CN108550174B - Coastline super-resolution mapping method and coastline super-resolution mapping system based on semi-global optimization - Google Patents
Coastline super-resolution mapping method and coastline super-resolution mapping system based on semi-global optimization Download PDFInfo
- Publication number
- CN108550174B CN108550174B CN201810189532.2A CN201810189532A CN108550174B CN 108550174 B CN108550174 B CN 108550174B CN 201810189532 A CN201810189532 A CN 201810189532A CN 108550174 B CN108550174 B CN 108550174B
- Authority
- CN
- China
- Prior art keywords
- coastline
- image
- sub
- pixel
- initial
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000013507 mapping Methods 0.000 title claims abstract description 25
- 238000005457 optimization Methods 0.000 title claims abstract description 18
- 230000008859 change Effects 0.000 claims abstract description 35
- 230000004927 fusion Effects 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 238000000605 extraction Methods 0.000 claims description 19
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 16
- 238000012937 correction Methods 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 4
- 230000004807 localization Effects 0.000 claims description 2
- 238000002474 experimental method Methods 0.000 description 9
- 238000003709 image segmentation Methods 0.000 description 8
- 238000011161 development Methods 0.000 description 6
- 230000018109 developmental process Effects 0.000 description 6
- 238000010586 diagram Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000001514 detection method Methods 0.000 description 3
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000003708 edge detection Methods 0.000 description 3
- 239000000284 extract Substances 0.000 description 3
- 230000004044 response Effects 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000009792 diffusion process Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 239000013049 sediment Substances 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000002902 bimodal effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000010339 dilation Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000013535 sea water Substances 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000002689 soil Substances 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000010792 warming Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T11/00—2D [Two Dimensional] image generation
- G06T11/20—Drawing from basic elements, e.g. lines or circles
- G06T11/203—Drawing of straight lines or curves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4053—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution
- G06T3/4069—Scaling of whole images or parts thereof, e.g. expanding or contracting based on super-resolution, i.e. the output image resolution being higher than the sensor resolution by subpixel displacements
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration using two or more images, e.g. averaging or subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
- G06T7/337—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods involving reference images or patches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20112—Image segmentation details
- G06T2207/20164—Salient point detection; Corner detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention provides a coastline super-resolution mapping method and a coastline super-resolution mapping system based on semi-global optimization. The method mainly comprises the following steps: acquiring an initial coastline image and a reference image, performing image preprocessing, image fusion and image registration, combining the change trend of the overall coastline form with the gray level change of the periphery of the initial coastline, extracting a coastline change control point, and dividing the initial coastline into a plurality of coastline segments by using the coastline change control point; in each section, obtaining a sub-pixel positioning result in a coastline neighborhood window, and fitting the sub-pixel positioning result into a smooth curve section by combining sub-pixel positioning coordinates of all points in one section; combining all the curve segments together to obtain a complete coastline vector result, and completing super-resolution drawing of the coastline; the invention provides a sub-pixel positioning method based on a local area, which can adapt to a coastline environment with high curvature, has high positioning precision, can adapt to coastlines with different curvatures, and improves the accuracy of results.
Description
Technical Field
The invention belongs to the field of coastline image processing, and relates to a coastline super-resolution mapping method and system based on semi-global optimization.
Background
The coastline is the boundary between the ocean and land and is one of the most important 27 surface features. The coastline is influenced by natural factors and human factors, resulting in changes in form and properties. The sea water power action (including tidal current and the like), geological meteorological disasters, sea level rise caused by climate warming and other natural condition changes cause the coast to be silted in and eroded away; the influence of human activities such as human reclamation, sea reclamation, ocean engineering and the like leads to the obvious change of the trend and the length of the coastline while the type of the coastline is changed.
Therefore, coastline mapping and change detection have been fundamental works for coast erosion monitoring, coastal zone resource management, coastal area environmental protection, and coastal area sustainable development. The method has the advantages that ocean resources in China are abundant, the coastline utilization degree is high, the fast and accurate extraction of the coastline is researched, comprehensive planning and treatment of the coastline in China and reasonable development and utilization of resources are facilitated, and the method has important practical significance for sustainable development of society, economy and nature in China. The main methods for coastline extraction are an image segmentation method and an edge detection method. The image segmentation method divides image data into a foreground part and a background part, and the boundary of the foreground part and the background part is the extracted coastline.
There are many image segmentation algorithms, such as extracting coastlines from SAR images using a level set model (certomere et. 2009; Ouyang, Chong, and wu.2010; Shu, Li, and gomes.2010); extracting a coastline based on a geometric active contour model (Niedermeier, Romaneessen, and Lehner.2000; Xing, Fu, and Zhou.2012; Zhang et al.2013); extracting a coastline (Di et al 2003; Liu et al 2011) from the high-resolution image by combining an image segmentation method and image local features; an integrated supervised and unsupervised classification method was used for coastline extraction (Sekovski et al 2014).
The edge detection method extracts the coastline by searching for energy intensity (luminance) discontinuities between adjacent pixels (Buono et al.2014; Fugura, Billa, and Pradhan 2011; Zhang et al.2013b).
