CN114821355B

CN114821355B - Coastline automatic identification method and device

Info

Publication number: CN114821355B
Application number: CN202210455364.3A
Authority: CN
Inventors: 毕京鹏; 马万栋; 申文明; 张文国; 吴玲; 郭艳玲; 史园莉; 毕晓玲; 陈绪慧; 任致华
Original assignee: Satellite Application Center for Ecology and Environment of MEE
Current assignee: Satellite Application Center for Ecology and Environment of MEE
Priority date: 2022-04-27
Filing date: 2022-04-27
Publication date: 2022-09-27
Anticipated expiration: 2042-04-27
Also published as: CN114821355A

Abstract

The invention discloses a coastline automatic identification method and a coastline automatic identification device, and belongs to the field of remote sensing monitoring. The method provided by the invention integrates the multi-scene remote sensing image information layer and the mirror image to extract the coastline, utilizes the land utilization data to divide coastline nodes and judge the coastline type attribute, and is suitable for coastline space position extraction and attribute updating under various image imaging conditions. The method overcomes the defects of uncertain position, low precision, difficult type attribute distinguishing and the like of the coastline caused by the influence of tide on the conventional single remote sensing image, simultaneously realizes the automatic study and judgment of the coastline type attribute, and improves the precision and the time efficiency of the automatic interpretation of the coastline. The method can accurately, quickly and automatically extract the spatial position and type attribute of the coastline, and provides basic data support for the protection and management of the natural coastline.

Description

Coastline automatic identification method and device

Technical Field

The invention relates to the field of remote sensing monitoring, in particular to a coastline automatic identification method and a coastline automatic identification device.

Background

The coastline is a boundary between the sea and the land, and the time sequence dynamic change of the coastline is position representation under the influence of comprehensive factors such as position evolution of the natural state of the coastline and development activities of the coastline, and is generally closely related to development activities such as seawater erosion, estuary siltation, sea reclamation activities, sea-enclosing cultivation, seawater salt pan, port and wharfs, and town expansion. The method can accurately acquire the spatial position and the type of the coastline, and is important basic data for protecting and managing the natural coasts and reasonably utilizing the artificial coastline.

Remote sensing has strong advantages in the aspect of coastline information extraction in a large-scale space range by virtue of the advantages of short-period observation, wide coverage range, rich spatial resolution and scale information, large spectral information amount and the like. The method has the advantages that the spatial distribution and type attribute information of the coastline are obtained by using a remote sensing technology, the spatial position form and long-term change of the coastline are mastered, and the method has important effects on the aspects of coastline safety, ecological protection, environmental management, coastline planning and the like.

The method for acquiring the spatial distribution position of the coastline by using the remote sensing image in the prior art mainly comprises an artificial visual interpretation method and a semi-automatic extraction method combining spectral characteristics with artificial visual interpretation. The artificial visual interpretation method includes the steps of overlaying satellite remote sensing images by means of geographic information system analysis software, manually drawing point positions such as spatial end points, nodes, inflection points, break points and the like of a coastline along sea-land boundaries, finally connecting the point positions into a line to obtain a coastline spatial distribution vector, interpreting ground object types attached to the coastline between the point positions, and giving attribute information to the coastline in a bank section.

The following disadvantages mainly exist in the manual visual interpretation method: the work load of drawing the coastline and the like is huge, the time consumption is long, and the updating is relatively slow; meanwhile, professionals are required to have rich knowledge of shoreline position definition and type judgment, and shoreline position and attribute judgment in complex situations is different from person to person and is influenced by different interpreters, so that the error of double eyelids of the shoreline is easily caused; when the coastline is interpreted and sketched, one region usually adopts one image, the influence of tide conditions on the spatial position of the coastline can be ignored, and a large amount of tide correction work is needed for later correction; meanwhile, the manual delineation method is easy to generate various errors such as omission, non-coincident end points, unsmooth land line form and the like, so that the precision of manually delineating the land line cannot meet the management requirement.

The semi-automatic extraction method combining spectral features with manual visual interpretation mainly comprises a single-waveband segmentation method, a water body index method, an edge detection method, an image classification method, an object-oriented method, a polarization method and the like. These methods typically extract the coastline, which is the instantaneous water line at the moment of imaging of the monoscopic image, by combining the spectral features of the monoscopic image with artificial visual interpretation. The actual shoreline is influenced by the tide, and its position changes with the fluctuation of the tide and water, and is in constant change. The instantaneous shoreline extracted from the single-scene image does not consider the tidal conditions, tidal correction, the difference of tidal types of all regions in the country and the like, and the instantaneous shoreline cannot be directly used as a coastline, so that a large amount of later-period tidal correction work is required. And the instantaneous water lines obtained by the single-scene images at different imaging moments are different, and are influenced by different remote sensing images, so that the error of double eyelids of the coastline is easily caused.

Disclosure of Invention

In order to solve the technical problems, the invention provides a coastline automatic identification method and a coastline automatic identification device, which can accurately, quickly and automatically extract the spatial position and the attribute of the coastline and provide basic data support for the protection and management of natural coastlines.

The technical scheme provided by the invention is as follows:

a coastline automated identification method, the method comprising:

s1: acquiring a remote sensing image of a region to be detected and coastal land utilization classification data of the region to be detected;

the remote sensing image of the region to be detected comprises a multi-scene remote sensing image imaged by the region to be detected at different time periods;

s2: generating a mirror image according to the remote sensing image of the area to be detected;

the image range of four to four of the mirror image, the image width, the image height, the pixel size and the coordinate system are the same as those of the remote sensing image of the area to be detected, and the initial value of each pixel of the mirror image is null;

s3: dividing the remote sensing image of the area to be detected into a plurality of sub-areas, wherein each sub-area comprises a multi-scene sub-image of the sub-area imaged at different time periods;

s4: respectively calculating the normalized water body index image of each scene image of each subarea for each subarea;

wherein, the pixel value of the normalized water body index image at the pixel (x, y) is the normalized water body index NDWI of the corresponding sub-image at the pixel (x, y);

NDWI＝(p(Green)-p(NIR))/(p(Green)+p(NIR))；

p (green) and p (nir) are the green band spectral value and the near infrared band spectral value of the sub-image at the pixel (x, y), respectively;

s5: for each subarea, performing binary segmentation on the normalized water body index image corresponding to the scene image by respectively using the segmentation threshold value of each scene image to obtain a binary image;

if the pixel value of the normalized water body index image at the pixel (x, y) is greater than the segmentation threshold value of the corresponding sub-image, the pixel value of the binary image at the pixel (x, y) is 1, otherwise, the pixel value is 0;

s6: for each subregion, calculating enumeration values of all pixels in the subregion, and assigning the enumeration values of all pixels of all subregions to pixels at the same position of the mirror image;

when in use

When M is in contact with _pixel ＝1

When in use

When, M _pixel ＝0

Wherein M is _pixel An enumerated value, N, representing a pixel (x, y) _pixel＝1 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 1, and N _pixel＝0 Representing all binary images in a sub-regionThe pixel value of the image at pixel (x, y) is 0, N _pixel ＝N _pixel＝1 +N _pixel＝0 P is a set interpretation threshold coefficient;

s7: taking the boundary of the pixel with the pixel value of 1 and the pixel with the pixel value of 0 in the mirror image as a coastline;

s8: and obtaining the type attribute of the coastline by using the coastal zone land utilization classification data of the area to be detected.

