CN115205682A - NDVI maximum value remote sensing data product seamless production processing method - Google Patents

NDVI maximum value remote sensing data product seamless production processing method Download PDF

Info

Publication number
CN115205682A
CN115205682A CN202210846115.7A CN202210846115A CN115205682A CN 115205682 A CN115205682 A CN 115205682A CN 202210846115 A CN202210846115 A CN 202210846115A CN 115205682 A CN115205682 A CN 115205682A
Authority
CN
China
Prior art keywords
ndvi
image
value
adjusted
image data
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
Application number
CN202210846115.7A
Other languages
Chinese (zh)
Other versions
CN115205682B (en
Inventor
李军
秦婷婷
张成业
马雪松
王雅颖
王金阳
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.)
Beijing Shulun Technology Co ltd
China University of Mining and Technology Beijing CUMTB
Original Assignee
Beijing Shulun Technology Co ltd
China University of Mining and Technology Beijing CUMTB
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 Beijing Shulun Technology Co ltd, China University of Mining and Technology Beijing CUMTB filed Critical Beijing Shulun Technology Co ltd
Priority to CN202210846115.7A priority Critical patent/CN115205682B/en
Publication of CN115205682A publication Critical patent/CN115205682A/en
Application granted granted Critical
Publication of CN115205682B publication Critical patent/CN115205682B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06V20/188—Vegetation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
    • G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
    • 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
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Health & Medical Sciences (AREA)
  • General Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Evolutionary Computation (AREA)
  • Databases & Information Systems (AREA)
  • Artificial Intelligence (AREA)
  • Astronomy & Astrophysics (AREA)
  • Remote Sensing (AREA)
  • Image Processing (AREA)

Abstract

The invention discloses a method for seamless production processing of NDVI maximum remote sensing data products, which comprises the following steps: A. acquiring image data and screening to obtain an image data set of a research area; B. calculating the NDVI value of pixels in the image data set of the research area; C. determining a reference image and an image to be adjusted in the same spatial region; D. building a cumulative distribution function model to count cumulative distribution data of the NDVI values; E. and establishing a network matching model corresponding to the reference image and the image to be adjusted to adjust the image to be adjusted, comparing the NDVI mean values of the image data of the same spatial region and obtaining the NDVI maximum value image data. According to the invention, the reference image and the image to be adjusted are extracted, the matching correction relation between the images is established, the NDVI data missing in the maximum green time of the vegetation is reconstructed, the problem of mosaic suture lines generated by inconsistent data production time is solved, the calculation result is closer to the maximum value of the NDVI year, and the method has the advantages of simple and efficient data processing flow, short processing time, high efficiency and the like.