Other methods include extracting shorelines from high resolution images using mathematical morphology (Puissant et al 2008); based on photogrammetry technology, a coastline (Lira et. 2016) is extracted by using a digital aerial photograph and a digital orthophoto; and automatic detection of coastlines (valencini.2017) based on remote sensing video systems.
To date, most research has been directed to finding the best way to determine the position of the coastline. Since most methods are based on hard classification, the coastline has only pixel-level positioning accuracy and cannot meet the actual accuracy requirement. Some scholars extract the coastline by using the high-resolution remote sensing image, and although the coastline with high precision can be obtained, the expensive price of the coastline cannot meet the extraction of the coastline with a large range in actual demands. Therefore, a method for extracting a high-precision coastline by using a medium-low resolution remote sensing image is needed. With the development of image processing technology, especially the development of sub-pixel positioning and super-resolution reconstruction technology, some scholars use the image processing technology in the coastline extraction of medium-low resolution remote sensing images, so that the extracted coastline has higher precision.
Foody et al (2005) evaluated soft classification methods of coastline from degraded simulated Landsat images, and the coastline and RMSE presented at the sub-pel scale were 2.25 meters. Pardo-pascal et al (2012) propose an automated method to extract the coastline of sub-pel accuracy from Landsat TM and ETM +, with the RMSE of the coastline position from 4.69 to 5.47 m. As the natural beach changes over time, Almonacid-cabaler et al (2016) applies the annual average coastline position extracted from Landsat images to substantially reduce short term variations of the coastline, and the extracted coastline deviates from the sea surface by about 4 to 5 meters.
These methods are often complex and difficult to implement in practical applications, and have certain limitations, such as being affected by suspended silt on the shore, being unable to adapt to the coastline environment of high curvature, being unable to remove the influence of the berthed vessel, etc.
Most methods are based on hard classification, so the coastline only has the positioning precision of a pixel level and cannot meet the requirement of actual precision. In the method for extracting the coastline at the pixel level, when the coastline is extracted by an image segmentation algorithm, in order to ensure the extraction precision, more post-processing steps are needed to determine the boundary pixels and the threshold value; the region growing algorithm must consider many criteria and thresholds, including splitting, merging, and starting point selection; a drawback of edge detection based methods is that these methods produce discontinuous lines that do not represent coastlines well. Some scholars extract the coastline by using the high-resolution remote sensing image, and although the coastline with high precision can be obtained, the expensive price of the coastline cannot meet the extraction of the coastline in a wide range in actual demands. Therefore, a method for extracting a high-precision coastline by using a medium-low resolution remote sensing image is needed. However, the soft classification method is often complex and difficult to implement in practical application, and has certain limitations, such as being affected by suspended silt on the shore, being unable to adapt to the coastline environment with high curvature, being unable to remove the influence of the berthed ship, and the like.
Disclosure of Invention
Aiming at the defects and shortcomings of the prior art, the invention provides a coastline super-resolution mapping method and system based on semi-global optimization, which are used for solving the technical problems that the positioning precision is not high and the coastline environment with high curvature cannot be adapted. The method mainly comprises the following steps:
s1, acquiring a Landsat8 land imager image and a GF-2 reference image, and respectively carrying out image preprocessing on the Landsat8 land imager image and the GF-2 reference image to obtain a Landsat8 fusion image, a water body index gray scale image corresponding to the Landsat8 fusion image and a GF-2 fusion image;
s2, taking the GF-2 fused image as a reference image, and aligning the reference image and the Landsat8 fused image to obtain an offset parameter between the reference image and the Landsat8 fused image;
s3, extracting a pixel-level coastline from the water body index gray-scale image to obtain an initial coastline; carrying out coastline extraction on the reference image to obtain a reference coastline;
s4, extracting a coastline change control point of the initial coastline to obtain an initial coastline change control point, and segmenting the initial coastline through the initial coastline change control point to obtain a plurality of segmented coastlines;
s5, performing sub-pixel positioning based on a local area on each pixel point in each section of coastline to obtain sub-pixel positioning coordinates of each point in each section of coastline; performing offset correction on the sub-pixel positioning coordinate of each point through the offset parameter in the step S2;
s6, carrying out coastline least square fitting on the sub-pixel positioning coordinates of all the points in the same coastline after deviation correction, and fitting the sub-pixel positioning coordinates into a smooth curve segment; and combining all the curve segments together to obtain a complete coastline vector result, and completing the super-resolution drawing of the coastline.
The coastline super-resolution mapping method based on semi-global optimization further comprises the following steps: and respectively calculating position errors between the sub-pixel positioning coordinates of the coastline and a reference coastline and between the coastline vector result and the reference coastline, and analyzing error results.
In the coastline super-resolution mapping method based on semi-global optimization, the sub-pixel positioning based on the local region in the step S5 comprises the following steps:
s51, calculating the edge direction of each pixel point in each section of initial coastline, and determining different neighborhood windows of each pixel point of the initial coastline according to different edge directions;
and S52, describing the boundary of the field window into a curve through a fitting function, calculating parameters of the fitting function, and determining the coordinates of the sub-pixel positioning points of the local region.
