CN114639014A - NDVI normalization method based on high-resolution remote sensing image - Google Patents

NDVI normalization method based on high-resolution remote sensing image Download PDF

Info

Publication number
CN114639014A
CN114639014A CN202210151368.2A CN202210151368A CN114639014A CN 114639014 A CN114639014 A CN 114639014A CN 202210151368 A CN202210151368 A CN 202210151368A CN 114639014 A CN114639014 A CN 114639014A
Authority
CN
China
Prior art keywords
ndvi
resolution
data
remote sensing
sensing image
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
CN202210151368.2A
Other languages
Chinese (zh)
Inventor
黄文丽
陶雨婷
沈焕锋
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.)
Wuhan University WHU
Original Assignee
Wuhan University WHU
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 Wuhan University WHU filed Critical Wuhan University WHU
Priority to CN202210151368.2A priority Critical patent/CN114639014A/en
Publication of CN114639014A publication Critical patent/CN114639014A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/253—Fusion techniques of extracted features

Abstract

The invention relates to an NDVI (normalized difference vegetation index) normalization method based on a high-resolution remote sensing image. Firstly, a series of preprocessing is carried out on a remote sensing image, then high-resolution to-be-normalized reflectivity data is down-sampled to the resolution of medium-resolution reference data, coefficients of a linear regression equation are obtained through solving according to a medium-resolution sample point pair, preliminary normalization results of all pixels of the to-be-normalized data are obtained through pixel-by-pixel calculation through linear relations, and finally the preliminary normalization results in the same region and the same time period are subjected to maximum value synthesis to obtain a final normalization result. The invention comprehensively utilizes the multi-source remote sensing data, eliminates the difference of sensors, imaging conditions and the like in the multi-source remote sensing data as much as possible, better solves the problem of obvious splicing seams when a large-range multi-scene image is embedded, has less input parameters required by the whole algorithm and higher operation efficiency, and can produce large-range high-resolution NDVI products.