Description

NDVI maximum value remote sensing data product seamless production processing method
Technical Field
The invention relates to the field of remote sensing cloud computing, remote sensing image processing and vegetation index product production, in particular to a NDVI maximum value remote sensing data product seamless production processing method.
Background
The vegetation is one of the most important components of a land ecosystem, has a key role of connecting ecological elements such as hydrology, soil, atmosphere and the like, and can provide powerful guarantee for a natural ecosystem and human productive life (related documents 1. The Vegetation Index established by using a remote sensing means can strengthen and extract Vegetation information only by simple band calculation, wherein a Normalized Difference Vegetation Index (NDVI) is one of the most widely used remote sensing Vegetation indexes at present, and an NDVI value is calculated by image data (such as remote sensing images), so that calculation errors exist due to cloud and cloud shadow occlusion. The index is based on the high absorptivity of chlorophyll in RED light (RED) wave band and the high reflectivity of plant body in Near Infrared (NIR) wave band, and can effectively reflect the growing density and relative activity of vegetation. The NDVI remote sensing data acquisition system is widely applied to vegetation monitoring and ecosystem change monitoring in a global range, a certain period maximum value (such as the NDVI annual maximum value) of the NDVI can well reflect vegetation conditions of vegetation in the most vigorous period, the annual vegetation change characteristics can be more comprehensively described by utilizing the long-time multi-year NDVI, and an important subject of research on the vegetation ecosystem is also provided, so that how to acquire a high-quality NDVI annual maximum value remote sensing data product is particularly critical, and a technical problem to be solved urgently is also provided.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a NDVI maximum remote sensing data product seamless production processing method, which reconstructs NDVI data missing in the maximum green time of vegetation by extracting reference images and images to be adjusted and establishing a matching correction relation between the images, solves the problem of mosaic suture caused by inconsistent data production time, enables the calculation result to be closer to the maximum value of NDVI year, and has the advantages of simple and efficient data processing flow, short processing time, high efficiency and the like.
The purpose of the invention is realized by the following technical scheme:
a method for processing NDVI maximum value remote sensing data products in seamless production comprises the following steps:
A. collecting image data to form an image data set, wherein the image data comprises an image ID number, wave band data and image information, and the image information comprises date information and space information; determining a research area and a research date of the research area, and screening by taking the research area as a spatial screening area and taking the research date as a date screening range to obtain a research area image data set; traversing pixels in image data of an image data set of a research area, selecting a wave band corresponding to cloud and cloud shadow, and removing a cloud and cloud shadow area in the image data;
B. calculating the normalized difference vegetation index NDVI of the pixels in the image data of the cloud and cloud shadow removing area in the image data set of the research area according to the following formula and correspondingly storing to obtain a pixel NDVI value data set:
Figure BDA0003728789810000021
wherein NIR represents the surface reflectivity of a near infrared band, RED represents the surface reflectivity of a RED light band;
C. counting the normalized difference vegetation index NDVI mean values of all pixels in the image data of the same spatial region of the image data set of the research region in the step B, obtaining the image data with the maximum normalized difference vegetation index NDVI mean value of all the image data of the same spatial region, selecting the image data as a reference image, and taking the rest image data of the same spatial region as an image to be adjusted;
D. building an accumulative distribution function model, wherein the accumulative distribution function model respectively carries out pixel NDVI value statistics on a reference image and an image to be adjusted in the same spatial region according to accumulative ratio from small to large order to obtain NDVI value accumulative distribution data of the reference image and NDVI value accumulative distribution data of the image to be adjusted, the NDVI value accumulative distribution data of the reference image and the image to be adjusted comprise NDVI values and accumulative ratio numbers corresponding to the NDVI values, and the accumulative ratio number corresponding to the NDVI values is the accumulative ratio number which is less than or equal to the NDVI value;
E. constructing a network matching model for correspondingly matching the reference image and the image to be adjusted;
e1, establishing an NDVI value a in the image to be adjusted by a network matching model i And NDVI value b in reference image j Is (a) of i ，b j ) Matching relationship (a) in network matching model i ，b j ) The matching method comprises the following steps:
according to the matching relation (a) i ，b j ) NDVI value a of the intermediate image to be adjusted i The corresponding cumulative percentage and the NDVI value b of the reference image j Corresponding accumulated proportion numbers are equal, and corresponding matching is carried out;
e2, matching relationship (a) i ，b j ) After finishing, the NDVI value b of the reference image is used j Adjusting the images to be adjusted in the same space region according to the following method:
will match the relationship (a) i ，b j ) NDVI value b of the reference image j Replacing the NDVI value a of the image to be adjusted with the corresponding assignment i And obtaining the adjusted to-be-adjusted image and the adjusted pixel NDVI value data set, comparing the NDVI mean values of the reference image and the adjusted to-be-adjusted image in the same spatial region, and selecting the image data with the largest NDVI mean value as the NDVI maximum value image data. The steps C to E mainly comprise processing the image data of the same spatial region and different time in the research area and adjusting the image data to obtain NDVIMaximum value image data.
In order to realize the image data processing of all spatial regions in the image data set of the research area so as to obtain the image data product of the NDVI maximum value of the research area, the invention also comprises the following method:
F. and D, sequentially processing the image data of all the spatial regions in the image data set of the research area according to the steps C to E to respectively obtain NDVI maximum value image data of all the spatial regions of the research area, and sequentially fusing the NDVI maximum value image data of all the spatial regions of the research area to obtain an NDVI maximum value image data product of the research area and an NDVI maximum value data set of the research area.
Preferably, the present invention further comprises, in step D, the following method:
the cumulative distribution function model is constructed by the NDVI value cumulative distribution data of the reference image to obtain a cumulative distribution function diagram of the NDVI value of the reference image, and the cumulative distribution function diagram of the NDVI value of the reference image takes the NDVI value as an abscissa and the cumulative percentage corresponding to the NDVI value as an ordinate;
the cumulative distribution function model is constructed by NDVI value cumulative distribution data of the image to be adjusted to obtain a cumulative distribution function graph of the NDVI value of the image to be adjusted, the NDVI value cumulative distribution function graph of the image to be adjusted takes the NDVI value as an abscissa, and the cumulative percentage corresponding to the NDVI value as an ordinate.
Within the scope of the inventive idea, the invention provides a second preferred solution: and B, removing the cloud shadow area from the image data of the cloud shadow area in the step A, constructing a cloud mask corresponding to the cloud shadow area, and obtaining cloud mask data, wherein the cloud mask data comprises position information of the cloud shadow area and the cloud shadow area.
The further preferred technical scheme is as follows: in step E2, before determining NDVI maximum image data, a supplementary correction process is further performed on a reference image in the same spatial region and a cloud and cloud shadow region in an adjusted image to be adjusted, where the method includes: and correspondingly replacing the NDVI mean values of the cloud and cloud shadow areas in the reference image and the adjusted image to be adjusted by using the NDVI mean values of the non-cloud and non-cloud shadow areas in the other adjusted image to be adjusted or the reference image under the same position information to realize supplementary correction, then comparing the NDVI mean values of the supplementary corrected reference image and the supplementary corrected image to be adjusted in the same space area, and selecting the image data with the largest NDVI mean value as the NDVI maximum value image data.
Preferably, the present invention also includes the following method: carrying out region division on the NDVI maximum value image data product in the research region according to the NDVI value range; dividing the region with the NDVI value range of 0.6-1 into a first type region, dividing the region with the NDVI value range of 0.2-0.6 into a second type region, dividing the region with the NDVI value range of 0-0.2 into a third type region, and dividing the region with the NDVI value range of-1-0 into a fourth type region.
The further preferred technical scheme is as follows: and extracting edge lines of the image data product with the maximum NDVI value in the research area according to the region classification and displaying the edge lines in a classification manner.
Compared with the prior art, the invention has the following advantages and beneficial effects:
the invention provides a seamless production processing method for remote sensing data products with maximum NDVI (normalized difference vegetation index) values, which is characterized in that a matching correction relation between images is established by extracting reference images and images to be adjusted, NDVI data missing in maximum green time of vegetation is reconstructed, the problem of inlaid stitches generated by inconsistent data production time is solved, the calculation result is closer to the maximum NDVI value, and the method has the advantages of simple and efficient data processing flow, short processing time, high efficiency and the like.
Drawings
FIG. 1 is a schematic flow chart of an embodiment of the present invention;
