 2, wherein the remote sensing image in the data and the ground class labels corresponding to the remote sensing image are cut into the same size for training the semantic segmentation model.
7. A computer-readable storage medium, in which a computer program is stored which, when being executed by a processor, carries out the steps of the method according to any one of claims 1 to 6.
CN202010602640.5A 2020-06-29 2020-06-29 Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology Pending CN111860204A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010602640.5A CN111860204A (en) 2020-06-29 2020-06-29 Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010602640.5A CN111860204A (en) 2020-06-29 2020-06-29 Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology

Publications (1)

Publication Number Publication Date
CN111860204A true CN111860204A (en) 2020-10-30

Family

ID=72989597

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010602640.5A Pending CN111860204A (en) 2020-06-29 2020-06-29 Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology

Country Status (1)

Country Link
CN (1) CN111860204A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112668494A (en) * 2020-12-31 2021-04-16 西安电子科技大学 Small sample change detection method based on multi-scale feature extraction
CN112819792A (en) * 2021-02-03 2021-05-18 杭州高斯洪堡科技有限公司 DualNet-based urban area change detection method
CN112950655A (en) * 2021-03-08 2021-06-11 甘肃农业大学 Land use information automatic extraction method based on deep learning
CN113298042A (en) * 2021-06-22 2021-08-24 中国平安财产保险股份有限公司 Method and device for processing remote sensing image data, storage medium and computer equipment
CN113408537A (en) * 2021-07-19 2021-09-17 中南大学 Adaptive semantic segmentation method for remote sensing image domain
CN113505636A (en) * 2021-05-25 2021-10-15 中国科学院空天信息创新研究院 Mining area change detection method based on attention mechanism and full convolution twin neural network

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103971115A (en) * 2014-05-09 2014-08-06 中国科学院遥感与数字地球研究所 Automatic extraction method for newly-increased construction land image spots in high-resolution remote sensing images based on NDVI and PanTex index
CN108089850A (en) * 2018-01-02 2018-05-29 北京建筑大学 A kind of ground mulching product increment updating method based on image collaboration segmentation with ecogeography zoning ordinance storehouse
CN108776772A (en) * 2018-05-02 2018-11-09 北京佳格天地科技有限公司 Across the time building variation detection modeling method of one kind and detection device, method and storage medium
CN109215038A (en) * 2018-09-29 2019-01-15 中国资源卫星应用中心 A kind of intelligent information retrieval method and system based on remote sensing image
CN109377480A (en) * 2018-09-27 2019-02-22 中国电子科技集团公司第五十四研究所 Arable land use change detection method based on deep learning
WO2019237646A1 (en) * 2018-06-14 2019-12-19 清华大学深圳研究生院 Image retrieval method based on deep learning and semantic segmentation

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103971115A (en) * 2014-05-09 2014-08-06 中国科学院遥感与数字地球研究所 Automatic extraction method for newly-increased construction land image spots in high-resolution remote sensing images based on NDVI and PanTex index
CN108089850A (en) * 2018-01-02 2018-05-29 北京建筑大学 A kind of ground mulching product increment updating method based on image collaboration segmentation with ecogeography zoning ordinance storehouse
CN108776772A (en) * 2018-05-02 2018-11-09 北京佳格天地科技有限公司 Across the time building variation detection modeling method of one kind and detection device, method and storage medium
WO2019237646A1 (en) * 2018-06-14 2019-12-19 清华大学深圳研究生院 Image retrieval method based on deep learning and semantic segmentation
CN109377480A (en) * 2018-09-27 2019-02-22 中国电子科技集团公司第五十四研究所 Arable land use change detection method based on deep learning
CN109215038A (en) * 2018-09-29 2019-01-15 中国资源卫星应用中心 A kind of intelligent information retrieval method and system based on remote sensing image

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
汪梓艺 等: "一种改进DeeplabV3 网络的烟雾分割算法", 《西安电子科技大学报》, vol. 46, no. 6, pages 1 - 8 *
胡永月 等: "面向对象变化检测中多时相图像分割模式影响评价", 《南京大学学报（自然科学）》, vol. 51, no. 5, pages 1049 - 1057 *