Description

NDVI normalization method based on high-resolution remote sensing image
Technical Field
The invention belongs to the technical field of remote sensing image processing, and particularly relates to an NDVI (normalized difference of variance) normalization method based on a high-resolution remote sensing image.
Background
The vegetation plays an important role in the ecological system, is closely related to natural environment factors such as soil, terrain, climate, hydrology and the like, has a profound influence on the energy balance of the ground gas system, and is deeply concerned by global scientists and governments of various countries for a long time. As one of the most effective means for global vegetation monitoring at present, satellite remote sensing can be free from the constraint of social and natural conditions, quickly acquire wide-range observation data and provide conditions for researching and monitoring the growth change of global or regional vegetation. Based on the spectral characteristics of the vegetation, linear or nonlinear combined operation is carried out on the visible light and the near infrared wave bands of the remote sensing data, and various vegetation indexes are generated. The Normalized Difference Vegetation Index (NDVI) is simple and convenient to calculate, can eliminate most of influences related to radiation calibration of the sensor, terrain, atmospheric conditions and observation angles, enhances sensitivity to vegetation, and becomes the vegetation index which is the most widely applied nowadays.
The vegetation index is one of important parameters for describing the vegetation cover characteristics of the earth surface, is a simple, effective and empirical earth surface vegetation condition measurement index, and is widely applied to qualitative and quantitative evaluation of vegetation cover and growth condition thereof. However, the existing vegetation index products mostly use single remote sensing data, and the products have certain defects in the aspects of time resolution, spatial resolution, precision and stability, and the defects can be made up to a certain extent by comprehensively utilizing the multi-source remote sensing data. Since there are differences in sensors, imaging conditions, etc. in multi-source remote sensing data, it is necessary to normalize it. The radiation normalization is divided into absolute radiation normalization and relative radiation normalization, and when the multi-temporal data of a single sensor is normalized, the simple relative radiation normalization can achieve higher precision; however, in quantitative application, absolute radiation normalization and relative radiation normalization are combined, the influence of reflection difference of different sensors caused by spectral response characteristics and atmospheric correction on normalization results is considered, multi-source data normalization is performed on the surface reflectivity level, and normalization of different images of different sensors is really realized.
Therefore, absolute radiation normalization and relative radiation normalization need to be combined, and medium-high resolution multi-source remote sensing data are comprehensively utilized for normalization so as to construct a high-resolution quantitative product data set and improve the dynamic monitoring level of urban vegetation resources.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides an NDVI normalization method based on a high-resolution remote sensing image. By comprehensively utilizing the multi-source remote sensing data, the differences of sensors, imaging conditions and the like in the multi-source remote sensing data are eliminated as much as possible, and the problem that splicing seams are obvious when large-range multi-scene images are embedded is solved well; the whole algorithm needs fewer input parameters and has higher operation efficiency, and the large-scale high-resolution NDVI product production can be carried out.
In order to achieve the above object, the technical solution provided by the present invention is an NDVI normalization method based on a high resolution remote sensing image, comprising the steps of:
step 1, preprocessing medium and high resolution remote sensing image data;
step 1.1, carrying out radiometric calibration, FLAASH atmospheric correction and orthographic correction with a reference image on the medium-resolution remote sensing image, and then carrying out combined calculation by utilizing a red light wave band and a near infrared wave band to obtain NDVI data used for reference;
step 1.2, after orthorectification, radiometric calibration and rapid atmospheric correction are carried out on the high-resolution remote sensing image, the NDVI data to be normalized is obtained by utilizing combined calculation of a red light wave band and a near infrared wave band;
step 1.3, applying IsoData Classification unsupervised Classification to the high-resolution remote sensing image preprocessed in the step 1.2 to obtain Classification data corresponding to NDVI data to be normalized;
step 1.4, applying a multi-scale convolution feature fused deep learning cloud and shadow detection method MSCFF to the high-resolution remote sensing image preprocessed in the step 1.2 to obtain cloud mask data corresponding to data to be normalized;