FIG. 2 is a data example of image data of a portion of the same spatial region of an exemplary region of interest in an embodiment;
FIG. 3 is an example of image data obtained by removing cloud and cloud shadows from image data of a portion of the same spatial region of an example study area in an embodiment;
FIG. 4 is a diagram illustrating NDVI distribution before adjustment of image data of a portion of the same spatial region in a study area in an example embodiment;
FIG. 5 is a graph of the cumulative distribution function of the NDVI values of the image data of FIG. 4;
FIG. 6 is a diagram illustrating an exemplary method for adjusting NDVI;
FIG. 7 is a graph of NDVI cumulative distribution function after adjustment of image data of a portion of the same spatial region of an exemplary study area in an embodiment
FIG. 8 is a diagram illustrating the NDVI distribution of the image data of the same spatial region of the study area after adjustment in the embodiment;
FIG. 9 is a graph of the annual maximum NDVI distribution of the exemplary study area in the examples.
Detailed Description
The present invention will be described in further detail with reference to the following examples:
examples
As shown in fig. 1 to 9, a method for processing seamless production of NDVI maximum remote sensing data products includes the following steps:
A. collecting image data to form an image data set, wherein the image data comprises an image ID (the image data is a unique identifier of an image searched by the image data set), wave band data (such as wave band information, the type of a wave band value, projection information and the like), and image information, and the image information comprises date information, space information, the width and the height of the image, description and the like; the image data (meeting the geometric and radiation quality requirements and having a spatial resolution of 30 m) of the embodiment is derived from Landsat series satellites or atmospheric corrected Earth surface reflectivity data generated by a Landsat 8 Operational Land Imager (OLI) sensor, and an image data set can be loaded by means of a GEE system (Google Earth Engine). In the embodiment, landsat series satellite remote sensing data is adopted, the data has the characteristics of high time resolution, high spatial resolution and long coverage time range, the spatial resolution is 30m, and the coverage time range is up to now in 1984; the invention can be obtained by the earth surface reflectivity data carried by the GEE system, which meets the requirements of geometric and radiation quality and is processed by a single-channel algorithm jointly created by Rochester's Institute of Technology (RIT) and the national aeronautics and astronautics administration (NASA) Jet Propulsion Laboratory (JPL) through atmospheric correction.
Determining a research area (generally, geographic coordinates of the research area are used as a spatial area) and a research date of the research area, and screening by using the research area as a spatial screening area and using the research date as a date screening range to obtain a research area image data set; for example, taking 2020 as an example, if the study date is determined to be from 1/4/09/30/2020, the date parameter is set to "filterDate (2020-04-01,2020-09-31)", i.e., all the remote sensing images from 1/4/2020 to 30/09/31/2020 are screened out.
Traversing pixels in image data (see the illustration of fig. 2) of an image data set of a research area, selecting a waveband containing cloud and cloud shadow corresponding to each pixel, and removing a cloud and cloud shadow area in the image data.
Figure BDA0003728789810000061
Table 1 illustrates pixel _ qa band data in a scene image data
The above table provides information about fill, transparency, water, cloud Shadow, snow, cloud confidence, and Cloud confidence of each pixel for the band data in the image data, and it can be known from the above table that, in the band data, the Cloud (Cloud) and the Cloud Shadow (Cloud Shadow) select a band including the Cloud and the Cloud Shadow corresponding to each pixel and remove the Cloud and the Cloud Shadow in the image data (see the illustration in fig. 3). Preferably, in the embodiment, the cloud mask is constructed and cloud mask data is obtained by removing the cloud and cloud shadow areas corresponding to the cloud and cloud shadow areas in the image data of the cloud and cloud shadow areas, and the cloud mask data includes position information of the cloud and cloud shadow areas.
B. Calculating the normalized difference vegetation index NDVI of the pixels in the image data of all the cloud and cloud shadow removal areas in the image data set of the research area according to the following formula and correspondingly storing to obtain a pixel NDVI value data set (sequentially and correspondingly storing according to the image ID number and the pixel coordinates, see the NDVI value distribution illustrated in fig. 4 and 5):
Figure BDA0003728789810000071
where NIR represents the surface reflectance in the near infrared band and RED represents the surface reflectance in the RED band.