Further, the S8 includes:

s81: intersecting the coastline with the coastal zone soil utilization classification data to obtain a series of intersection points;

s82: at each intersection point, a line segment perpendicular to the coastline is made to pass through the intersection point, and the lengths of the line segment on the two sides of the coastline are respectively a first length and a second length;

s83: for each intersection point, taking the line segment of the intersection point as one side of a rectangle, and making a closed rectangle passing through the next intersection point of the intersection points to obtain a series of closed rectangular areas;

s84: for each closed rectangular area, determining the type attribute of the closed rectangular area according to the coastal zone land utilization classification data in the closed rectangular area;

s85: and taking the type attribute of the closed rectangular region as the type attribute of a section of coastline between two intersection points on the closed rectangular region to obtain the type attribute of the coastline.

Further, the S84 includes:

s841: for each closed rectangular area, taking an intersection surface of the closed rectangular area and the coastal zone land utilization classification data, and acquiring a land utilization type set of the intersection surface on the coastal zone land utilization classification data;

s842: if the land utilization type set of the closed rectangular area comprises 'sand beach or beach', 'beach or wetland or besmearing', 'mangrove or seaweed bed or coral reef' and 'river', the type attributes of the closed rectangular area are respectively set as 'sandy shoreline', 'argillaceous shoreline', 'biological shoreline', 'estuary shoreline';

or if the land use type set of the closed rectangular region comprises land use types which are not influenced by artificial activities, setting the type attribute of the closed rectangular region as a 'natural shoreline';

s843: if the land utilization type set of the closed rectangular area comprises grassland, forest land or bare rock, and the occupation ratio of the grassland, forest land or bare rock in the closed rectangular area is greater than a first ratio, setting the type attribute of the closed rectangular area as a bedrock shoreline, or setting the type attribute of the closed rectangular area as a natural shoreline;

s844: if the set of land use types of the closed rectangular area includes "dry land or paddy field", "residential land or coastal dike", "dock or harbor or fishing port", "land for transportation", "farming", "salt pan", and the occupation ratio of the "dry land or paddy field", "residential land or coastal dike", "dock or harbor or fishing port", "land for transportation", "farming", "salt pan" in the closed rectangular area is greater than the second ratio, the type attribute of the closed rectangular area is set to "farmland dike", "construction dike", "dock dike", "traffic dike", "farming dike", "salt pan dike", respectively;

or if the land use type set of the closed rectangular region comprises land use types influenced by artificial activities, setting the type attribute of the closed rectangular region as an artificial shoreline.

Further, the S85 includes:

s851: for each closed rectangular area, associating a section of coastline between two of the intersections on the closed rectangular area with the closed rectangular area;

s852: for each closed rectangular area, copying a type attribute of the closed rectangular area to a section of coastline associated with the closed rectangular area.

Further, for any scene image, the segmentation threshold is determined by the following method:

s51: traversing each value of the segmentation threshold value on a value interval of the normalized water body index according to a set step length;

s52: comparing the pixel value of each pixel of the normalized water body index image corresponding to the sub-image with the value of each segmentation threshold, and classifying the pixels of the normalized water body index image with the pixel value larger than the value of the segmentation threshold into a first area, or classifying the pixels into a second area;

s53: calculating the difference g between a first area and a second area corresponding to the value of each segmentation threshold value for the value of each segmentation threshold value;

g＝w ₀ w ₁ (μ ₀ -μ ₁ ) ²

wherein N is ₀ And N ₁ The number of pixels, mu, of the first region and the second region respectively ₀ Is the average value of pixel values of all pixels of the first region, mu ₁ The average value of pixel values of all pixels of the second area is obtained;

s54: and finding out the maximum value of the difference corresponding to the values of all the segmentation threshold values, wherein the value of the segmentation threshold value corresponding to the maximum value is the segmentation threshold value of the sub-image.

Further, the S1 includes:

s11: acquiring satellite remote sensing images, a range vector of an area to be detected and coastal land utilization classification data;

s12: when the satellite remote sensing image, the range vector of the area to be detected and the projection coordinates of the coastal land utilization classification data are inconsistent, coordinate transformation enabling the projection coordinates to be consistent is carried out;

s13: and cutting the satellite remote sensing image by using the range vector of the area to be detected to obtain the remote sensing image of the area to be detected.

Further, the S7 includes:

s71: merging the pixels with the pixel value of 1 in the mirror image into a first object, and merging the pixels with the pixel value of 0 in the mirror image into a second object;

s72: the boundary of the first object and the second object is taken as the coastline.

Further, the method further comprises:

s9: carrying out smoothing processing and topology inspection on the coastline;

s10: and counting the geographic spatial position, the length and the type attribute of each section of coastline of the coastline.