In the coastline super-resolution mapping method based on semi-global optimization, the coastline minimum two-times fitting in the step S6 comprises the following steps:
s61, detecting and extracting sub-pixel positioning coordinates on each section of coastline;
and S62, dividing the sub-pixel positioning coordinates into different point sets, and performing least-squares fitting on the sub-pixel positioning coordinates in the same point set to obtain sub-pixel level coastlines.
Preferably, the invention also provides a coastline super-resolution mapping system based on semi-global optimization, which comprises the following modules:
the image preprocessing image fusion module is used for acquiring a Landsat8 land imager image and a GF-2 reference image, and respectively preprocessing the Landsat8 land imager image and the GF-2 reference image to obtain a Landsat8 fusion image, a water index gray-scale image corresponding to the Landsat8 fusion image and a GF-2 fusion image;
the image registration module is used for taking the GF-2 fused image as a reference image, registering the reference image and the Landsat8 fused image and obtaining an offset parameter between the reference image and the Landsat8 fused image;
the coastline extraction module is used for extracting the pixel level coastline of the water body index gray level image to obtain an initial coastline; carrying out coastline extraction on the reference image to obtain a reference coastline;
extracting a coastline change control point, wherein the coastline change control point is extracted from the initial coastline to obtain an initial coastline change control point, and the initial coastline is segmented by the initial coastline change control point to obtain a plurality of segmented coastlines;
the sub-pixel positioning module is used for performing sub-pixel positioning based on a local area on each pixel point in each section of coastline to obtain a sub-pixel positioning coordinate of each point in each section of coastline; offset correction is carried out on the sub-pixel positioning coordinate of each point through the offset parameters of the image registration module;
the positioning point fitting module is used for performing coastline least square fitting on the sub-pixel positioning coordinates of all the points in the same coastline, which are subjected to offset correction, and fitting the coastline least square fitting into a smooth curve segment; and combining all the curve segments together to obtain a complete coastline vector result, and completing the super-resolution drawing of the coastline.
The coastline super-resolution mapping system based on semi-global optimization further comprises an error analysis module: and the method is used for respectively calculating the position errors between the sub-pixel positioning coordinates of the coastline and the reference coastline and between the coastline vector result and the reference coastline and analyzing the error results.
In the coastline super-resolution mapping system based on semi-global optimization, the sub-pixel positioning module based on the local area comprises the following modules:
the neighborhood window determining module is used for calculating the edge direction of each pixel point in each section of the initial coastline and determining different neighborhood windows of each pixel point of the initial coastline according to different edge directions;
and the sub-pixel positioning point coordinate acquisition module is used for describing the field window boundary into a curve through a fitting function, calculating parameters of the fitting function and determining the coordinates of the sub-pixel positioning points in the local area.
In the coastline super-resolution mapping system based on semi-global optimization, the coastline least square fitting in the positioning point fitting module comprises the following modules:
the sub-pixel positioning coordinate extracting module is used for detecting and extracting sub-pixel positioning coordinates on each section of coastline;
and the sub-pixel level coastline acquisition module is used for dividing the sub-pixel positioning coordinates into different point sets, and performing least square fitting on the sub-pixel positioning coordinates in the same point set to obtain the sub-pixel level coastline.
The invention provides a coastline super-resolution mapping method and a coastline super-resolution mapping system based on semi-global optimization, wherein an initial coastline image and a reference image are obtained, image preprocessing, image fusion and image registration are carried out, the change trend of the whole coastline state is combined with the gray level change of the periphery of the initial coastline, a coastline change control point is extracted, and the initial coastline is divided into a plurality of coastline segments by using the coastline change control point; in each section, obtaining a sub-pixel positioning result in a coastline neighborhood window, and fitting the sub-pixel positioning result into a smooth curve section by combining sub-pixel positioning coordinates of all points in one section; combining all the curve segments together to obtain a complete coastline vector result, and completing super-resolution drawing of the coastline; the invention provides a sub-pixel positioning method based on a local area, which can adapt to a coastline environment with high curvature, has high positioning precision, can adapt to coastlines with different curvatures, and improves the accuracy of results.
Drawings
The invention will be further described with reference to the accompanying drawings and examples, in which:
FIG. 1 is a flow chart of a method according to an embodiment of the present invention;
FIG. 2 is an image of experiment area 1 according to an embodiment of the present invention;
FIG. 3 is an image of experiment area 2 according to the present invention;
FIG. 4 is an initial shoreline extraction diagram according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of the inflection point extraction of a coastline according to an embodiment of the present invention;
FIG. 6 is a diagram illustrating the positioning result of sub-pixels according to the embodiment of the present invention;
FIG. 7 is a piecewise fit graph of an embodiment of the present invention;
FIG. 8 is a graph showing the results in experimental section 1 according to the example of the present invention;
FIG. 9 is a graph showing the results in experiment zone 2 of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail below with reference to the accompanying drawings and examples.