step 2, obtaining a preliminary normalization result by adopting global linear fitting normalization;
step 2.1, after removing the cloud, the shadow and the background through the classification data obtained in the step 1.3 and the cloud mask obtained in the step 1.4, the high-resolution NDVI data to be normalized is down-sampled to the medium resolution of the reference data;
step 2.2, according to the high-resolution remote sensing image classification data obtained in the step 1.3, calculating the purity of each pixel of the medium-resolution remote sensing image, and screening out the pixels with the purity smaller than a set threshold value T as pure pixels;
step 2.3, based on the pure pixels screened out in the step 2.2, solving linear relation coefficients of the NDVI data to be normalized and the reference NDVI data by utilizing Huber type M estimation;
step 2.4, obtaining a normalization value of the high-resolution NDVI data to be normalized by using the coefficient of the linear relation obtained in the step 2.3;
step 3, obtaining a final normalization result by adopting a maximum synthesis method;
step 3.1, carrying out geometric registration on the primary normalization results of the images in different time in the same region;
and 3.2, after the primary normalization results in the same region and the same time period are superposed, selecting the maximum value at each pixel as the value of the final result.
Furthermore, in step 1.1, the NDVI data is calculated as follows:
Figure BDA0003506733950000031
in the formula, ρNIRFor near infrared band data, pREDIs data of red light wave band.
Moreover, the purity calculation formula of the pixels in the step 2.2 is as follows:
Figure BDA0003506733950000032
in the formula, r is the pixel purity; k is a radical ofcThe number of pixels belonging to the class c in the high-resolution classification data in any medium-resolution pixel range is, and c is the ground object class with the largest proportion in any medium-resolution pixel range; and m is the ratio of the medium resolution to the high resolution of the remote sensing image, and m multiplied by m is the number of high resolution pixels in the pixel range of the medium resolution remote sensing image.
Furthermore, the linear relationship between the NDVI data to be normalized and the reference NDVI data in step 2.3 is as follows:
yn＝a×xn+b (3)
in the formula, xnFor the medium resolution of the pure pixels down-sampled in step 2.1 NDVI data to be normalized, ynAnd (3) referring to NDVI data for the medium resolution of the pure pixel obtained in the step 1.1, wherein a and b are coefficients of a linear relation.
A more robust Huber type M estimation solution is used:
euv＝a×xuv+b-yuv (4)
by minimizing
Figure BDA0003506733950000033
To solve for coefficients, where ρ () is a Huber type influence function:
Figure BDA0003506733950000034
wherein c is the Huber parameter.
Moreover, in step 2.4, the normalized value calculation formula of the high-resolution NDVI data to be normalized is as follows:
y′m＝a×xm+b (6)
in the formula, xmThe m pixel, y 'in the high-resolution data to be normalized obtained in the step 1.2'mAnd a and b are linear relation coefficients obtained by solving in the step 2.3 for the corresponding high-resolution normalization result values.
Furthermore, the maximum synthesis method in step 3.2 has the following calculation formula:
NDVImax＝max(NDVI1,NDVI2,...,NDVIn) (7)
in the formula, NDVImaxIs the maximum value of NDVI in this time period, NDVI1，NDVI2，NDVInA plurality of NDVI values for different periods of time within the time period.
Compared with the prior art, the invention has the following advantages:
1) the medium and high resolution multi-source remote sensing data are comprehensively utilized, and the defects of the existing vegetation index product in the aspects of time resolution, spatial resolution, precision and stability are made up to a certain extent;
2) differences of sensors, imaging conditions and the like of multi-source remote sensing data are eliminated as much as possible, and the problem that splicing seams are obvious when large-range multi-scene images are embedded is solved well;
3) the absolute radiation normalization and the relative radiation normalization are combined, the influence of reflection difference of different sensors caused by spectral response characteristics and atmospheric correction on normalization results is considered, the normalization among different images of different sensors is really realized, and the quantitative application requirements are met;
4) the algorithm needs fewer input parameters and has higher operation efficiency, and is suitable for the production of large-range high-resolution NDVI products.
Drawings
FIG. 1 is a flow chart of an embodiment of the present invention.
Fig. 2 is a comparison graph before and after normalization of a multi-scene mosaic, in which fig. 2(a) is a mosaic result graph before normalization, fig. 2(b) is standard data, and fig. 2(c) is a mosaic result graph before and after normalization.
Detailed Description
The invention provides an NDVI normalization method based on a high-resolution remote sensing image, and the technical scheme of the invention is further explained by combining the accompanying drawings and an embodiment.