C. Counting the normalized difference vegetation index NDVI mean values of all pixels in the image data of the same spatial region of the image data set of the research region in the step B, obtaining the image data with the maximum normalized difference vegetation index NDVI mean value of all the image data of the same spatial region, selecting the image data as a reference image, and taking the rest image data of the same spatial region as an image to be adjusted; each scene image in the image data set of the research area has a unique image ID number, and the reference image and the image to be adjusted respectively comprise the image ID numbers. In the present embodiment, the image of the part from 16 days 4 to 28 days 7 of 2020 is selected as an example for detailed description (see fig. 2 to 5); the GEE system selects the 'NDVI' wave band of each scene image by using a 'select' function, calculates each Jing Xiangyuan NDVI value and each scene image NDVI mean value by using a 'reduce region' method, counts each scene image NDVI mean value and outputs an image with the maximum mean value. The NDVI mean values of the scene images are shown in table 2 below, and it can be seen from the table that the NDVI mean value of the "LC08_129042 \u20200728" image is the largest, 0.783 and is used as the reference image, and the other scene NDVI is used as the image to be adjusted.
Figure BDA0003728789810000072
Table 2 NDVI average value table for selecting images from 16 days 4/month to 28 days 7/month 2020
D. An accumulative distribution function model is constructed, the accumulative distribution function model respectively carries out pixel NDVI value statistics on a reference image and an image to be adjusted in the same space region according to the accumulative ratio of the order from small to large to obtain the NDVI value accumulative distribution data of the reference image and the NDVI value accumulative distribution data of the image to be adjusted, the NDVI value accumulative distribution data of the reference image and the image to be adjusted comprise the NDVI value and the accumulative ratio corresponding to the NDVI value (see the NDVI value distribution situation illustrated in fig. 4 and 5), the accumulative ratio corresponding to the NDVI value is the accumulative ratio less than or equal to the NDVI value (as shown in fig. 5, the NDVI of an abscissa is the NDVI value, the probability of an ordinate is the accumulative ratio corresponding to the NDVI value, the accumulative ratio of the NDVI value is represented by '1', the accumulative ratio is 100% by the '1', for example, the accumulative ratio of the abscissa is 0.8, and the ordinate is 0.5, the accumulative ratio of the NDVI value less than or equal to 0.8 is 0.5).
In some embodiments, in step D, the following method is further included:
the cumulative distribution function model is constructed by using the NDVI value cumulative distribution data of the reference image to obtain a reference image NDVI value cumulative distribution function graph, and the NDVI value cumulative distribution function graph of the reference image NDVI value takes the NDVI value as a horizontal coordinate (corresponding to NDVI in the graph) and the cumulative percentage (corresponding to probability in the graph) corresponding to the NDVI value as a vertical coordinate.
The cumulative distribution function model is constructed by NDVI value cumulative distribution data of the image to be adjusted to obtain a cumulative distribution function graph of the NDVI value of the image to be adjusted, and the cumulative distribution function graph of the NDVI value of the image to be adjusted takes the NDVI value (NDVI in a corresponding graph) as an abscissa and the cumulative percentage (probability in the corresponding graph) corresponding to the NDVI value as an ordinate.
E. Constructing a network matching model for correspondingly matching the reference image and the image to be adjusted;
e1, establishing an NDVI value a in the image to be adjusted by a network matching model i And the NDVI value b in the reference image j Is (a) of i ，b j ) Matching relationship (a) in network matching model i ，b j ) The matching method comprises the following steps:
according to the matching relation (a) i ，b j ) NDVI value a of the intermediate image to be adjusted i The corresponding cumulative percentage and the NDVI value b of the reference image j Corresponding cumulative percentage numbers are equal to perform corresponding matching;
e2, matching relationship (a) i ，b j ) After finishing, the NDVI value b of the reference image is used j Adjusting the images to be adjusted in the same space region according to the following method:
will match the relationship (a) i ，b j ) NDVI value b of the reference image j Replacing the NDVI value a of the image to be adjusted by corresponding assignment i And obtaining the adjusted to-be-adjusted image and the adjusted pixel NDVI value data set, comparing the NDVI mean values of the reference image and the adjusted to-be-adjusted image in the same spatial region, and selecting the image data with the largest NDVI mean value as the NDVI maximum value image data. As shown in fig. 6, the NDVI value of the image to be adjusted is 620 (actually 0.620,in this embodiment, the NDVI value is enlarged by 1000 times for comparison, the probability is 0.625, the NDVI value is 800 (actually 0.800, the NDVI value is enlarged by 1000 times for comparison), and the probability is 0.625 in the NDVI value cumulative distribution function graph of the reference image, so the matching relationship (a) (a i ，b j ) If so, (620, 800) the NDVI value 800 is assigned to replace the NDVI value 620 of the image to be adjusted, and so on, to obtain the adjusted image to be adjusted and the adjusted pixel NDVI value data set (see fig. 7 and fig. 8 for example of the distribution of the adjusted NDVI values).
In some embodiments, in the step a, a cloud mask is constructed and cloud mask data is obtained by removing the cloud and cloud shadow areas corresponding to the cloud and cloud shadow areas from the image data, and the cloud mask data includes position information of the cloud and cloud shadow areas. In some embodiments, in step E2, step E2 may be replaced as follows:
e2, matching relationship (a) i ，b j ) After finishing, the NDVI value b of the reference image is used j Adjusting the images to be adjusted in the same space region according to the following method: will match the relationship (a) i ，b j ) NDVI value b of reference image j Corresponding to the NDVI value a of the image to be adjusted i Obtaining an image to be adjusted after adjustment and an adjusted pixel NDVI value data set; the method comprises the following steps of performing supplementary correction processing on a reference image in the same space region and a cloud and cloud shadow region in an adjusted image to be adjusted, wherein the supplementary correction processing comprises the following steps: the NDVI mean values of cloud and cloud shadow areas in the reference image and the adjusted image to be adjusted are correspondingly replaced by NDVI mean value maximum values of non-cloud and non-cloud shadow areas in other adjusted images or non-cloud and non-cloud shadow areas in the reference image under the same position information to realize supplementary correction (the method specifically comprises the steps of determining the cloud and cloud shadow areas needing to be corrected and determining the position range (geographical coordinate range), finding the corresponding position range in the reference image and the adjusted image to be adjusted and obtaining all non-cloud and non-cloud shadow areas in the same position range, then calculating the NDVI mean values of all the non-cloud and non-cloud shadow areas and selecting the NDVI maximum mean value of the non-cloud and non-cloud shadow areas, and adopting the NDVI mean value maximum values of the non-cloud and non-cloud shadow areas of the same position informationTo replace and correct the NDVI mean values of the cloud and cloud shadow regions), and then compare the NDVI mean values of the reference image after the complementary correction and the image to be adjusted after the complementary correction in the same spatial region, and select the image data with the largest NDVI mean value as the NDVI maximum value image data. This example illustrates the maximum NDVI data product obtained in the study area 2020 as shown in fig. 9.
F. And D, sequentially processing the image data of all the spatial regions in the image data set of the research region according to the steps C to E to respectively obtain NDVI maximum value image data of all the spatial regions of the research region, and sequentially fusing the NDVI maximum value image data of all the spatial regions of the research region to obtain an NDVI maximum value image data product of the research region and an NDVI maximum value data set of the research region (see, for example, a annual maximum NDVI distribution diagram of the research region in FIG. 9). In this embodiment, 300 sample points are randomly selected from the NDVI maximum image data product in the research area, the NDVI value of the original image corresponding to the 300 sample points is actually calculated, and the accuracy is verified according to the Root Mean Square Error (RMSE) of the NDVI product and the maximum NDVI in the real year,
Figure BDA0003728789810000101
where n is the number of sample points,
Figure BDA0003728789810000102
for the maximum NDVI value, y, obtained according to the method of the invention for the i-th sample point i The true year maximum NDVI value for the ith sample point; RMSE =0.007 was obtained by calculation, and the accuracy thereof was very ideal.
In some embodiments, when the NDVI value is close to 1, it means that the reflectance in the red band of the investigation region is low and the reflectance in the near infrared band is high; according to the fact that the red light wave band is absorbed and the near infrared wave band is reflected, the fact that the green vegetation in the research area is dense or the vegetation growth condition is healthy is shown. When the NDVI value is close to 0, it means that the reflectance of the red band of the investigation region is almost equal to the reflectance of the near infrared band. The research area reflects red wave band and near infrared wave band, the vegetation grows badly or has no vegetation at all, and the vegetation may be rocks, bare soil, water or urban built-up area, etc. When the NDVI value is close to-1, the reflectivity of a red light wave band in a research area is high, and the reflectivity of a near infrared wave band is low; the research area has strong absorption to near infrared wave band, and reflects all visible light wave band including red light wave band, and no vegetation in the research area may be ice, snow and cloud. Carrying out region division on the NDVI maximum value image data product in the research region according to the NDVI value range; the area with the NDVI value range of 0.6 to 1 is divided into a first type area (high value area, generally temperate rainforest or tropical rainforest), the area with the NDVI value range of 0.2 to 0.6 is divided into a second type area (middle value area, generally shrub and grassy land), the area with the NDVI value range of 0 to 0.2 is divided into a third type area (low value area, generally vegetation-poor area such as rock, sand stone or bare soil), and the area with the NDVI value range of-1 to 0 is divided into a fourth type area (generally cloud or snow coverage area). Preferably, edge lines are extracted from the NDVI maximum image data products of the research area according to the region classification category and are displayed in a classified manner.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.