The invention provides a coastline super-resolution mapping method and a coastline super-resolution mapping system based on semi-global optimization, wherein a complete flow chart of the method is shown in a figure 1, and a research area of an embodiment of the invention is divided into two parts: the canochien dian harbor and mansion-quanzhou peripheral coastal areas. The Caofei Dian harbor is located at 118.5 DEG E and around 39 DEG N, is adjacent to Beijing jin Ji city group in China, and is one of the important ore transportation ports in China. The caochien county is only one sand island in the past and becomes a port after artificial construction, so that the main shoreline type of the caochien county is an artificial shoreline, and the position of the shoreline is stable for a long time and is not changed for many years. The Xiamen-Quanzhou peripheral experimental area is located between 118 degrees E-118.5 degrees E and 24.35 degrees N-24.6 degrees N, and is close to Taiwan strait, and the Xiamen is an important port for international economic and cultural communication. The increasingly close connection with the world pushes the development of buildings and quays around the coast, which leads to the rapid change of the position of the coastline in recent decades. The shoreline types around mansion doors and quanzhou mainly include bedrock shores, artificial shores and flat sandy shores.
In the study of the inventive example, all experimental images were downloaded from the USGS database, and the images were acquired by landlocked 8 imager sensors. Landolt-8 land imager imagery detailed parameters are shown in table 1 using WGS84 ellipsoid model Using Transverse Mercator (UTM) projection. Table 1 summarizes the error statistics (mean error, standard deviation) for each date, each experimental area and each data type. The average error is obtained by averaging all errors, and since all errors are obtained by calculating the absolute value of the distance from the final coastline point to the reference coastline, the average error is used to interpret the level of deviation from the reference coastline. Standard deviation (STDEV) represents the variability around the mean error.
TABLE 1 Landsat-8 land imager parameter table
The hong Kong of Caofei is located in a block of 122 rows-033 columns, 12 images of the area from 2013 to 2016 are processed, and three artificial coastlines with different water and soil distributions are selected from each image to serve as an experimental area 1 for determining a wave band combination mode with the best noise resistance, as shown in FIG. 2. The xiamen-quan perimeter coastal region was acquired from the block of column 119 row-043, the image acquisition date was 33 minutes (UTC) at 2 am on day 13 month 10 of 2015, and 5 coastlines with different curvatures were selected from the image, as shown in fig. 3, for verifying the algorithm's universality.
In addition, a reference coastline covering the study area was extracted using high-resolution number 2 optical remote sensing satellite (GF-2) images, as shown. The high-resolution second satellite is successfully transmitted in 2014 within 19 months, is the first civil optical remote sensing satellite with spatial resolution superior to 1 meter in China, is provided with two high-resolution 1-meter panchromatic and 4-meter multispectral cameras, and has the characteristics of sub-meter spatial resolution, high positioning precision, quick attitude maneuvering capability and the like. Since the three experimental regions in hong Kong of Caofiedian are artificial coastlines and the positions are stable for a long period of time and are not affected by sea tides, GF-2 images obtained at 05, month and 31 of 2015 were used as reference images. And a sandy coast and a bedrock coast exist in the coastal area around the Xiamen-Quanzhou, and are easily influenced by sea tides, so that a high-resolution No. 2 image close to the acquisition time of an experimental area is selected as a reference image to eliminate or reduce the influence of tides on the result precision. Since the spatial resolution of the top-scoring No. 2 image is 1 meter/pixel, the uncertainty of the coastline reference location is estimated to be + -1.5 meters. Table 1 lists more information about GF-2 images.
1. And (5) image preprocessing. Due to the sensor, the atmosphere, the terrain and other reasons, the remote sensing satellite generates errors in the data acquisition process, the image quality is affected by the errors, and the image precision is affected, so that the Landsat8 OLI experiment image and the GF-2 reference image need to be subjected to image preprocessing respectively before the experiment. Firstly, the RPC parameters carried by GF-2 image data are utilized to respectively carry out orthorectification on GF-2 image multispectral data and panchromatic data. Because the multispectral data and the panchromatic data are well geographically registered after the high-resolution data orthorectification, image registration is not needed, image fusion is directly carried out on the GF-2 reference image by adopting a Neorest neighboring Diffusion (NNDiffuse) pan sharing algorithm, and the spatial resolution of the GF-2 reference image after fusion is 1 m. And then performing image fusion on the Landsat8 OLI experimental image by adopting a Neorest Neighbor Diffusion (NNDiffuse) pan sharpening algorithm, wherein the spatial resolution of the Landsat8 OLI experimental image after fusion is 15 m. And more pixels can be obtained on the coastline with the same length at a smaller sampling interval, and more primary coastline pixels are favorable for gradient calculation and accuracy of a sub-pixel positioning result, and the smaller the area of a neighborhood window is, the more the principle of ground object proximity is met. And then, carrying out radiometric calibration on the Landsat8 OLI experimental image through radiometric calibration parameters to obtain radiance value data. And then, carrying out atmospheric correction (carrying out rapid atmospheric correction on Landsat8 OLI experimental images) on the radiation brightness value data by using a FLAASH model, and eliminating the influence of atmospheric scattering on the spectrum.