A coastline automated identification apparatus, the apparatus comprising:

the data acquisition module is used for acquiring remote sensing images of the area to be detected and coastal land utilization classification data of the area to be detected;

the mirror image generation module is used for generating a mirror image according to the remote sensing image of the area to be detected;

the image range of four to four, the image width, the image height, the pixel size and the coordinate system of the mirror image are all the same as those of the remote sensing image of the area to be detected, and the initial value of each pixel of the mirror image is null;

the image segmentation module is used for segmenting the remote sensing image of the area to be detected into a plurality of sub-areas, and each sub-area comprises a plurality of scene sub-images imaged by the sub-area at different time periods;

the water body index calculation module is used for calculating the normalized water body index image of each scene image of each subarea for each subarea;

NDWI＝(p(Green)-p(NIR))/(p(Green)+p(NIR))；

the binary segmentation module is used for performing binary segmentation on the normalized water body index image corresponding to each scene image by using the segmentation threshold value of each scene image to obtain a binary image;

the mirror image assignment module is used for calculating enumeration values of all pixels in each sub-region for each sub-region and assigning the enumeration values of all the pixels of all the sub-regions to the pixels at the same position of the mirror image;

when in use

When M is in contact with _pixel ＝1

When in use

When M is in contact with _pixel ＝0

Wherein M is _pixel An enumerated value, N, representing a pixel (x, y) _pixel＝1 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 1, and N _pixel＝0 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 0, and N _pixel ＝N _pixel＝1 +N _pixel＝0 P is a set interpretation threshold coefficient;

the coastline determining module is used for taking the boundary of the pixel with the pixel value of 1 and the pixel with the pixel value of 0 in the mirror image as a coastline;

the coastline type attribute determining module is used for obtaining the type attribute of the coastline by utilizing the coastal zone land utilization classification data of the area to be detected.

Further, the shoreline type attribute determination module includes:

the data intersection unit is used for intersecting the coastline and the coastal zone soil utilization classification data to obtain a series of intersection points;

a line segment determining unit, configured to make a line segment perpendicular to the coastline through the intersection at each intersection point, where lengths of the line segment on both sides of the coastline are a first length and a second length, respectively;

a closed rectangular region determining unit, configured to, for each intersection point, use a line segment of the intersection point as one side of a rectangle, and make a closed rectangle passing through a next intersection point of the intersection point to obtain a series of closed rectangular regions;

the closed rectangular area type attribute determining unit is used for determining the type attribute of each closed rectangular area according to the coastal zone land utilization classification data in the closed rectangular area;

a coastline type attribute determining unit, configured to obtain a type attribute of the coastline by using the type attribute of the closed rectangular area as a type attribute of a coastline segment between two intersection points on the closed rectangular area.

The invention has the following beneficial effects:

the invention provides a method for extracting a coastline by fusing a multi-scene remote sensing image information layer and a mirror image, and judges the coastline type attribute by dividing coastline nodes by land utilization data aiming at the problems of high labor consumption, incapability of ensuring precision and low efficiency of the conventional manual identification and delineation method. The invention relates to a coastline extraction algorithm with strong adaptability, which solves the problems of coastline space position extraction and attribute updating under various image imaging conditions. The method utilizes the multi-temporal image to identify the spatial position of the coastline, and overcomes the defects of uncertain position, low precision, difficult type attribute distinction and the like of the coastline caused by the influence of tide on the conventional single remote sensing image; meanwhile, the automatic study and judgment of the shoreline type attribute are realized, and the accuracy and the time efficiency of automatic interpretation of the shoreline are improved. The method can accurately, quickly and automatically extract the spatial position and the attribute of the coastline, provides basic data support for the protection and management of the natural coastline, can meet the requirements of coastline ecological destruction supervision, natural coastline retention rate accounting, coastline engineering supervision, coastline protection and management and the like by analyzing the change of the spatial position and the type attribute of the coastline extracted by the method, and has a good application prospect in the aspects of natural coastline protection, supervision and management and the like.

Drawings

FIG. 1 is a flow chart of a coastline automated identification method of the present invention;

FIG. 2 is a schematic diagram of a remote sensing image of a detection area and its segmentation into sub-areas;

FIG. 3 is a schematic view of a mirror image;

FIG. 4 is a schematic view of a multi-scene normalized water body index image of a sub-region;

FIG. 5 is a schematic diagram of a binarized image obtained by binary segmentation of a normalized water index image;

FIG. 6 is a schematic diagram of assignment of mirror images;

FIG. 7 is a schematic diagram of a shoreline-type attribute determination process;

FIG. 8 is a diagram illustrating the results of shoreline extraction;

fig. 9 is a schematic view of an automatic coastline identification device according to the present invention.

Detailed Description

In order to make the technical problems, technical solutions and advantages of the present invention more apparent, the following detailed description is given with reference to the accompanying drawings and specific embodiments.

Example 1:

an embodiment of the present invention provides a coastline automatic identification method, as shown in fig. 1, the method including:

the remote sensing image of the area to be detected comprises a multi-scene remote sensing image formed by the area to be detected in different time periods.

In one example, the specific implementation manner of this step may include:

s11: and acquiring satellite remote sensing images, the range vectors of the areas to be detected and classification data of the coastal land utilization.

The satellite remote sensing image can be an original image, and can also be a high-level image processed by radiometric calibration, atmospheric correction, geometric correction and other image enhancement methods. When the satellite remote sensing image is obtained, basic parameter information of the satellite remote sensing image, such as spatial scale (resolution), wave band quantity, spectral range, projection coordinates, non-imaging region values and the like, is also obtained, and the satellite remote sensing image generally at least comprises 4 spectral wave bands of Red, Green, Blue and Near Infrared.

The region to be detected of the coastal zone is generally in a vector format, namely a region range vector to be detected, and relevant vector information is attached to the region to be detected.

Coastal zone land use classification data typically includes class1 … class n land use types.

S12: and when the projection coordinates of the satellite remote sensing image, the range vector of the area to be detected and the coastal land utilization classification data are inconsistent, carrying out coordinate transformation for enabling the projection coordinates to be consistent.

And if the projection coordinates are consistent, performing coordinate conversion processing. The method of coordinate transformation may adopt various methods known to those skilled in the art, and the present invention is not limited thereto.

And if the whole satellite remote sensing image is the area to be detected, removing the non-imaging area according to the basic parameter information of the satellite remote sensing image obtained in the S11, or classifying the non-imaging area as an invalid area.

S2: and generating a mirror image according to the remote sensing image of the area to be detected.

The method comprises the following steps of generating a mirror image according to a to-be-detected area, wherein the range of four to four images, the image width, the image height, the pixel size and the coordinate system of the mirror image are the same as those of a remote sensing image of the to-be-detected area, and the initial value of each pixel of the mirror image is set to be null.

S3: and dividing the remote sensing image of the area to be detected into a plurality of sub-areas, wherein each sub-area comprises a plurality of scene sub-images of the sub-area imaged at different time periods.

