CN113128388A - Optical remote sensing image change detection method based on space-time spectrum characteristics - Google Patents
Optical remote sensing image change detection method based on space-time spectrum characteristics Download PDFInfo
- Publication number
- CN113128388A CN113128388A CN202110399135.XA CN202110399135A CN113128388A CN 113128388 A CN113128388 A CN 113128388A CN 202110399135 A CN202110399135 A CN 202110399135A CN 113128388 A CN113128388 A CN 113128388A
- Authority
- CN
- China
- Prior art keywords
- remote sensing
- sensing image
- change
- features
- space
- 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.)
- Granted
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
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/40—Analysis of texture
- G06T7/41—Analysis of texture based on statistical description of texture
- G06T7/45—Analysis of texture based on statistical description of texture using co-occurrence matrix computation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
-
- 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
- G06T2207/10036—Multispectral image; Hyperspectral image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/58—Extraction of image or video features relating to hyperspectral data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/194—Terrestrial scenes using hyperspectral data, i.e. more or other wavelengths than RGB
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Probability & Statistics with Applications (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to a method for detecting changes of optical remote sensing images based on space-time spectral characteristics, which belongs to the technical field of optical remote sensing image processing and comprises the steps of firstly obtaining a first mask of a change point and an invariant point by utilizing a CCDC algorithm; judging each pixel point by utilizing a time sequence fitting curve, and screening out a change point and an invariant point to obtain a second mask; taking intersection of the two masks to obtain a training sample point with high reliability; extracting features from the time series remote sensing image data set to construct a stacking feature matrix; and then training the selected training samples by using machine learning to obtain a classification model, and predicting all pixel points to obtain a final change detection result. The invention introduces spectral characteristics, texture characteristics and statistical characteristics to carry out change detection, and the algorithm can automatically generate a training sample label, effectively utilizes spatial information among pixels and is beneficial to improving the accuracy of remote sensing change detection.
Description
Technical Field
The invention relates to the technical field of remote sensing image processing, in particular to an optical remote sensing image change detection method based on space-time spectrum characteristics.
Background
Methods for change detection can be divided into two categories, namely, two-phase change detection and time series change detection. The two-phase change detection method uses only two images that must be sharp (i.e., free of clouds and cloud shadows) and collected at nearly the same time during the year. This approach is relatively simple to implement, but is not always applicable, since it may take years to collect the ideal two images, especially in cloudy and rainy areas, whereas time-series change detection utilizes all available images for change detection, even if a portion of the images are covered by clouds and shadows. It requires much less input data and can provide more comprehensive information about the changes in the terrain. The time series approach may require significant storage and computational costs.
The invention with the reference application number of CN201910789051.X discloses a target library generation system suitable for multi-source heterogeneous remote sensing data, which comprises: typical object retrieval function: performing object-level target detection on standard product data by adopting an object-level target detection method based on deep learning to obtain a labeling file of an object-level target; advanced product production function module: performing image processing operation on the standard product data corresponding to the object-level target according to the label file of the object-level target to obtain an object-level product containing the object-level target; the sample library production function module: and identifying a target level target in the target level product by adopting a target level target detection method based on deep learning to obtain the target level product. The invention realizes the functions of typical target retrieval, high-grade product production, sample library production and the like by data mining and image processing of the remote sensing image, greatly improves the typical target retrieval speed of mass remote sensing data, and realizes the automatic production of high-grade products.
However, when the training sample data set is produced, the above patent needs manual production and interpretation, so that the efficiency of the above change detection method in the application program is low; in addition, the above patent does not consider spatial information and correlation between pixels in the detection process, so that the accuracy cannot be improved.
Disclosure of Invention
The invention aims to provide a method for detecting changes of optical remote sensing images based on space-time spectral characteristics, which aims to solve the technical problems that a training sample data set cannot be automatically made in the prior art and spatial information and mutual correlation among pixels are not considered in the detection process.