As shown in fig. 1, the process of the embodiment of the present invention includes the following steps:
step 1, preprocessing the intermediate and high resolution remote sensing image data, which comprises the following substeps:
step 1.1, carrying out radiometric calibration, FLAASH atmospheric correction and orthorectification with reference images on the medium-resolution remote sensing image, and then carrying out combined calculation by using a red light wave band and a near infrared wave band to obtain NDVI data used for reference, wherein the calculation formula is as follows:
Figure BDA0003506733950000051
in the formula, ρNIRFor near infrared band data, pREDIs data of red light wave band.
And step 1.2, after orthorectification, radiometric calibration and rapid atmospheric correction are carried out on the high-resolution remote sensing image, calculating according to the formula (1) to obtain NDVI data to be normalized.
And step 1.3, applying IsoData Classification unsupervised Classification to the high-resolution remote sensing image preprocessed in the step 1.2 to obtain Classification data corresponding to the NDVI data to be normalized.
And step 1.4, applying a multi-scale convolution feature fused deep learning cloud and shadow detection method MSCFF to the high-resolution remote sensing image preprocessed in the step 1.2 to obtain cloud mask data corresponding to the data to be normalized.
Step 2, obtaining a preliminary normalization result by adopting global linear fitting normalization, wherein the preliminary normalization result comprises the following substeps:
and 2.1, removing the cloud, the shadow and the background through the classification data obtained in the step 1.3 and the cloud mask obtained in the step 1.4, and then down-sampling the high-resolution NDVI data to be normalized to the medium resolution of the reference data.
And 2.2, calculating the purity of each pixel of the medium-resolution remote sensing image according to the high-resolution remote sensing image classification data obtained in the step 1.3, and screening out the pixels with the purity smaller than a set threshold value T as pure pixels.
The pixel purity calculation formula is as follows:
Figure BDA0003506733950000052
in the formula, r is the pixel purity; k is a radical ofcThe number of pixels belonging to the class c in the high-resolution classification data in any medium-resolution pixel range is, and c is the ground object class with the largest proportion in any medium-resolution pixel range; and m is the ratio of the medium resolution to the high resolution of the remote sensing image, and m multiplied by m is the number of high resolution pixels in the pixel range of the medium resolution remote sensing image.
And 2.3, based on the pure pixels screened out in the step 2.2, solving the coefficient of the linear relation between the NDVI data to be normalized and the reference NDVI data by utilizing Huber type M estimation.
The linear relationship between the NDVI data to be normalized and the reference NDVI data is as follows:
yn＝a×xn+b (3)
in the formula, xnNDVI data to be normalized for medium resolution of the pure pixels down-sampled in step 2.1nAnd (2) referring to NDVI data for the medium resolution of the pure pixel obtained in the step 1.1, wherein a and b are coefficients of a linear relation.
Because operations such as geometric correction, resampling, unsupervised classification and the like have errors, noise and abnormal values are brought certainly, and the least square method is very easily influenced by the noise and the abnormal values, so that the coefficient solving has errors, and a more robust Huber type M estimation solving is used:
euv＝a×xuv+b-yuv (4)
by minimizing
Figure BDA0003506733950000061
To solve for coefficients where ρ () is the impact function, a Huber-type impact function is chosen in this embodiment:
Figure BDA0003506733950000062
where c is the Huber parameter, in this example c is 1.345.
And 2.4, obtaining a normalization value of the high-resolution NDVI data to be normalized by using the coefficient of the linear relation obtained in the step 2.3.
The calculation formula is as follows:
y′m＝a×xm+b (6)
in the formula, xmThe m pixel, y 'in the high-resolution data to be normalized obtained in the step 1.2'mAnd a and b are linear relation coefficients obtained by solving in the step 2.3 for the corresponding high-resolution normalization result values.
And 3, obtaining a final normalization result by adopting a maximum synthesis method.
And 3.1, performing geometric registration on the primary normalization results of the images in different time in the same region.
Step 3.2, after the preliminary normalization results in the same area and the same time period are superposed, the maximum value is selected at each pixel as the value of the final result, and the maximum synthesis method has the following calculation formula:
NDVImax＝max(NDVI1,NDVI2,...,NDVIn) (7)
in the formula, NDVImaxIs the maximum value of NDVI in this time period, NDVI1，NDVI2，NDVInA plurality of NDVI values for different periods of time within the time period.
The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.