The invention fully utilizes the spectral characteristics of multispectral remote sensing data to respectively calculate the normalized difference water body index (NDWI) and the improved normalized difference water body index (MNDWI), and the specific expressions are shown in a formula 1 and a formula 2. The two water body indexes can inhibit image noise, and the image histogram is a bimodal histogram and meets the initial assumption of a subsequent algorithm. In consideration of the effect of the shore-side suspended sediment on the results, the Short Wave Infrared (SWIR) band is selected as another initial image because the Short Wave Infrared (SWIR) wavelength is longer and can penetrate through the suspended sediment. These options are consistent with the authors of other similar studies.
NDWI ═ (p (Green) -p (NIR))/(p (Green) + p (NIR)) (formula 1)
MNDWI ═ (p (Green) -p (MIR))/(p (Green)) + p (MIR)) (formula 2)
2. And (5) image registration. The geometric displacement between the Landsat8 fused image and the GF-2 fused image (reference image) is calculated. Because the image resolutions of the Landsat8 image and the GF-2 fused image (reference image) are different, a sift algorithm with scale invariance is selected for image registration to obtain the geometric displacement (dx, dy) between the Landsat8 image and the reference image.
3. And (4) extracting the coastline. The extraction schematic diagram is shown in fig. 4, because the spectral responses of water and land in different wave bands are different, the histograms of the three different wave band combined images show bimodality and meet the initial assumption of the OTSU algorithm, and therefore the maximum inter-class variance method (OTSU) is adopted for image segmentation. The OTSU algorithm divides the original image into a foreground part and a background part by calculating an optimal threshold value. Assuming that the threshold is T, the optimal threshold T may be obtained by the following algorithm:
W0＝N0/(N0+N1)
W1＝N1/(N0+N1)
W0+W1＝1
U＝W0*U0+W1*U1
G＝W0*(U0-U)2+W1*(U1-U)2
wherein the number of the pixels with the pixel gray value less than the threshold value T is N0The number of pixels with the pixel gray level larger than the threshold value T is N1，W0The number of foreground pixels accounts for the image proportion, U0Is the foreground mean gray, W1The number of background pixels accounts for the proportion of the image, U1The average gray level of the background, U, and G are the average gray levels of the whole image and the inter-class variance.
When the inter-class variance G is maximum, the threshold T at this time is the optimal threshold T. The larger the inter-class variance between the background and the foreground is, the larger the difference between the two parts constituting the image is, and the smaller the difference between the two parts is caused when part of the foreground is mistaken for the background or part of the background is mistaken for the foreground. Therefore, the optimal threshold T that maximizes the inter-class variance can obtain an image segmentation result with the least probability of misclassification.
Then, the segmentation result is processed by using a region growing algorithm and a morphological dilation filter, so that impurities in land and water are removed, and finally a series of pixels representing the position of the initial coastline are obtained.
4. And extracting a coastline change control point. And carrying out coastline change control point extraction on the initial coastline to obtain initial coastline change control points, and segmenting the initial coastline through the initial coastline change control points to obtain a plurality of segmented coastlines. Extracting an initial coastline change control point by using a Harris angular point detection algorithm, namely calculating a Harris response value R of each pixel on the initial coastline in 3; and selecting a local maximum value in the R by threshold judgment and neighborhood inhibition screening and combining the overall form of the initial coastline, wherein the local maximum value is the change control point on the initial coastline. When the Harris response value R is calculated, the pixels in the initial coastline are sequentially calculated according to the connection relation, and therefore the acquired positions of the change control points are sequentially stored according to the connection relation. The coastline change control point extraction diagram is shown in fig. 5.
5. And (5) sub-pixel positioning and deviation correction. On the basis of 4, the initial coastline point (pixel level) in each segment coastline is sub-pixel positioned. The sub-pixel positioning algorithm based on the gray level of the neighborhood pixels is based on an assumption that: according to the proximity principle, the grey value of each pixel is weighted by the grey values of the different objects in the pixel, and the probability of the same object between adjacent pixels is maximal. There are only two types of terrain, land and water, surrounding the preliminary coastline pixels, so we can consider the gray value of each preliminary coastline pixel to be weighted by the neighboring pure pixels. Therefore, the sub-pel positioning algorithm is divided into two stages-determining the size of the neighborhood window and calculating the sub-pel coordinate position. The sub-pixel localization result is shown in FIG. 6.
Therefore, on the basis of a waterway binary image of initial image segmentation, the edge direction of each initial coastline pixel is calculated, different neighborhood windows of the initial coastline pixels are given according to different edge directions, when the absolute value of the edge direction is less than or equal to 1, a neighborhood window of 5 × 3 is used, and when the absolute value of the edge direction is greater than 1, a neighborhood window of 3 × 5 is used.