Claims (7)

1. A NDVI maximum value remote sensing data product seamless production processing method is characterized by comprising the following steps: the method comprises the following steps:
A. collecting image data to form an image data set, wherein the image data comprises an image ID number, wave band data and image information, and the image information comprises date information and space information; determining a research area and a research date of the research area, and screening by taking the research area as a spatial screening area and taking the research date as a date screening range to obtain a research area image data set; traversing pixels in image data of an image data set of a research area, selecting a wave band corresponding to cloud and cloud shadow, and removing a cloud and cloud shadow area in the image data;
B. calculating the normalized difference vegetation index NDVI of the pixels in the image data of the cloud and cloud shadow removing area in the image data set of the research area according to the following formula and correspondingly storing to obtain a pixel NDVI value data set:
Figure FDA0003728789800000011
wherein NIR represents the surface reflectivity of a near infrared band, RED represents the surface reflectivity of a RED light band;
C. counting the normalized difference vegetation index NDVI mean values of all pixels in the image data of the same spatial region of the image data set of the research region in the step B, obtaining the image data with the maximum normalized difference vegetation index NDVI mean value of all the image data of the same spatial region, selecting the image data as a reference image, and taking the rest image data of the same spatial region as an image to be adjusted;
D. building an accumulative distribution function model, wherein the accumulative distribution function model respectively carries out pixel NDVI value statistics on a reference image and an image to be adjusted in the same spatial region according to accumulative ratio from small to large order to obtain NDVI value accumulative distribution data of the reference image and NDVI value accumulative distribution data of the image to be adjusted, the NDVI value accumulative distribution data of the reference image and the image to be adjusted comprise NDVI values and accumulative ratio numbers corresponding to the NDVI values, and the accumulative ratio number corresponding to the NDVI values is the accumulative ratio number which is less than or equal to the NDVI value;
E. constructing a network matching model for correspondingly matching the reference image and the image to be adjusted;
e1, establishing an NDVI value a in the image to be adjusted by a network matching model i And the NDVI value b in the reference image j Is (a) of i ，b j ) Matching relationship (a) in network matching model i ，b j ) The matching method comprises the following steps:
according to the matching relation (a) i ，b j ) NDVI value a of the intermediate image to be adjusted i The corresponding cumulative percentage and the NDVI value b of the reference image j Corresponding cumulative percentage numbers are equal to perform corresponding matching;
e2, matching relationship (a) i ，b j ) After finishing, the NDVI value b of the reference image is used j Adjusting the images to be adjusted in the same space region according to the following method:
will match the relationship (a) i ，b j ) NDVI value b of the reference image j Replacing the NDVI value a of the image to be adjusted by corresponding assignment i And obtaining the adjusted to-be-adjusted image and the adjusted pixel NDVI value data set, comparing the NDVI mean values of the reference image and the adjusted to-be-adjusted image in the same spatial region, and selecting the image data with the largest NDVI mean value as the NDVI maximum value image data.
2. The NDVI maximum remote sensing data product seamless production processing method according to claim 1, wherein:
F. and D, sequentially processing the image data of all the spatial regions in the image data set of the research area according to the steps C to E to respectively obtain NDVI maximum value image data of all the spatial regions of the research area, and sequentially fusing the NDVI maximum value image data of all the spatial regions of the research area to obtain an NDVI maximum value image data product of the research area and an NDVI maximum value data set of the research area.
3. The NDVI maximum remote sensing data product seamless production processing method according to claim 1, wherein: in step D, the following method is also included:
the cumulative distribution function model is constructed by the NDVI value cumulative distribution data of the reference image to obtain a cumulative distribution function diagram of the NDVI value of the reference image, and the cumulative distribution function diagram of the NDVI value of the reference image takes the NDVI value as an abscissa and the cumulative percentage corresponding to the NDVI value as an ordinate;
the cumulative distribution function model is constructed by NDVI value cumulative distribution data of the image to be adjusted to obtain a cumulative distribution function graph of the NDVI value of the image to be adjusted, the NDVI value cumulative distribution function graph of the image to be adjusted takes the NDVI value as an abscissa, and the cumulative percentage corresponding to the NDVI value as an ordinate.
4. The NDVI maximum remote sensing data product seamless production processing method according to claim 1, wherein: and B, removing the cloud and cloud shadow areas in the image data of the cloud and cloud shadow areas in the step A to construct a cloud mask and obtain cloud mask data, wherein the cloud mask data comprises position information of the cloud and cloud shadow areas.
5. The NDVI maximum remote sensing data product seamless production processing method according to claim 4, wherein: in step E2, before determining NDVI maximum image data, a supplementary correction process is further performed on a reference image in the same spatial region and a cloud and cloud shadow region in an adjusted image to be adjusted, where the method includes:
and correspondingly replacing the NDVI mean values of the cloud and cloud shadow areas in the reference image and the adjusted image to be adjusted by using the NDVI mean values of the non-cloud and non-cloud shadow areas in the other adjusted image to be adjusted or the reference image under the same position information to realize supplementary correction, then comparing the NDVI mean values of the supplementary corrected reference image and the supplementary corrected image to be adjusted in the same space area, and selecting the image data with the largest NDVI mean value as the NDVI maximum value image data.
6. The NDVI maximum remote sensing data product seamless production processing method according to claim 2, wherein: carrying out region division on the NDVI maximum value image data product in the research region according to the NDVI value range; dividing the region with the NDVI value range of 0.6-1 into a first type region, dividing the region with the NDVI value range of 0.2-0.6 into a second type region, dividing the region with the NDVI value range of 0-0.2 into a third type region, and dividing the region with the NDVI value range of-1-0 into a fourth type region.
7. The NDVI maximum remote sensing data product seamless production processing method according to claim 6, wherein: and extracting edge lines of the image data product with the maximum NDVI value in the research area according to the region classification and displaying the edge lines in a classification manner.
CN202210846115.7A 2022-07-04 2022-07-04 NDVI maximum value remote sensing data product seamless production processing method Active CN115205682B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210846115.7A CN115205682B (en) 2022-07-04 2022-07-04 NDVI maximum value remote sensing data product seamless production processing method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210846115.7A CN115205682B (en) 2022-07-04 2022-07-04 NDVI maximum value remote sensing data product seamless production processing method