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112668494A (en) * 2020-12-31 2021-04-16 西安电子科技大学 Small sample change detection method based on multi-scale feature extraction
CN112819792A (en) * 2021-02-03 2021-05-18 杭州高斯洪堡科技有限公司 DualNet-based urban area change detection method
CN112950655A (en) * 2021-03-08 2021-06-11 甘肃农业大学 Land use information automatic extraction method based on deep learning
CN113505636A (en) * 2021-05-25 2021-10-15 中国科学院空天信息创新研究院 Mining area change detection method based on attention mechanism and full convolution twin neural network
CN113505636B (en) * 2021-05-25 2023-11-17 中国科学院空天信息创新研究院 Mining area change detection method based on attention mechanism and full convolution twin neural network
CN113298042A (en) * 2021-06-22 2021-08-24 中国平安财产保险股份有限公司 Method and device for processing remote sensing image data, storage medium and computer equipment
CN113298042B (en) * 2021-06-22 2024-02-02 中国平安财产保险股份有限公司 Remote sensing image data processing method and device, storage medium and computer equipment
CN113408537A (en) * 2021-07-19 2021-09-17 中南大学 Adaptive semantic segmentation method for remote sensing image domain

Similar Documents

Publication Publication Date Title
CN111860204A (en) Multi-temporal remote sensing image change detection method and medium based on semantic segmentation technology
Lin et al. Measuring the relationship between morphological spatial pattern of green space and urban heat island using machine learning methods
Zhang et al. Scale Sequence Joint Deep Learning (SS-JDL) for land use and land cover classification
Martínez-Casasnovas et al. Mapping multi-year cropping patterns in small irrigation districts from time-series analysis of Landsat TM images
CN108846832B (en) Multi-temporal remote sensing image and GIS data based change detection method and system
Hu et al. Impervious surface area extraction from IKONOS imagery using an object-based fuzzy method
Sui et al. Modeling the dynamics of landscape structure in Asia’s emerging desakota regions: a case study in Shenzhen
CN111028255B (en) Farmland area pre-screening method and device based on priori information and deep learning
WO2024020744A1 (en) High-resolution image-based ecological patch extraction method
CN111986099A (en) Tillage monitoring method and system based on convolutional neural network with residual error correction fused
CN106778605A (en) Remote sensing image road net extraction method under navigation data auxiliary
CN113223042B (en) Intelligent acquisition method and equipment for remote sensing image deep learning sample
Hu et al. Estimating impervious surfaces from medium spatial resolution imagery: A comparison between fuzzy classification and LSMA
Ghanea et al. Automatic building extraction in dense urban areas through GeoEye multispectral imagery
CN104282026A (en) Distribution uniformity assessment method based on watershed algorithm and minimum spanning tree
Liu et al. Domain adaptation for land use classification: A spatio-temporal knowledge reusing method
CN114463623A (en) Method and device for detecting farmland change based on multi-scale remote sensing image
CN116822980A (en) Base vector guided typical earth surface element intelligent monitoring method and system
CN115858840B (en) Remote sensing image mosaic method based on scene
Zhang et al. Impervious surface extraction from high-resolution satellite image using pixel-and object-based hybrid analysis
Xia et al. Refined extraction of buildings with the semantic edge-assisted approach from very high-resolution remotely sensed imagery
Doustfatemeh et al. Comprehensive urban area extraction from multispectral medium spatial resolution remote-sensing imagery based on a novel structural feature
CN101887522A (en) Method for extracting remote sensing image road based on multipoint simultaneous and rapid advance
Erharter et al. Machine learning–an approach for consistent rock glacier mapping and inventorying–example of Austria
Albert et al. Land use classification using conditional random fields for the verification of geospatial databases

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
CB02 Change of applicant information

Address after: 610000 No. 270, floor 2, No. 8, Jinxiu street, Wuhou District, Chengdu, Sichuan

Applicant after: Chengdu shuzhilian Technology Co.,Ltd.

Address before: 610000 No.2, 4th floor, building 1, Jule Road intersection, West 1st section of 1st ring road, Wuhou District, Chengdu City, Sichuan Province

Applicant before: CHENGDU SHUZHILIAN TECHNOLOGY Co.,Ltd.

CB02 Change of applicant information