After the confirmation of the neighborhood window, the sub-pixel coordinate position is calculated. The boundary in the field window is described as a curve through a fitting function, and the coordinate position of the sub-pixel positioning point is determined by calculating the parameters of the fitting function. And (3) carrying out error correction on the geographic coordinates of the sub-pixel coastline positioning point based on the geometric displacement (dx, dy) between the Landsat8 fusion image and the reference image calculated in the step 2.
6. And fitting the positioning points. Since most of the coastlines are basically a smooth line segment except the corner points (the coastline change control points), the quadratic fitting of the least square method can be performed on the sub-pixel level positioning points in each segmented coastline (namely, the positioning results under the sub-pixel level scale are corrected by using the whole form of the coastline under the pixel level scale). Regarding the sub-pixel positioning points extracted from each segment of the segmented coastline in 5 as a point set, we can consider that the curvatures of the sub-pixel positioning points in the same point set have no large change, and then perform the least-squares fitting on the sub-pixel positioning points in the same point set, wherein a schematic diagram of the piecewise fitting is shown in fig. 7. And setting connectivity constraint at the joint of adjacent line segments, and performing smooth processing to obtain a complete continuous sub-pixel-level coastline so as to realize super-resolution drawing. The results of experiment 1 are shown in FIG. 8, and the results of experiment 2 are shown in FIG. 9.
7. And (5) error analysis. And respectively calculating position errors between the sub-pixel positioning coordinates of the coastline and the reference coastline and between the coastline vector result and the reference coastline, and analyzing error results. The results of the error analysis are shown in tables 2 and 3.
TABLE 2 error statistics
TABLE 3 error statistics
Table 2 summarizes the error statistics (mean error, standard deviation) for each date, each experimental area and each data type. The average error is obtained by averaging all errors, and since all errors are obtained by calculating the absolute value of the distance from the final coastline point to the reference coastline, the average error is used to interpret the level of deviation from the reference coastline. Standard deviation (STDEV) represents the variability around the mean error.
As can be seen from Table 3, the sub-pixel positioning algorithm based on local adjacent domain gray scale effectively improves the accuracy of coastline extraction, the average error value is between 3.15 and 5.87, the standard deviation is between 1.07 and 2.06, and the improvement range of positioning accuracy is between 30% and 60%. And comparing the sub-pixel coastline precisions of three different wave band combinations (selections). Obviously, the coastline obtained by MNDWI data has the highest precision, the total average error value is the smallest, and most average errors are the smallest in different fields. This indicates that MNDWI is the water body index for particularly optimal noise immunity. By combining the results of the experiment area 1 and the experiment area 2, the semi-global super-resolution mapping method provided by the invention has good universality and can adapt to coastlines with different curvatures.
While the present invention has been described with reference to the embodiments illustrated in the drawings, the present invention is not limited to the embodiments, which are illustrative and not restrictive, and it will be apparent to those skilled in the art that various changes and modifications can be made herein without departing from the spirit and scope of the invention as defined in the appended claims.
Claims (4)
1. A coastline super-resolution mapping method based on semi-global optimization is characterized by comprising the following steps:
s1, acquiring a Landsat8 land imager image and a GF-2 reference image, and respectively carrying out image preprocessing on the Landsat8 land imager image and the GF-2 reference image to obtain a Landsat8 fusion image, a water body index gray-scale image corresponding to the Landsat8 fusion image and a GF-2 fusion image;
s2, taking the GF-2 fused image as a reference image, and registering the reference image and the Landsat8 fused image to obtain an offset parameter between the reference image and the Landsat8 fused image;
s3, extracting a pixel-level coastline from the water body index gray-scale image to obtain an initial coastline; carrying out coastline extraction on the reference image to obtain a reference coastline;
s4, extracting a coastline change control point of the initial coastline to obtain an initial coastline change control point, and segmenting the initial coastline through the initial coastline change control point to obtain a plurality of segmented coastlines;
s5, performing sub-pixel positioning based on a local area on each pixel point in each section of coastline to obtain sub-pixel positioning coordinates of each point in each section of coastline; performing offset correction on the sub-pixel positioning coordinate of each point through the offset parameter in the step S2;
s6, carrying out coastline least square fitting on the sub-pixel positioning coordinates of all the points in the same coastline after deviation correction, and fitting the sub-pixel positioning coordinates into a smooth curve segment; combining all the curve segments together to obtain a complete coastline vector result, and completing super-resolution drawing of the coastline;
the local area based sub-pixel localization in step S5 comprises the steps of:
s51, calculating the edge direction of each pixel point in each section of initial coastline, and determining different neighborhood windows of each pixel point of the initial coastline according to different edge directions;
s52, describing the boundary of the adjacent domain window into a curve through a fitting function, calculating parameters of the fitting function, and determining coordinates of sub-pixel positioning points of a local region;
the shoreline least squares fit in step S6 comprises the steps of:
s61, detecting and extracting sub-pixel positioning coordinates on each section of coastline;
and S62, dividing the sub-pixel positioning coordinates into different point sets, and performing least square fitting on the sub-pixel positioning coordinates in the same point set to obtain a sub-pixel level coastline.