This step can divide the remote sensing image of the region to be detected into { subregion 1 … subregion i … subregion n }, where i is 1, 2, …, n, and when the blocks are partitioned, the blocks are generally performed in a manner of overlapping blocks according to the basic parameter information (plane range, imaging width, etc.) in S11. Each sub-region includes satellite images (i.e., the sub-images) imaged in multiple time periods in the sub-region, and the multi-view satellite images in the sub-region can be represented as a set of satellite images imaged at different times under the same satellite column number. The set of multi-view sub-images for each sub-region is { sub-image 1 … sub-image j … sub-image m }, where j is 1, 2, …, m.

wherein, the pixel value of the Normalized Water body Index image at the pixel (x, y) is the NDWI (Normalized Difference Water Index) of the corresponding sub-image at the pixel (x, y);

NDWI＝(p(Green)-p(NIR))/(p(Green)+p(NIR))；

p (green) and p (nir) are the green band spectral value and the near infrared band spectral value of the sub-image at the pixel (x, y), respectively.

Each subregion comprises m scene images, each scene image is calculated to obtain a normalized water body index image, the set of the normalized water body index images of one subregion is { normalized water body index image 1 … normalized water body index image j … normalized water body index image m }, j is 1, 2, …, m.

S5: for each subregion, performing binary segmentation on the normalized water body index image corresponding to each scene image by using the segmentation threshold value of each scene image to obtain a binary image;

if the pixel value of the normalized water body index image at the pixel (x, y) is greater than the segmentation threshold value of the corresponding sub-image, the pixel value of the binary image at the pixel (x, y) is 1, and otherwise, the pixel value is 0.

Each scene subimage j of one subregion is provided with a corresponding segmentation threshold Tj, the set of the segmentation thresholds of one subregion is { T1 … Tj … Tm }, and the normalized water body index image j corresponding to the subimage j is subjected to binary segmentation through the segmentation threshold Tj of the subimage j.

The segmentation threshold Tj is a boundary threshold between the ocean (or water, the same below) and the land (or non-water, the same below) on the normalized water body index image j, and the pixel-by-pixel of the normalized water body index image j is judged through the segmentation threshold Tj to classify the ocean and land categories at the pixel level. Specifically, the pixels of the normalized water body index image j larger than the segmentation threshold Tj are classified as oceans, and the pixel values of the pixels are set to be 1, otherwise, the pixels are classified as lands, and the pixel values of the pixels are set to be 0, so that a binary image is obtained.

Each scene normalized water body index image j of one subregion is subjected to binary segmentation to obtain a binary image, and the set of the binary images of one subregion is a { binary image 1 … binary image j … binary image m }, wherein j is 1, 2, … and m.

In the prior art, a single-image sea-land segmentation threshold is selected, and due to the fact that factors influencing the position accuracy of a coastline are more, the defects and condition limitations of the coastline such as data discontinuity and false edges are easily caused, and the later correction work is required, so that the detection accuracy and speed of the coastline are severely limited. In order to solve the problems, the invention respectively sets the segmentation threshold values for the multi-scene images for segmentation, and comprehensively considers the remote sensing images of multiple time phases to determine the spatial position of the coastline, so that the coastline is more accurate.

The segmentation threshold of each scene image may be set empirically or may be determined according to the image characteristics of the sub-image itself, and for example, for any scene image, the segmentation threshold may be determined as follows:

s51: and traversing each value of the segmentation threshold value on the value interval of the normalized water body index according to a set step length.

The value interval of the normalized water body index is [0,1], and each value can be traversed on [0,1] by taking 0.01 as a step length.

S52: and comparing the pixel value of each pixel of the normalized water body index image corresponding to the sub-image with the value of the segmentation threshold, and classifying the pixels of which the pixel values are larger than the value of the segmentation threshold into a first area (namely, an ocean area), or else, classifying the pixels into a second area (namely, a land area).

g＝w ₀ w ₁ (μ ₀ -μ ₁ ) ²

wherein N is ₀ And N ₁ The number of pixels, mu, of the first region and the second region respectively ₀ Is the average value of pixel values of all pixels of the first region, mu ₁ Is the average value of the pixel values of all the pixels of the second area.

S54: finding out the maximum value of the difference corresponding to the values of all the segmentation threshold values, wherein the value of the segmentation threshold value corresponding to the maximum value is the segmentation threshold value of the sub-image.

The segmentation threshold is calculated in a self-adaptive mode through the normalized water body index image corresponding to each scene image, each calculated segmentation threshold is most suitable for the sub-image, the value of the segmentation threshold enables the difference between the ocean area and the land area to be maximum, the distinguishing degree between the ocean area and the land area to be maximum, the ocean area and the land area obtained through segmentation are enabled to be more accurate, and the accuracy of the coastline is improved.

when in use

When M is in contact with _pixel ＝1

When in use

When M is in contact with _pixel ＝0

Wherein M is _pixel An enumerated value, N, representing a pixel (x, y) _pixel＝1 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 1, N _pixel＝0 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 0, and N _pixel ＝N _pixel＝1 +N _pixel＝0 (ii) a P is a set interpretation threshold coefficient, which represents the percentage of the total pixels, and may be, for example, 10%.

One subregion comprises m binary images, and for a pixel with the coordinate (x, y) of one subregion, the same coordinate (x, y) has m pixel values on the m binary images, namely the m binary images have m pixel values at the same coordinate (x, y), and the enumeration value of the pixel with the coordinate (x, y) is calculated through the m pixel values. The calculation formula is as above, wherein m is N _pixel ＝N _pixel＝1 +N _pixel＝0 。

And (3) calculating an enumeration value of a pixel with coordinates (x, y) for one subregion, assigning the enumeration value to the pixel with the same coordinate position (x, y) of the mirror image, and operating all subregions according to the same method to finish assignment of all pixels of the mirror image.

And after the assignment is completed, the pixel value at the image coordinate (x, y) of the mirror image is the statistical value of the pixel value at the multi-scene binary image coordinate (x, y). As mentioned above, the pixel values 1 and 0 of the binary image represent the ocean and the land, respectively, and thus the pixel values 1 and 0 at the image coordinates (x, y) represent the statistical values of the ocean and the land at the image coordinates (x, y).

S7: and taking the boundary of the pixel with the pixel value of 1 and the pixel with the pixel value of 0 in the mirror image as a coastline.

The pixel values 1 and 0 of the image coordinates represent the statistic values of the ocean and land of the multi-scene sub-image, and the boundary of the ocean 1 and the land 0 is the coastline.

The NDWI of the multi-scene images is calculated, multi-scene binary images of ocean and land segmentation are generated, pixel values of the same positions of the multi-scene binary images are counted and then assigned to the mirror image images, and the coastline is determined through the mirror image images. The invention comprehensively considers the remote sensing images of multiple time phases to determine the spatial position of the coastline, solves the problems of uncertain position and low extraction precision of the coastline caused by the influence of tide on the conventional single remote sensing image, and does not need the later tide correction work.