Claims (7)

1. An NDVI normalization method based on a high-resolution remote sensing image is characterized by comprising the following steps:
step 1, preprocessing medium and high resolution remote sensing image data;
step 2, obtaining a preliminary normalization result by adopting global linear fitting normalization;
step 3, obtaining a final normalization result by adopting a maximum synthesis method;
step 3.1, carrying out geometric registration on the primary normalization results of the images in different time in the same region;
and 3.2, after the primary normalization results in the same region and the same time period are superposed, selecting the maximum value at each pixel as the value of the final result.
2. The NDVI normalization method based on the high-resolution remote sensing image according to claim 1, characterized in that: the pretreatment of the medium and high resolution remote sensing image data in the step 1 comprises the following substeps:
step 1.1, carrying out radiometric calibration, FLAASH atmospheric correction and orthorectification with reference images on the medium-resolution remote sensing image, and then carrying out combined calculation by using a red light wave band and a near infrared wave band to obtain NDVI data used for reference, wherein the calculation formula is as follows:
Figure FDA0003506733940000011
in the formula, ρNIRFor near infrared band data, pREDData of red light wave band;
step 1.2, after orthorectification, radiometric calibration and rapid atmospheric correction are carried out on the high-resolution remote sensing image, the NDVI data to be normalized is obtained by utilizing combined calculation of a red light wave band and a near infrared wave band;
step 1.3, applying IsoData Classification unsupervised Classification to the high-resolution remote sensing image preprocessed in the step 1.2 to obtain Classification data corresponding to NDVI data to be normalized;
and step 1.4, applying a multi-scale convolution feature fused deep learning cloud and shadow detection method MSCFF to the high-resolution remote sensing image preprocessed in the step 1.2 to obtain cloud mask data corresponding to the data to be normalized.
3. The NDVI normalization method based on the high-resolution remote sensing image according to claim 2, characterized in that: the step 2 of obtaining the preliminary normalization result by adopting global linear fitting normalization comprises the following substeps:
step 2.1, after removing the cloud, the shadow and the background through the classification data obtained in the step 1.3 and the cloud mask obtained in the step 1.4, the high-resolution NDVI data to be normalized is down-sampled to the medium resolution of the reference data;
step 2.2, according to the high-resolution remote sensing image classification data obtained in the step 1.3, calculating the purity of each pixel of the medium-resolution remote sensing image, and screening out the pixels with the purity smaller than a set threshold value T as pure pixels;
step 2.3, based on the pure pixels screened out in the step 2.2, solving linear relation coefficients of the NDVI data to be normalized and the reference NDVI data by utilizing Huber type M estimation;
and 2.4, obtaining a normalization value of the high-resolution NDVI data to be normalized by using the coefficient of the linear relation obtained in the step 2.3.
4. The NDVI normalization method based on the high-resolution remote sensing image according to claim 3, characterized in that: the purity calculation formula of the pixels in the step 2.2 is as follows:
Figure FDA0003506733940000021
in the formula, r is the pixel purity; k is a radical ofcThe number of pixels belonging to the class c in the high-resolution classification data in any medium-resolution pixel range is, and c is the ground object class with the largest proportion in any medium-resolution pixel range; and m is the ratio of the medium resolution to the high resolution of the remote sensing image, and m multiplied by m is the number of high resolution pixels in the pixel range of the medium resolution remote sensing image.
5. The NDVI normalization method based on the high-resolution remote sensing image according to claim 4, characterized in that: the linear relationship between the NDVI data to be normalized and the reference NDVI data in step 2.3 is as follows:
yn＝a×xn+b (3)
in the formula, xnFor the medium resolution of the pure pixels down-sampled in step 2.1 NDVI data to be normalized, ynReferring to NDVI data for the medium resolution of the pure pixel obtained in the step 1.1, wherein a and b are coefficients of a linear relation;
a more robust Huber type M estimation solution is used:
euv＝a×xuv+b-yuv (4)
by minimizing
Figure FDA0003506733940000022
To solve for coefficients, where ρ () is a Huber-type influence function:
Figure FDA0003506733940000023
wherein c is the Huber parameter.
6. The NDVI normalization method based on the high-resolution remote sensing image according to claim 5, characterized in that: step 2.4, the calculation formula of the normalization value of the high-resolution NDVI data to be normalized is as follows:
y′m＝a×xm+b (6)
in the formula, xmThe m pixel, y 'in the high-resolution data to be normalized obtained in the step 1.2'mAnd a and b are linear relation coefficients obtained by solving in the step 2.3 for the corresponding high-resolution normalization result values.
7. The NDVI normalization method based on the high-resolution remote sensing image according to claim 1, characterized in that: the maximum synthesis in step 3.2 is calculated as follows:
NDVImax＝max(NDVI1,NDVI2,...,NDVIn) (7)
in the formula, NDVImaxIs the maximum value of NDVI in this time period, NDVI1，NDVI2，NDVInA plurality of NDVI values for different periods of time within the time period.
CN202210151368.2A 2022-02-16 2022-02-16 NDVI normalization method based on high-resolution remote sensing image Pending CN114639014A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210151368.2A CN114639014A (en) 2022-02-16 2022-02-16 NDVI normalization method based on high-resolution remote sensing image