2. The coastline super-resolution mapping method based on semi-global optimization according to claim 1, further comprising: and respectively calculating position errors between the sub-pixel positioning coordinates of the coastline and a reference coastline and between the coastline vector result and the reference coastline, and analyzing error results.
3. The coastline super-resolution mapping system based on semi-global optimization is characterized by comprising the following modules:
the image preprocessing image fusion module is used for acquiring a Landsat8 land imager image and a GF-2 reference image, and respectively preprocessing the Landsat8 land imager image and the GF-2 reference image to obtain a Landsat8 fusion image, a water index gray-scale image corresponding to the Landsat8 fusion image and a GF-2 fusion image;
the image registration module is used for taking the GF-2 fused image as a reference image, registering the reference image and the Landsat8 fused image and obtaining an offset parameter between the reference image and the Landsat8 fused image;
the coastline extraction module is used for extracting the pixel level coastline of the water body index gray level image to obtain an initial coastline; carrying out coastline extraction on the reference image to obtain a reference coastline;
extracting a coastline change control point, wherein the coastline change control point is extracted from the initial coastline to obtain an initial coastline change control point, and the initial coastline is segmented by the initial coastline change control point to obtain a plurality of segments of segmented coastlines;
the sub-pixel positioning module is used for performing sub-pixel positioning based on a local area on each pixel point in each section of coastline to obtain a sub-pixel positioning coordinate of each point in each section of coastline; offset correction is carried out on the sub-pixel positioning coordinate of each point through the offset parameters of the image registration module;
the positioning point fitting module is used for performing coastline least square fitting on the sub-pixel positioning coordinates of all the points in the same coastline after deviation correction to fit the coastline least square fitting into a smooth curve segment; combining all the curve segments together to obtain a complete coastline vector result, and completing super-resolution drawing of the coastline;
the sub-pel positioning based on local area in the sub-pel positioning module comprises the following modules:
the neighborhood window determining module is used for calculating the edge direction of each pixel point in each section of the initial coastline and determining different neighborhood windows of each pixel point of the initial coastline according to different edge directions;
the sub-pixel positioning point coordinate acquisition module is used for describing the boundary of the adjacent domain window into a curve through a fitting function, calculating parameters of the fitting function and determining the coordinate of the sub-pixel positioning point of the local region;
the shoreline least squares fit in the anchor point fitting module comprises the following modules:
the sub-pixel positioning coordinate extracting module is used for detecting and extracting sub-pixel positioning coordinates on each section of coastline;
and the sub-pixel level coastline acquisition module is used for dividing the sub-pixel positioning coordinates into different point sets, and performing least square fitting on the sub-pixel positioning coordinates in the same point set to obtain the sub-pixel level coastline.
4. The coastline super-resolution mapping system based on semi-global optimization as claimed in claim 3, further comprising an error analysis module: and the method is used for respectively calculating the position errors between the sub-pixel positioning coordinates of the coastline and the reference coastline and between the coastline vector result and the reference coastline and analyzing the error results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810189532.2A CN108550174B (en) | 2018-03-08 | 2018-03-08 | Coastline super-resolution mapping method and coastline super-resolution mapping system based on semi-global optimization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810189532.2A CN108550174B (en) | 2018-03-08 | 2018-03-08 | Coastline super-resolution mapping method and coastline super-resolution mapping system based on semi-global optimization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108550174A CN108550174A (en) | 2018-09-18 |
| CN108550174B true CN108550174B (en) | 2020-06-26 |
Family
ID=63515926
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810189532.2A Active CN108550174B (en) | 2018-03-08 | 2018-03-08 | Coastline super-resolution mapping method and coastline super-resolution mapping system based on semi-global optimization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108550174B (en) |
Families Citing this family (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109360362A (en) * | 2018-10-25 | 2019-02-19 | 中国铁路兰州局集团有限公司 | A kind of railway video monitoring recognition methods, system and computer-readable medium |
| CN110648347A (en) * | 2019-09-24 | 2020-01-03 | 北京航天宏图信息技术股份有限公司 | Coastline extraction method and device based on remote sensing image |
| CN111027446B (en) * | 2019-12-04 | 2020-09-01 | 中国地质环境监测院 | Coastline automatic extraction method of high-resolution image |
| CN110968714B (en) * | 2019-12-23 | 2020-11-06 | 自然资源部国土卫星遥感应用中心 | Satellite remote sensing image instant service method and instant service platform |