In one example, the specific implementation manner of this step may include:

s71: and combining the pixels with the pixel value of 1 in the mirror image into a first object, and combining the pixels with the pixel value of 0 in the mirror image into a second object.

The first object is a sea area, the second object is a land area, and the boundary between the sea area and the land area is a coastline.

The coastal zone soil utilization classification data of the area to be detected comprises the land utilization type of the coastal zone of the area to be detected, and the type attribute of the coastline can be determined according to the land utilization type.

In one example, the specific implementation manner of this step may include:

s81: and intersecting the coastline with the coastal zone soil utilization classification data to obtain a series of intersection points.

Assuming that the number of intersections is k, the set of intersections can be represented as q1, q2 … qk, each intersection serving as a segment node of the coastline.

S82: at each intersection point, a line segment perpendicular to the coastline is made across the intersection point, and the lengths of the line segment on both sides of the coastline are a first length L1 and a second length L2, respectively.

And each intersection point is used as a line segment, k line segments are shared, and a set consisting of the k line segments can be expressed as { l1, l2 … lk }. Each line segment has 2 endpoints, and the set of endpoints of { l1, l2 … lk } can be represented as { (q1_ a, q1_ b), ((q2_ a, q2_ b)) … (qk _ a, qk _ b) }. The values of L1 and L2 may be determined as desired.

S83: for each intersection point, the line segment of the intersection point is used as one side of the rectangle, and the next intersection point passing through the intersection point is used as a closed rectangle to obtain a series of closed rectangular areas.

For example, starting from the first intersection point q1, a line segment l1 is taken as one side of a rectangle, a straight line perpendicular to the line segment l1 is respectively made at the end points (q1_ a and q1_ b) of the line segment l1 to obtain a straight line 1 and a straight line 2, a straight line parallel to the line segment l1 is made at the second intersection point q2 to obtain a straight line 3, and a closed rectangular region composed of the line segment l1, the straight line 1, the straight line 2 and the straight line 3 is a first closed rectangular region and is marked as Area 1.

The process is then repeated for a second intersection point, resulting in a second closed rectangular Area, denoted as Area2, and so on. For the non-closed circular coastline, the k intersection points together result in k-1 closed rectangular regions (since the last intersection point can not be used as a closed rectangular region with the next intersection point, there are k-1 closed rectangular regions), and the set is marked as { Area1, Area2 … Area-1 }.

In a special case, if the coastline is closed-loop such as a sea island, k closed rectangular regions may be generated, and the last closed rectangular region is a closed rectangle formed by the k-th intersection and the first intersection, and the set of the k-th intersection and the first intersection is represented as { Area1, Area2 … Area }.

S84: for each closed rectangular area, determining a type attribute of the closed rectangular area according to the coastal zone land utilization classification data within the closed rectangular area.

In each closed rectangular area, a plurality of land utilization types in the coastal zone land utilization classification data are contained, and the type attribute of the closed rectangular area can be determined according to the land utilization types contained in the closed rectangular area.

In one example, the specific implementation manner of this step may include:

s841: for each closed rectangular area, taking the intersection surface of the closed rectangular area and the coastal zone land utilization classification data, and acquiring a land utilization type set of the intersection surface on the coastal zone land utilization classification data.

The set of land use types of each closed rectangular region is part or all of land use types { class1, class2 … class N }, and the land use types { class1, class2 … class N } correspond to Type attributes { Type1, Type2 … Type N }.

S842: if the land use type set of the closed rectangular area includes "sand beach or beach", "mudflat or wetland or mudflat", "mangrove or seaweed bed or coral reef" and "river", the type attribute of the closed rectangular area is set to "sandy shoreline", "muddy shoreline", "biological shoreline" and "estuary shoreline", respectively.

Or if the land use type set of the closed rectangular area comprises land use types which are not influenced by human activities, setting the type attribute of the closed rectangular area as a natural shoreline.

The method comprises the steps of directly assigning Type attributes, and if the land utilization Type set comprises the land utilization Type class, directly assigning the corresponding Type attribute Type to the closed rectangular area. Or, if the land use Type set of the closed rectangular region includes land use types which are not affected by artificial activities, including but not limited to the "sandy shoreline", "argillaceous shoreline", "biological shoreline", and "estuary shoreline" described above are all "natural shorelines", the Type attribute Type may be set to "natural shoreline" directly and uniformly.

S843: if the set of land use types of the closed rectangular area includes "grassland or woodland or bare rock" and the occupation ratio of the "grassland or woodland or bare rock" in the closed rectangular area is greater than the first ratio1, the type attribute of the closed rectangular area is set to "bedrock shoreline", or the type attribute of the closed rectangular area is set to "natural shoreline".

S844: if the set of land use types of the closed rectangular area includes "dry land or paddy field", "residential land or coastal dike", "dock or port or fishing port", "transportation land", "farming", "salt pan", and the occupation ratio of the "dry land or paddy field", "residential land or coastal dike", "dock or port or fishing port", "transportation land", "farming", "salt pan" in the closed rectangular area is greater than the second ratio2, the type attribute of the closed rectangular area is set as "farmland dike", "construction dike", "dock dike", "transportation dike", "farming dike", "salt pan dike", respectively.

Alternatively, if the set of land use types of the closed rectangular area includes land use types affected by human activities, including, but not limited to, the above-mentioned dry land or paddy field "," residential land or coastal embankment "," dock or harbor or fishing port "," transportation land "," farming "," salt pan ", etc., the Type attribute Type thereof may be uniformly set to" artificial shoreline "directly.

Steps S843-S844 first determine the occupation ratio of each of class1, class2 … class n in the closed rectangular area, and then determine the Type attribute Type of the closed rectangular area according to the occupation ratio. Since the above-described "bedrock shoreline" is "natural shoreline", its Type attribute Type may also be set to "natural shoreline". Since the above-mentioned "farmland dike", "construction dike", "dock dike", "traffic dike", "cultivation dike", and "salt pan dike" are all "artificial shorelines", it is also possible to set the Type attribute Type to "artificial shorelines" in a unified manner.