Publications (2)

Publication Number Publication Date
CN115205682A true CN115205682A (en) 2022-10-18
CN115205682B CN115205682B (en) 2022-12-13

Family

ID=83581811

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210846115.7A Active CN115205682B (en) 2022-07-04 2022-07-04 NDVI maximum value remote sensing data product seamless production processing method

Country Status (1)

Country Link
CN (1) CN115205682B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101948107B1 (en) * 2018-11-13 2019-02-14 전남대학교산학협력단 Estimation method for ripening ratio of rice
CN113869370A (en) * 2021-09-03 2021-12-31 中林信达（北京）科技信息有限责任公司 Tree species refined classification method based on deep learning algorithm and time sequence sentinel image
CN114639014A (en) * 2022-02-16 2022-06-17 武汉大学 NDVI normalization method based on high-resolution remote sensing image

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101948107B1 (en) * 2018-11-13 2019-02-14 전남대학교산학협력단 Estimation method for ripening ratio of rice
CN113869370A (en) * 2021-09-03 2021-12-31 中林信达（北京）科技信息有限责任公司 Tree species refined classification method based on deep learning algorithm and time sequence sentinel image
CN114639014A (en) * 2022-02-16 2022-06-17 武汉大学 NDVI normalization method based on high-resolution remote sensing image

Also Published As

Publication number Publication date
CN115205682B (en) 2022-12-13

Similar Documents

Publication Publication Date Title
CN108764255B (en) Method for extracting winter wheat planting information
Mucher et al. Land cover characterization and change detection for environmental monitoring of pan-Europe
CN110298211B (en) River network extraction method based on deep learning and high-resolution remote sensing image
Chen et al. Mapping forest and their spatial–temporal changes from 2007 to 2015 in tropical hainan island by integrating ALOS/ALOS-2 L-Band SAR and landsat optical images
US20220392215A1 (en) System and Method for Mapping Land Cover Types with Landsat, Sentinel-1, and Sentinel-2 Images
CN114092835B (en) Normalized vegetation index data space-time fusion method based on different space-time resolutions
CN114926748A (en) Soybean remote sensing identification method combining Sentinel-1/2 microwave and optical multispectral images
CN110310246B (en) Sugarcane planting area remote sensing information extraction method based on three-linear array image
Liu et al. Comparative analysis of fractional vegetation cover estimation based on multi-sensor data in a semi-arid sandy area
CN111222539B (en) Method for optimizing and expanding supervision classification samples based on multi-source multi-temporal remote sensing image
Grigillo et al. Automated building extraction from IKONOS images in suburban areas
CN113869370A (en) Tree species refined classification method based on deep learning algorithm and time sequence sentinel image
CN115294460B (en) Method for determining degradation degree of phyllostachys praecox forest, medium and electronic device
Zhu et al. Estimating and mapping mangrove biomass dynamic change using WorldView-2 images and digital surface models
Liu et al. An efficient approach based on UAV orthographic imagery to map paddy with support of field-level canopy height from point cloud data
CN114120101A (en) Soil moisture multi-scale comprehensive sensing method
WO2023043317A1 (en) Method and system for delineating agricultural fields in satellite images
Jiang et al. Unmixing-based spatiotemporal image fusion accounting for complex land cover changes
Liu et al. Thick cloud removal under land cover changes using multisource satellite imagery and a spatiotemporal attention network
CN115272860A (en) Method and system for determining rice planting area, electronic device and storage medium
Tian et al. Impacts of satellite revisit frequency on spring phenology monitoring of deciduous broad-leaved forests based on vegetation index time series
Zhang et al. A 250m annual alpine grassland AGB dataset over the Qinghai-Tibetan Plateau (2000–2019) based on in-situ measurements, UAV images, and MODIS Data
CN115205682B (en) NDVI maximum value remote sensing data product seamless production processing method
Papasaika et al. Fusion of digital elevation models from various data sources
Tian et al. A process-oriented method for rapid acquisition of canopy height model from RGB point cloud in semiarid region

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