| CN111399503A (en) * | 2020-03-09 | 2020-07-10 | 西南科技大学 | Automatic mooring method for small unmanned ship |
| CN111914909B (en) * | 2020-07-15 | 2021-01-12 | 南京审计大学 | Hyperspectral change detection method based on space-spectrum combined three-direction convolution network |
| CN112579688B (en) * | 2020-12-11 | 2022-10-18 | 山东科技大学 | Mining method, device, equipment and medium of spatial association rule |
| CN113739788B (en) * | 2021-08-03 | 2023-05-23 | 中山大学 | Geographic position correction method and device for bright temperature data |
| CN116863010B (en) * | 2023-05-18 | 2023-12-19 | 珠江水利委员会珠江流域水土保持监测中心站 | Water and soil loss map spot comprehensive drawing method based on space aggregation analysis |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8463568B1 (en) * | 2011-04-21 | 2013-06-11 | The United States Of America, As Represented By The Secretary Of The Interior | Measuring seawater capacitance in three dimensions as well as time |
| CN103870608A (en) * | 2014-04-08 | 2014-06-18 | 武汉大学 | Method for mapping knowledge construction and formalized expression based on map symbols |
| CN104732215A (en) * | 2015-03-25 | 2015-06-24 | 广西大学 | Remote-sensing image coastline extracting method based on information vector machine |
| CN105374041A (en) * | 2015-11-18 | 2016-03-02 | 国家海洋局第一海洋研究所 | Method of extracting sandy coastline by using multiple periods of remote sensing images |
| CN107301649A (en) * | 2017-06-15 | 2017-10-27 | 大连海事大学 | A kind of region merging technique SAR image coastline Detection Method algorithm based on super-pixel |
| DE102017206592A1 (en) * | 2016-04-27 | 2017-11-02 | Codar Ocean Sensors, Ltd. | Coastal HF Radar System for Tsunami Warning |
-
2018
- 2018-03-08 CN CN201810189532.2A patent/CN108550174B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8463568B1 (en) * | 2011-04-21 | 2013-06-11 | The United States Of America, As Represented By The Secretary Of The Interior | Measuring seawater capacitance in three dimensions as well as time |
| CN103870608A (en) * | 2014-04-08 | 2014-06-18 | 武汉大学 | Method for mapping knowledge construction and formalized expression based on map symbols |
| CN104732215A (en) * | 2015-03-25 | 2015-06-24 | 广西大学 | Remote-sensing image coastline extracting method based on information vector machine |
| CN105374041A (en) * | 2015-11-18 | 2016-03-02 | 国家海洋局第一海洋研究所 | Method of extracting sandy coastline by using multiple periods of remote sensing images |
| DE102017206592A1 (en) * | 2016-04-27 | 2017-11-02 | Codar Ocean Sensors, Ltd. | Coastal HF Radar System for Tsunami Warning |
| CN107301649A (en) * | 2017-06-15 | 2017-10-27 | 大连海事大学 | A kind of region merging technique SAR image coastline Detection Method algorithm based on super-pixel |
Non-Patent Citations (1)
| Title |
|---|
| 利用可控分形插值实现海岸线多分辨率可视化;江宝得;《计算机辅助设计与图形学学报》;20171130;第2015-2021页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108550174A (en) | 2018-09-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108550174B (en) | Coastline super-resolution mapping method and coastline super-resolution mapping system based on semi-global optimization | |
| Di et al. | Automatic shoreline extraction from high-resolution IKONOS satellite imagery | |
| Dai et al. | Coastline extraction from repeat high resolution satellite imagery | |
| CN107610164B (en) | High-resolution four-number image registration method based on multi-feature mixing | |
| CN108765488B (en) | Shadow-based high-resolution remote sensing image building height estimation method | |
| Jawak et al. | A rapid extraction of water body features from antarctic coastal oasis using very high-resolution satellite remote sensing data | |
| KR20180094702A (en) | Apparatus and method for extraction of tidal creeks in tidal flat zones using drone | |
| CN103984945A (en) | Optical remote sensing image ship detection method | |
| Ghorai et al. | Extracting shoreline from satellite imagery for GIS analysis | |
| Kaiser | Environmental changes, remote sensing, and infrastructure development: The case of Egypt's East Port Said harbour | |
| CN111597930A (en) | Coastline extraction method based on remote sensing cloud platform | |
| Di et al. | Coastal mapping and change detection using high-resolution IKONOS satellite imagery | |
| Zhao et al. | Stability evaluation of tidal flats based on time-series satellite images: A case study of the Jiangsu central coast, China | |
| Zhu et al. | Coastline extraction from high-resolution multispectral images by integrating prior edge information with active contour model | |
| JP2005346664A (en) | Coastline extraction method and coastline extraction system | |
| CN114519824A (en) | Rapid detection method for SAR image flood inundation area | |
| Wang et al. | Coastline extraction from SAR images using robust ridge tracing | |
| Yu et al. | Coastline detection using optical and synthetic aperture radar images | |
| Wang et al. | An automatic algorithm for estimating tropical cyclone centers in synthetic aperture radar imagery | |
| Tsai et al. | Monitoring multidecadal coastline change and reconstructing tidal flat topography | |
| CN103530628A (en) | High-resolution remote sensing image ortho-rectification method based on floating control point | |
| CN112166693B (en) | Regional surface water resource remote sensing monitoring method based on small satellite | |
| Wang et al. | Deriving natural coastlines using multiple satellite remote sensing images | |
| CN111696054B (en) | Rubber dam body detection method based on full-polarization SAR image | |
| Chen et al. | Method of Water Body Information Extraction in Complex Geographical Environment from Remote Sensing Images. |
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 |