Wherein, according to class of land utilization classification of coastal zone, the aforementioned "forest land" includes, but is not limited to, "deciduous broadleaf forest, coniferous mixed forest, evergreen coniferous bush forest, evergreen broadleaf bush forest, deciduous broadleaf bush forest, evergreen coniferous forest, shrub garden, shrub greenbelt, evergreen broadleaf forest, deciduous broadleaf forest, arbor garden, arbor greenbelt,"; "wetlands" include, but are not limited to, "forest wetlands, bush wetlands," and "grasslands" include, but are not limited to, "meadows, bushes, herbaceous greens, herbaceous wetlands.

S85: and taking the type attribute of the closed rectangular area as the type attribute of a section of coastline between two intersection points on the closed rectangular area to obtain the type attribute of the coastline.

Each closed rectangular Area corresponds to a section of coastline between two intersection points, the { Area1, Area2 … Area-1 } corresponds to a section of coastline between { (q1, q2), (q2, q3) … (qk-1, pk) }, the type attribute of each section of coastline is set as the type attribute of the corresponding closed rectangular Area, and the determination of the type attribute of the whole coastline is completed.

In one example, the specific implementation manner of this step may include:

s851: for each closed rectangular area, associating a section of coastline between two of the intersections on the closed rectangular area with the closed rectangular area.

In the step, a section of coastline between { (q1, q2), (q2, q3) … (qk-1, pk) } is taken to be associated with the corresponding closed rectangular areas { Area1, Area2 … Area-1 }.

S852: for each closed rectangular area, copying the type attribute of the closed rectangular area to a section of coastline associated with the closed rectangular area.

Illustratively, the Type attribute Type of the closed rectangular Area1 is copied to a section of coastline between (q1, q2), and so on until the setting of the Type attribute of each section of coastline is completed, resulting in coastline vector data containing spatial location information and Type attribute information.

As an improvement of the present invention, the present invention may further include:

s9: and carrying out smoothing processing and topology inspection on the coastline.

The coastline is mostly in a zigzag shape, and is smoothed in order to reduce the coastline length and form errors.

The coastline does not overlap or cross, and the number of coastline end points is 2. Therefore, the coastline can be subjected to topology inspection by setting topology rules such as coastline overlapping, crossing and endpoint >2, and the topology errors of the coastline can be corrected.

S10: and counting the geographic spatial position and length of the coastline and the type attribute of each section of the coastline.

In this step, the length of the coastline is calculated and fields such as the geographical position of the coastline and the type of the coastline of each segment are updated under the projection coordinates acquired in S11, and the final coastline data is acquired.

The method comprises the steps of obtaining a satellite remote sensing image of a region to be detected, obtaining key information such as a spatial scale, a wave band number, a spectral range, a projection coordinate and a non-imaging region value of the satellite remote sensing image, and generating a mirror image covering the region to be detected; calculating an NDWI water body characteristic information layer of each scene of remote sensing image, identifying ocean and land information on each NDWI water body characteristic information layer by setting a segmentation threshold value of a water body, assigning to a mirror image after superposition operation of a plurality of ocean and land distinguishing information, and generating a coastline; the coastline and coastline land utilization classification data are superposed to generate segmentation nodes, then closed rectangular areas used for judging coastline type attributes are generated at adjacent nodes, coastline type attribute information is generated through the coastline land utilization data in the closed rectangular areas, and the coastline type attribute information is assigned to the coastline sections between the adjacent nodes of the coastline; finally, coastline vector data containing information such as geographic spatial positions, lengths and type attributes are obtained after boundary smoothing, topology processing and field attribute updating.

The invention has the following beneficial effects:

1. the image of the mirror image is generated by covering the multi-scene remote sensing images in the same area at different imaging times, the image plane covering condition of the area is represented, and the coastline double-eyelid error of different single-scene remote sensing images during the coastline extraction is reduced.

2. Calculating NDWI of the multi-view images, generating multi-view binary images of sea and land segmentation, performing statistics through pixel values of the same positions of the multi-view binary images, then assigning the image values to mirror images, and determining coastlines through the mirror images; the method comprehensively considers the remote sensing images in multiple time phases to determine the spatial position of the coastline, solves the problems of uncertain coastline position, low extraction precision and high complexity caused by the influence of tide on the conventional single remote sensing image, and has the advantages of more accurate extracted coastline, capability of being most close to the position definition of the average large tide height coastline of the coastline and no need of later tide correction work.

3. And obtaining shoreline type segmentation nodes by utilizing the coastline and the coastal zone soil utilization classification data, establishing a closed rectangular area for identifying the shoreline type attribute according to the nodes, and judging and adding the type attribute of the shoreline according to the land utilization type. The method reduces the selection time of the coastline segmentation nodes, reduces the misjudgment and the missing judgment of the coastline type attribute interpretation, improves the time efficiency, and improves the interpretation precision and efficiency of the coastline space position and the attribute characteristics.

4. Through boundary smoothing and topology processing, the accuracy of the coastline is improved, and then coastline vector data containing geographic spatial position and type attribute information is obtained after field attributes are updated.

5. The segmentation threshold of each scene image is calculated in a self-adaptive mode, so that the ocean area and the land area obtained by segmentation are more accurate, and the accuracy of the coastline is improved.

The invention is explained in detail below by means of a specific example:

1. and acquiring satellite remote sensing satellite images, coastline to-be-detected region range vectors and classification data of the coastal land utilization. The satellite remote sensing image at least comprises a space scale (resolution), a wave band number, a spectrum range, a projection coordinate and a non-imaging area value), and the method comprises the following steps:

spatial resolution 2m, band number bands 4, band1 p (blue), band2 p (green), band3 p (red), band4 p (nir), projection coordinate spatial reference WGS84 UTM51N, and non-imaging region value DataIgnoreValue 0.

2. The method comprises the steps of conducting vector cutting on a satellite remote sensing image to obtain a remote sensing image of a region to be detected (namely a target region of a coastline to be extracted), dividing the remote sensing image into a plurality of sub-regions, and generating a mirror image.

The remote sensing image of the area to be detected and the sub-area thereof are shown in fig. 2, the generated mirror image is shown in fig. 3, and the four-to-four range, the Width, the Height, the pixel size, the coordinate system and the like of the images are the same.

3. And calculating the NDWI image of the multi-scene sub-image in each sub-region, and dividing the NDWI image into an ocean image and a land image, wherein the land image is assigned to be 0, and the ocean image is assigned to be 1 to obtain a binary image.

Fig. 4 is an NDWI image of a multi-scene sub-image, and fig. 5 is a binary image obtained by assigning each scene NDWI image to land and sea.