The technical scheme of the invention is as follows: the method comprises the following steps:
step 1: extracting spectral features, textural features and statistical features from the time series remote sensing image data set;
step 2: stacking all the features extracted in the step (1) to construct a high-dimensional stacked feature matrix, wherein the stacked feature matrix is composed of normalized spatio-temporal spectral information of the remote sensing image;
and step 3: obtaining a first mask of a variable point and a constant point by using a CCDC algorithm;
and 4, step 4: fitting the time sequences of all the pixel points by utilizing a formula by utilizing an OLS (least squares method), and performing threshold operation on residual errors calculated by utilizing a time sequence model to generate a second mask of a variable point and a non-variable point;
and 5: taking an intersection of the two masks in the step 3 and the step 4 to obtain a training sample with high credibility;
step 6: putting the characteristics of the training samples into an SVM classifier for classification to obtain a training model;
and 7: and predicting all the pixel points by utilizing the SVM according to the training model to obtain a final change detection result.
As a further optimization of the technical solution, the band information of the remote sensing image is extracted as a spectral feature, and the spectral feature includes but is not limited to the following bands: blue (Blue), Green (Green), Red (Red), NIR (near infrared), SWIR1 (short wave Infrared 1), SWIR2 (short wave Infrared 2) and TIR (thermal Infrared)
As further optimization of the technical scheme, texture features can be extracted from the remote sensing image through the gray level co-occurrence matrix, and the feature values of the gray level co-occurrence matrix are obtained by taking the mean value in four directions for calculation.
As a further optimization of the technical scheme, the four directions are respectively 0 degree, 45 degrees, 90 degrees and 135 degrees; and for each remote sensing image, calculating the texture characteristic value of the central pixel point of the window in a sliding window mode. The stride is set to 1 pixel and the sliding window size is set to 3 pixels.
As a further optimization of the technical scheme, the statistical characteristics consist of local means and variances;
where n is the window size, V (i, j) is the spectral value in the ith row and jth column in the moving window, and n is set to 3.
As a further optimization of the technical scheme, the normalized formula is
Wherein PV and PV' are the values of each feature before and after normalization, respectively; PVmax and PVmin are the respective maximum and minimum values of each feature before normalization, respectively.
As a further optimization of the technical scheme, the formula is
Wherein x is julian date, i is ith spectral band, T is number of days per year (T is 365), a (0, i) is coefficient of general variation trend, a (1, i) and b (1, i) are coefficients of annual variation, c (1, i) is coefficient of annual variation, and the fitted coefficients and real time series values are calculated to obtain root mean square error.
As a further optimization of the present technical solution, the CCDC algorithm is used to roughly classify images into two categories: changed and unchanged pixels.
The invention provides an optical remote sensing image change detection method based on space-time spectral characteristics through improvement, and compared with the prior art, the method has the following improvements and advantages:
firstly, the invention introduces spatial features (namely texture features and statistical features) to carry out change detection, effectively utilizes spatial information among pixels, and is beneficial to improving the accuracy of remote sensing change detection, so that the change detection result is more accurate, and scattered misjudgment points are fewer. The method and the device solve the technical problem that in the prior art, spatial information and mutual correlation among pixels are not considered in the detection process, so that the accuracy cannot be improved.
Secondly, the CCDC algorithm can obtain a mask of a change point and an invariant point, a formula is used for fitting time sequences of all pixel points by using an OLS (least squares) method to obtain a second mask, and the two masks are intersected to obtain a training sample with high reliability; the characteristics of the training samples are put into an SVM classifier for classification to obtain a training model, and then all pixel points are predicted to obtain a final change detection result. By automatically generating these labels through the above algorithm, the time taken to mark the labels can be greatly reduced.
Drawings
The invention is further explained below with reference to the figures and examples:
FIG. 1 is a flow chart of the CCDC algorithm;
FIG. 2 is a flow chart of the algorithm of the present invention.
Detailed Description
The present invention will be described in detail with reference to fig. 1 to 2, and the technical solutions in the embodiments of the present invention will be clearly and completely described, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides a method for detecting changes of optical remote sensing images based on space-time spectral characteristics through improvement, as shown in figures 1-2, the method comprises the following steps:
step 1: extracting spectral features, textural features and statistical features from the time series remote sensing image data set;
step 2: stacking all the features extracted in the step (1) together to construct a high-dimensional stacked feature matrix, wherein the stacked feature matrix is composed of normalized spatio-temporal spectral information of the remote sensing image;
and step 3: obtaining a first mask of a variable point and a constant point by using a CCDC algorithm;
and 4, step 4: fitting the time sequences of all the pixel points by utilizing a formula by utilizing an OLS (least squares method), and performing threshold operation on residual errors calculated by utilizing a time sequence model to generate a second mask of a variable point and a non-variable point;
specifically, the formula is
Wherein x is julian date, i is ith spectral band, T is number of days per year (T is 365), a (0, i) is coefficient of general variation trend, a (1, i) and b (1, i) are coefficients of annual variation, c (1, i) is coefficient of annual variation, and the fitted coefficients and real time series values are calculated to obtain root mean square error.
And 5: taking an intersection of the two masks in the step 3 and the step 4 to obtain a training sample with high credibility;
step 6: and (5) training the training samples selected in the step (5) by using an SVM classifier to obtain a training model. In this patent, the optimal parameters of the SVM classifier are adjusted through cross validation, the SVM is a support vector machine, which is the prior art and is not described herein any more, and the SVM classifier can also be replaced by other machine learning classifiers, such as a Random forest (Random forest) classifier or other well-known classifiers, and is not described herein in detail.
And 7: and predicting all the pixel points by utilizing the SVM according to the training model to obtain a final change detection result.
As shown in fig. 1, fig. 1 is a flow chart of a CCDC algorithm, which is a time-series model generated for each pixel point based on time-series pixel values of satellite images, and has seasonal, trend and interruption components, can capture all three categories of surface changes by automatically generating thresholds, and can continuously detect various land cover changes as new images are collected, providing a land cover change map at any given time.
Specifically, in step 1, the features include spectral features, texture features, and statistical features. In this patent, spectral, textural and statistical features of time series images are extracted.
In spectral feature extraction, 7 bands, Blue, Green, Red, NIR (near infrared), SWIR1 (short wave infrared 1), SWIR2 (short wave infrared 2) and TIR (temperature reflectance) are used in this patent.
In the texture feature extraction, the texture features can be extracted from the remote sensing image through a gray level co-occurrence matrix, and the feature values of the gray level co-occurrence matrix are obtained by taking the mean value in four directions for calculation. The gray level co-occurrence matrix (GLCM) can effectively extract texture features from the remote sensing image. It has been widely used to characterize land cover types. In this patent, GLCM-based texture features are constructed in eight categories, mean, variance, contrast, distance, second moment, homogeneity, and cluster shade, respectively. Notably, GLCM eigenvalues are calculated by averaging over four directions (0 °, 45 °, 90 ° and 135 °). And for each remote sensing image, calculating the texture characteristic value of the central pixel point of the window in a sliding window mode. The stride is set to 1 pixel and the sliding window size is set to 3 pixels.
Specifically, the four directions are respectively 0 °, 45 °, 90 ° and 135 °; and for each remote sensing image, calculating the texture characteristic value of the central pixel point of the window in a sliding window mode. The stride is set to 1 pixel and the sliding window size is set to 3 pixels.
Specifically, the statistical features consist of local means and variances;
where n is the window size, V (i, j) is the spectral value in the ith row and jth column in the moving window, and n is set to 3. Variations in the type of ground cover typically involve variations in strength and area detail. In this patent, statistical features (i.e., local means and variances) are used to describe the intensity and detail information of the remotely sensed image. It is noted that the mean and variance of the statistical features are calculated using surface reflectivity or temperature, while the mean and variance in the texture features are calculated using the GLCM matrix for feature values.
Specifically, since the values of the spectral feature, the texture feature and the statistical feature are in different orders of magnitude, in order to reduce the influence of the values in the classification, normalization processing needs to be performed on the spectral feature, the texture feature and the statistical feature, and the normalization formula is
Wherein PV and PV' are the values of each feature before and after normalization, respectively; PVmax and PVmin are the respective maximum and minimum values of each feature before normalization, respectively.
Specifically, the CCDC algorithm is used to roughly classify images into two categories: changed and unchanged pixels.
The invention introduces spatial features (namely texture features and statistical features) to carry out change detection, effectively utilizes spatial information among pixels, and is beneficial to improving the accuracy of remote sensing change detection, so that the change detection result is more accurate, and scattered misjudgment points are fewer.
Under the general condition, the fitting curve of the changed pixel points has larger difference with the real time sequence, the RMSE value is larger, similarly, the RMSE of the unchanged pixel points is smaller, the changed and unchanged pixel points are judged according to the RMSE size, each pixel point finally has a corresponding RMSE value, the RMSE values are calculated by an Otsu method, a threshold value is automatically generated, the pixel points which are larger than the threshold value are considered to be changed points, the pixel points which are smaller than the threshold value are considered to be unchanged points, and as seven spectral bands are used, each single band can obtain a change detection result by using the judgment method, so that the total number of the change detection results is 7; different land changes may behave differently in different wavebands, i.e. one land change does not change significantly in one waveband but changes significantly in another waveband; the method and the device overlap change detection results obtained by 7 wave bands respectively (namely, the pixel point is considered to be changed as long as the pixel point is judged to be changed under one wave band), and finally the pixel point is changed into a mask of a change point and an invariant point.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. A method for detecting changes of optical remote sensing images based on space-time spectral characteristics is characterized by comprising the following steps:
step 1: extracting spectral features, textural features and statistical features from the time series remote sensing image data set;
step 2: stacking all the features extracted in the step (1) to construct a high-dimensional stacked feature matrix, wherein the stacked feature matrix is composed of normalized spatio-temporal spectral information of the remote sensing image;
and step 3: obtaining a first mask of a variable point and a constant point by using a CCDC algorithm;
and 4, step 4: fitting the time sequences of all the pixel points by utilizing a formula by utilizing an OLS (on-line analytical system), and performing threshold operation on residual errors calculated by utilizing a time sequence model to generate a second mask of the variable points and the invariable points;
and 5: taking an intersection of the two masks in the step 3 and the step 4 to obtain a training sample with high credibility;
step 6: putting the characteristics of the training samples into an SVM classifier for classification to obtain a training model;
and 7: and predicting all the pixel points by utilizing the SVM according to the training model to obtain a final change detection result.
2. The method for detecting the change of the optical remote sensing image based on the space-time spectral features as claimed in claim 1, wherein: extracting waveband information of the remote sensing image as spectral characteristics, wherein the spectral characteristics comprise but are not limited to the following wavebands: blue, Green, Red, NIR, SWIR1, SWIR2, and TIR.
3. The method for detecting the change of the optical remote sensing image based on the space-time spectral features as claimed in claim 1, wherein: texture features can be extracted from the remote sensing image through the gray level co-occurrence matrix, and the feature values of the gray level co-occurrence matrix are obtained by taking the mean value in four directions for calculation.
4. The method for detecting the change of the optical remote sensing image based on the space-time spectral features as claimed in claim 3, characterized in that: the four directions are respectively 0 degree, 45 degrees, 90 degrees and 135 degrees; for each remote sensing image, the texture characteristic value of a central pixel point of a window is calculated in a sliding window mode, the step pitch is set to be 1 pixel, and the size of the sliding window is set to be 3 pixels.
5. The method for detecting the change of the optical remote sensing image based on the space-time spectral features as claimed in claim 3, characterized in that: the statistical features consist of local means and variances;
where n is the window size, V (i, j) is the spectral value in the ith row and jth column in the moving window, and n is set to 3.
6. The method for detecting the change of the optical remote sensing image based on the space-time spectral characteristics as claimed in claim 1 or 4, wherein: the normalized formula is
Wherein PV and PV' are the values of each feature before and after normalization, respectively; PVmax and PVmin are the respective maximum and minimum values of each feature before normalization, respectively.
7. The method for detecting the change of the optical remote sensing image based on the space-time spectral features as claimed in claim 1, wherein: the formula is
Wherein x is julian date, i is ith spectrum band, T is number of days per year, a is coefficient of overall change trend, a and b are coefficients of annual change, c is coefficient of annual change, and the fitted coefficients and real time sequence values are calculated to obtain root mean square error.
8. The method for detecting the change of the optical remote sensing image based on the space-time spectral features as claimed in claim 1, wherein: using the CCDC algorithm, images are roughly classified into two categories: a changed pixel and a non-changed pixel.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110399135.XA CN113128388B (en) | 2021-04-14 | 2021-04-14 | Optical remote sensing image change detection method based on space-time spectrum characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110399135.XA CN113128388B (en) | 2021-04-14 | 2021-04-14 | Optical remote sensing image change detection method based on space-time spectrum characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113128388A true CN113128388A (en) | 2021-07-16 |
| CN113128388B CN113128388B (en) | 2022-09-02 |
Family
ID=76776227
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110399135.XA Active CN113128388B (en) | 2021-04-14 | 2021-04-14 | Optical remote sensing image change detection method based on space-time spectrum characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113128388B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114120137A (en) * | 2021-10-19 | 2022-03-01 | 桂林理工大学 | Wetland element space-time evolution monitoring method based on time sequence main remote sensing image |
| CN115841435A (en) * | 2023-02-24 | 2023-03-24 | 湖南大学 | Remote sensing image banding removing method based on sorting domain direction gradient histogram constraint |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110170768A1 (en) * | 2010-01-11 | 2011-07-14 | Tandent Vision Science, Inc. | Image segregation system with method for handling textures |
| CN103500450A (en) * | 2013-09-30 | 2014-01-08 | 河海大学 | Multi-spectrum remote sensing image change detection method |
| US20160307073A1 (en) * | 2015-04-20 | 2016-10-20 | Los Alamos National Security, Llc | Change detection and change monitoring of natural and man-made features in multispectral and hyperspectral satellite imagery |
| CN109934154A (en) * | 2019-03-08 | 2019-06-25 | 北京科技大学 | A kind of remote sensing image variation detection method and detection device |
| CN110443822A (en) * | 2019-07-16 | 2019-11-12 | 浙江工业大学 | A kind of high score remote sensing target fine extracting method of semanteme edge auxiliary |
| CN110472184A (en) * | 2019-08-22 | 2019-11-19 | 电子科技大学 | A kind of cloudy misty rain area rice recognition methods based on Landsat remotely-sensed data |
| CN110717531A (en) * | 2019-09-27 | 2020-01-21 | 华东师范大学 | Method for detecting classified change type based on uncertainty analysis and Bayesian fusion |
| CN111640159A (en) * | 2020-05-11 | 2020-09-08 | 武汉大学 | Remote sensing image change detection method based on twin convolutional neural network |
| CN112560826A (en) * | 2021-02-24 | 2021-03-26 | 航天宏图信息技术股份有限公司 | Training method of image classification model, and image classification method and device |
-
2021
- 2021-04-14 CN CN202110399135.XA patent/CN113128388B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110170768A1 (en) * | 2010-01-11 | 2011-07-14 | Tandent Vision Science, Inc. | Image segregation system with method for handling textures |
| CN103500450A (en) * | 2013-09-30 | 2014-01-08 | 河海大学 | Multi-spectrum remote sensing image change detection method |
| US20160307073A1 (en) * | 2015-04-20 | 2016-10-20 | Los Alamos National Security, Llc | Change detection and change monitoring of natural and man-made features in multispectral and hyperspectral satellite imagery |
| CN109934154A (en) * | 2019-03-08 | 2019-06-25 | 北京科技大学 | A kind of remote sensing image variation detection method and detection device |
| CN110443822A (en) * | 2019-07-16 | 2019-11-12 | 浙江工业大学 | A kind of high score remote sensing target fine extracting method of semanteme edge auxiliary |
| CN110472184A (en) * | 2019-08-22 | 2019-11-19 | 电子科技大学 | A kind of cloudy misty rain area rice recognition methods based on Landsat remotely-sensed data |
| CN110717531A (en) * | 2019-09-27 | 2020-01-21 | 华东师范大学 | Method for detecting classified change type based on uncertainty analysis and Bayesian fusion |
| CN111640159A (en) * | 2020-05-11 | 2020-09-08 | 武汉大学 | Remote sensing image change detection method based on twin convolutional neural network |
| CN112560826A (en) * | 2021-02-24 | 2021-03-26 | 航天宏图信息技术股份有限公司 | Training method of image classification model, and image classification method and device |
Non-Patent Citations (5)
| Title |
|---|
| CHENGBIN DENG ET AL.: "Continuous subpixel monitoring of urban impervious surface using Landsat time series", 《ELSEVIER》 * |
| JINING YAN ET AL.: "A time-series classification approach based on change detection for rapid land cover mapping", 《ELSEVIER》 * |
| YUNFENG HU ET AL.: "An automatic approach for land-change detection and land updates based on integrated NDVI timing analysis and the sGVAPS method with GEE support", 《ELSEVIER》 * |
| ZHE ZHU: "Continuous change detection and classification of land cover using all available Landsat data", 《ELSEVIER》 * |
| 林蕾: "基于循环神经网络模型的遥感影像时间序列分类及变化检测方法研究", 《中国博士学位论文全文数据库工程科技Ⅱ辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114120137A (en) * | 2021-10-19 | 2022-03-01 | 桂林理工大学 | Wetland element space-time evolution monitoring method based on time sequence main remote sensing image |
| CN115841435A (en) * | 2023-02-24 | 2023-03-24 | 湖南大学 | Remote sensing image banding removing method based on sorting domain direction gradient histogram constraint |
| CN115841435B (en) * | 2023-02-24 | 2023-04-25 | 湖南大学 | Remote sensing image strip removing method based on ordering domain direction gradient histogram constraint |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113128388B (en) | 2022-09-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Tan et al. | Recognising blueberry fruit of different maturity using histogram oriented gradients and colour features in outdoor scenes | |
| Yu et al. | A litchi fruit recognition method in a natural environment using RGB-D images | |
| Wang et al. | A review on weed detection using ground-based machine vision and image processing techniques | |
| Majeed et al. | Deep learning based segmentation for automated training of apple trees on trellis wires | |
| Liu et al. | A detection method for apple fruits based on color and shape features | |
| Lin et al. | In-field citrus detection and localisation based on RGB-D image analysis | |
| Qureshi et al. | Machine vision for counting fruit on mango tree canopies | |
| Castillo-Martínez et al. | Color index based thresholding method for background and foreground segmentation of plant images | |
| Kale et al. | A research review on hyperspectral data processing and analysis algorithms | |
| CN113128388B (en) | Optical remote sensing image change detection method based on space-time spectrum characteristics | |
| CN112101271A (en) | Hyperspectral remote sensing image classification method and device | |
| Liu et al. | Automatic segmentation of overlapped poplar seedling leaves combining Mask R-CNN and DBSCAN | |
| Miao et al. | Classification of farmland images based on color features | |
| CN112990314B (en) | Hyperspectral image anomaly detection method and device based on improved isolated forest algorithm | |
| CN111007013B (en) | Crop rotation fallow remote sensing monitoring method and device for northeast cold region | |
| Liu et al. | Small unopened cotton boll counting by detection with MRF-YOLO in the wild | |
| CN115410096B (en) | Satellite remote sensing image multi-scale fusion change detection method, medium and electronic device | |
| CN110705449A (en) | Land utilization change remote sensing monitoring analysis method | |
| CN103366373B (en) | Multi-time-phase remote-sensing image change detection method based on fuzzy compatible chart | |
| Bawa et al. | A support vector machine and image processing based approach for counting open cotton bolls and estimating lint yield from UAV imagery | |
| CN110135479A (en) | The high spectrum image object detection method and system of study are estimated based on random forest | |
| CN108960005B (en) | Method and system for establishing and displaying object visual label in intelligent visual Internet of things | |
| Khanday et al. | Change detection in hyper spectral images | |
| Ma et al. | Spectral difference analysis and identification of different maturity blueberry fruit based on hyperspectral imaging using spectral index | |
| da Silva et al. | Multispectral image classification using multilayer perceptron and principal components analysis |
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 |