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210151368.2A CN114639014A (en) 2022-02-16 2022-02-16 NDVI normalization method based on high-resolution remote sensing image

Publications (1)

Publication Number Publication Date
CN114639014A true CN114639014A (en) 2022-06-17

Family

ID=81945988

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210151368.2A Pending CN114639014A (en) 2022-02-16 2022-02-16 NDVI normalization method based on high-resolution remote sensing image

Country Status (1)

Country Link
CN (1) CN114639014A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115205682A (en) * 2022-07-04 2022-10-18 中国矿业大学(北京) NDVI maximum value remote sensing data product seamless production processing method

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115205682A (en) * 2022-07-04 2022-10-18 中国矿业大学(北京) NDVI maximum value remote sensing data product seamless production processing method

Similar Documents

Publication Publication Date Title
Bastin et al. Aboveground biomass mapping of African forest mosaics using canopy texture analysis: toward a regional approach
Rakwatin et al. Restoration of Aqua MODIS band 6 using histogram matching and local least squares fitting
Dadon et al. Use of derivative calculations and minimum noise fraction transform for detecting and correcting the spectral curvature effect (smile) in Hyperion images
CN105809148B (en) Crop drought recognition and risk assessment method based on remote sensing space-time spectrum fusion
Zhu et al. Cloud and cloud shadow detection for Landsat images: The fundamental basis for analyzing Landsat time series
US10054438B2 (en) Method and system for geometric referencing of multi-spectral data
AU2016230926A1 (en) Method and apparatus for processing spectral images
CN107610050B (en) High-spatial-temporal-resolution land surface temperature fusion method based on multi-source satellite data
CN112183209A (en) Regional crop classification method and system based on multi-dimensional feature fusion
Kumar et al. Comparison of efficient techniques of hyper-spectral image preprocessing for mineralogy and vegetation studies
CN113033670A (en) Method for extracting rice planting area based on Sentinel-2A/B data
CN109308688B (en) Visible light and near-infrared band thick cloud and shadow removing method
Sagan et al. Data-driven artificial intelligence for calibration of hyperspectral big data
CN115372282A (en) Farmland soil water content monitoring method based on hyperspectral image of unmanned aerial vehicle
CN114639014A (en) NDVI normalization method based on high-resolution remote sensing image
Scheffler et al. Preprocessing of hyperspectral images: a comparative study of destriping algorithms for EO1-hyperion
Zhou et al. A geometric misregistration resistant data fusion approach for adding red-edge (RE) and short-wave infrared (SWIR) bands to high spatial resolution imagery
Chen et al. Improved relative radiometric normalization method of remote sensing images for change detection
Elsharkawy et al. Improvement in the detection of land cover classes using the Worldview-2 imagery
Pal et al. Destriping of Hyperion images using low-pass-filter and local-brightness-normalization
Kong et al. Super resolution of historic Landsat imagery using a dual generative adversarial network (GAN) model with CubeSat constellation imagery for spatially enhanced long-term vegetation monitoring
Pal et al. Noise reduction and destriping using local spatial statistics and quadratic regression from Hyperion images
Toro Herrera et al. Semi-automated production and filtering of satellite derived water quality parameters
CN113902759B (en) Space-spectrum information combined satellite-borne hyperspectral image segmentation and clustering method
Ouerghemmi et al. Urban vegetation mapping using hyperspectral imagery and spectral library

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