4. According to a calculation formula of enumeration values, pixel values representing ocean and land of each pixel in the sub-area n of { sub-area 1 … } in the mirror image are calculated in a traversing manner, as shown in fig. 6.

5. According to the pixel values of the sea and land of the mirror image, the boundary line is taken as the spatial distribution position of the coastline, as shown in fig. 6.

6. The coastline is intersected by the coastline using the classification data of the coastline, and the resulting intersection points { q1, q2, q3, q4} serve as the segment nodes of the coastline, as shown in fig. 7.

7. A closed rectangular region { Area1, Area2, Area3} for determining the attribute of the coastline type is generated on the coastline neighboring segment node, as shown in fig. 7.

8. The attribute of the type of the closed rectangular area is judged in the closed rectangular area, and the value of the attribute is assigned to a corresponding section of the coastline, as shown in fig. 7.

9. And obtaining coastline vector data containing geospatial position and type attribute information after performing boundary smoothing, topology processing and field attribute updating on the obtained coastline containing the spatial position and attribute information, as shown in fig. 8.

10. And (4) carrying out statistical analysis on the space length, the type attribute and the like of the coastline data.

In summary, the invention provides a method for extracting a coastline by fusing a multi-scene remote sensing image information layer and a mirror image, and judges the coastline type attribute by dividing coastline nodes by land utilization data, aiming at the problems of high manpower consumption, incapability of ensuring precision and low efficiency of the conventional manual identification and delineation method. The invention relates to a coastline extraction algorithm with strong adaptability, which solves the problems of coastline space position extraction and attribute updating under various image imaging conditions. The method utilizes the multi-temporal image to identify the spatial position of the coastline, and overcomes the defects of uncertain position, low precision, difficult type attribute distinction and the like of the coastline caused by the influence of tide on the conventional single remote sensing image; meanwhile, the automatic study and judgment of the shoreline type attribute is realized, and the accuracy and the time efficiency of automatic interpretation of the shoreline are improved. The method can accurately, quickly and automatically extract the spatial position and the attribute of the coastline, provides basic data support for the protection and management of the natural coastline, can meet the requirements of coastline ecological destruction supervision, natural coastline retention rate accounting, coastline engineering supervision, coastline protection and management and the like by analyzing the change of the spatial position and the type attribute of the coastline extracted by the method, and has a good application prospect in the aspects of natural coastline protection, supervision and management and the like.

Example 2:

an embodiment of the present invention provides an automatic coastline identification apparatus, as shown in fig. 9, the apparatus including:

the data acquisition module 1 is used for acquiring a remote sensing image of a region to be detected and coastal zone soil utilization classification data of the region to be detected;

the remote sensing images of the area to be detected comprise multi-scene remote sensing images of the area to be detected, which are imaged at different time periods.

The mirror image generation module 2 is used for generating a mirror image according to the remote sensing image of the area to be detected;

the image range of four to four, the image width, the image height, the pixel size and the coordinate system of the mirror image are all the same as those of the remote sensing image of the area to be detected, and the initial value of each pixel of the mirror image is null.

The image segmentation module 3 is configured to segment the remote sensing image of the to-be-detected region into a plurality of sub-regions, where each sub-region includes a multi-scene sub-image of the sub-region imaged at different time periods.

The water body index calculation module 4 is used for calculating the normalized water body index image of each scene image of each subarea for each subarea;

NDWI＝(p(Green)-p(NIR))/(p(Green)+p(NIR))；

The binary segmentation module 5 is used for performing binary segmentation on the normalized water body index image corresponding to each scene image by using the segmentation threshold value of each scene image to obtain a binary image;

The mirror image assignment module 6 is configured to calculate an enumeration value of each pixel in each sub-region for each sub-region, and assign the enumeration values of each pixel of all sub-regions to pixels at the same position of the mirror image;

when in use

When M is in contact with _pixel ＝1

When in use

When, M _pixel ＝0

Wherein M is _pixel An enumerated value, N, representing a pixel (x, y) _pixel＝1 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 1, and N _pixel＝0 The number of pixel values of all binary images in a sub-area at the pixel (x, y) is 0, and N _pixel ＝N _pixel＝1 +N _pixel＝0 And P is a set interpretation threshold coefficient.

And the coastline determining module 7 is configured to use a boundary between the pixel with the pixel value of 1 and the pixel with the pixel value of 0 in the mirror image as a coastline.

And the coastline type attribute determining module 8 is used for obtaining the type attribute of the coastline by using the coastal zone land utilization classification data of the area to be detected.

The shoreline type attribute determination module includes:

and the data intersection unit is used for intersecting the coastline and the coastal zone soil utilization classification data to obtain a series of intersection points.

And the line segment determining unit is used for making a line segment perpendicular to the coastline through the intersection point at each intersection point, and the lengths of the line segment on two sides of the coastline are respectively a first length and a second length.

And the closed rectangular area determining unit is used for taking the line segment of each intersection point as one side of the rectangle and taking the next intersection point passing through the intersection point as a closed rectangle to obtain a series of closed rectangular areas.

And the closed rectangular area type attribute determining unit is used for determining the type attribute of each closed rectangular area according to the coastal zone land utilization classification data in the closed rectangular area.

The closed rectangular region type attribute determining unit includes:

and the land utilization type set determining subunit is used for taking the intersection surface of each closed rectangular area and the coastal zone land utilization classification data and acquiring a land utilization type set of the intersection surface on the coastal zone land utilization classification data.

A first determination subunit of type attribute, configured to set, if the set of land utilization types of the closed rectangular region includes "sand beach or beach", "beach or wetland or enclosure", "mangrove or seaweed bed or coral reef", and "river", the type attribute of the closed rectangular region to "sandy shoreline", "muddy shoreline", "biological shoreline", "estuary shoreline", respectively;

And the type attribute second determination subunit is used for setting the type attribute of the closed rectangular area as a "bedrock shoreline" or setting the type attribute of the closed rectangular area as a "natural shoreline" if the land utilization type set of the closed rectangular area comprises "grassland or woodland or bare rock" and the occupation ratio of the "grassland or woodland or bare rock" in the closed rectangular area is greater than a first ratio.

A type attribute third determination subunit operable to set the type attribute of the closed rectangular region to "farmland dike", "construction dike", "wharf dike", "traffic dike", "cultivation dike", and "salt field" respectively, if the set of land use types of the closed rectangular region includes "dry land or paddy field", "residential land or coast dike", "wharf or harbor or fishing port", "transportation land", "cultivation", "salt field", and the occupation ratio of the "dry land or paddy field", "residential land or seacoast dike", "wharf or harbor", "transportation land", "cultivation", "salt field" in the closed rectangular region is larger than the second ratio.

Or if the land use type set of the closed rectangular area comprises land use types influenced by artificial activities, setting the type attribute of the closed rectangular area as an artificial shoreline.

The shoreline type attribute determination unit includes:

a data association subunit, configured to associate, for each closed rectangular area, a section of coastline between two of the intersections on the closed rectangular area with the closed rectangular area.

And the type attribute replication unit is used for replicating the type attribute of each closed rectangular region to a section of coastline associated with the closed rectangular region.

For any one of the foregoing scene images, the segmentation threshold can be determined by the following process:

and traversing each value of the segmentation threshold value on the value interval of the normalized water body index according to a set step length.

And comparing the pixel value of each pixel of the normalized water body index image corresponding to the sub-image with the value of the segmentation threshold, and classifying the pixels of which the pixel values are greater than the value of the segmentation threshold into a first region, otherwise, classifying the pixels into a second region.

Calculating the difference g between a first area and a second area corresponding to the value of each segmentation threshold value for the value of each segmentation threshold value;

g＝w ₀ w ₁ (μ ₀ -μ ₁ ) ²

wherein, N ₀ And N ₁ The number of pixels of the first region and the second region respectively，μ ₀ Is the average value of pixel values of all pixels of the first region, mu ₁ Is the average value of the pixel values of all the pixels of the second area.

And finding out the maximum value of the difference corresponding to the values of all the segmentation threshold values, wherein the value of the segmentation threshold value corresponding to the maximum value is the segmentation threshold value of the sub-image.

The data acquisition module comprises:

and the data acquisition unit is used for acquiring the satellite remote sensing image, the range vector of the area to be detected and the classification data of the coastal land utilization.

And the coordinate transformation unit is used for carrying out coordinate transformation for enabling the projection coordinates to be consistent when the satellite remote sensing image, the range vector of the area to be detected and the projection coordinates of the coastal earthy utilization classification data are inconsistent.

And the cutting unit is used for cutting the satellite remote sensing image by using the range vector of the area to be detected to obtain the remote sensing image of the area to be detected.

The shoreline determination module comprises:

and the merging and classifying unit is used for merging the pixels with the pixel value of 1 in the mirror image into a first object and merging the pixels with the pixel value of 0 in the mirror image into a second object.

A coastline determining unit for regarding a boundary of the first object and the second object as the coastline.

The apparatus may further include:

and the correction module is used for carrying out smoothing processing and topology inspection on the coastline.

And the statistical module is used for carrying out statistics on the geographic spatial position and the length of the coastline and the type attribute of each section of coastline.

The invention provides a method for extracting a coastline by fusing a multi-scene remote sensing image information layer and a mirror image, and judges the coastline type attribute by dividing coastline nodes by land utilization data aiming at the problems of high labor consumption, incapability of ensuring precision and low efficiency of the conventional manual identification and delineation method. The invention relates to a coastline extraction algorithm with strong adaptability, which solves the problems of coastline space position extraction and attribute updating under various image imaging conditions. The method utilizes the multi-temporal image to identify the spatial position of the coastline, and overcomes the defects of uncertain position, low precision, difficult type attribute distinction and the like of the coastline caused by the influence of tide on the conventional single remote sensing image; meanwhile, the automatic study and judgment of the shoreline type attribute is realized, and the accuracy and the time efficiency of automatic interpretation of the shoreline are improved. The method can accurately, quickly and automatically extract the spatial position and attribute of the coastline, provides basic data support for the protection and management of the natural coastline, can meet the requirements of coastline ecological destruction supervision, natural coastline survival rate accounting, coastline engineering supervision, coastline protection and management and the like by analyzing the change of the spatial position and type attribute of the coastline extracted by the method, and has better application prospects in the aspects of natural coastline protection, supervision and management and the like.

The device provided by the embodiment of the present invention has the same implementation principle and technical effect as the method embodiment, and for the sake of brief description, reference may be made to the corresponding content in the method embodiment 1 without reference to the device embodiment. It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the apparatus and the unit described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

Finally, it should be noted that: the above-mentioned embodiments are merely specific embodiments of the present invention, which are used to illustrate the technical solutions of the present invention, but not to limit the technical solutions, and the scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the above-mentioned embodiments. Those of ordinary skill in the art will understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention. Are intended to be covered by the scope of the present invention.

Claims

1. A coastline automated identification method, the method comprising:

s1: acquiring a remote sensing image of a region to be detected and coastal zone land utilization classification data of the region to be detected;

NDWI＝(p(Green)-p(NIR))/(p(Green)+p(NIR))；

when in use

When M is in contact with _pixel ＝1

When in use

When M is in contact with _pixel ＝0

s8: obtaining the type attribute of the coastline by utilizing the coastal zone land utilization classification data of the area to be detected;

the S8 includes:

2. The coastline automated identification method of claim 1, wherein the S84 comprises:

or if the land utilization type set of the closed rectangular area comprises land utilization types which are not influenced by human activities, setting the type attribute of the closed rectangular area as a 'natural shoreline';

3. The coastline automated identification method of claim 2, wherein the S85 comprises:

4. The coastline automatic identification method of any one of claims 1 to 3, wherein the segmentation threshold for any one of the scene images is determined by:

g＝w ₀ w ₁ (μ ₀ -μ ₁ ) ²

wherein N is ₀ And N ₁ The number of pixels, mu, of the first region and the second region respectively ₀ Is the average value of pixel values of all pixels of the first region, mu ₁ The average value of the pixel values of all the pixels of the second area is obtained;

5. The coastline automated identification method of claim 4, wherein the S1 comprises:

6. The coastline automated identification method of claim 5, wherein the S7 comprises:

7. The coastline automated identification method of claim 6, further comprising:

s9: carrying out smoothing processing and topology inspection on the coastline;

8. An automatic coastline identification device, the device comprising:

the image segmentation module is used for segmenting the remote sensing image of the area to be detected into a plurality of sub-areas, and each sub-area comprises a plurality of scene sub-images of the sub-area imaged at different time periods;

NDWI＝(p(Green)-p(NIR))/(p(Green)+p(NIR))；

when in use

When M is in contact with _pixel ＝1

When in use

When, M _pixel ＝0

the coastline type attribute determining module is used for obtaining the type attribute of the coastline by utilizing the coastal zone land utilization classification data of the area to be detected;

the shoreline type attribute determination module includes: