A method for monitoring and analyzing muddy coastline changes based on remote sensing technology

Abstract

The present invention discloses a muddy coastline change monitoring and analysis method based on remote sensing technology. The method comprises the following steps: firstly, pre-processing a remote sensing image to obtain an initial remote sensing image of a target coastal area; constructing a water body index based on the remote sensing image to obtain a water body index image; then determining a target segmentation threshold to divide the water body index image to obtain a binary image of the water body index image; then obtaining the instantaneous water edge line and target coastline category of the target coastal area; adopting tidal gully-based tide level correction to correct the instantaneous water edge line in the binary image of the target coastal area to obtain the target coastline; finally, inputting the target coastline collected at different periods into a digital coastline analysis system, and performing end point change rate and linear regression change rate analysis on the multiple target coastlines collected by the digital coastline analysis system, thereby dynamically monitoring the target coastline and obtaining the change of the target coastline.

Description

Technical Field

[0001] The invention relates to a muddy coastline change monitoring and analysis method based on remote sensing technology, belonging to the technical field of coastline monitoring. Background Art

[0002] Coastlines are undergoing dramatic changes due to the combined influence of global and regional environmental processes and human activities. These changes have significant implications for ecology, the environment, and the economy and society. Consequently, research on coastline change has garnered widespread attention. Coastline changes not only reflect the characteristics and evolution of coastal environments but also reveal the interplay between socioeconomic development, ecological and environmental changes, and policy orientations. Furthermore, coastline changes can significantly impact marine engineering. Therefore, research on the dynamics of coastline change is fundamental to coastal environmental monitoring, resource development, and management, and is of great significance to coastal development.

[0003] Manual field collection of shoreline data presents drawbacks such as difficulty, high cost, and long cycles. It also fails to meet the coastline update rate required for marine management. Remote sensing technology, however, offers large-scale, simultaneous monitoring and a shorter update cycle, making it more conducive to dynamic coastline monitoring. Currently, there are two methods for extracting coastlines using remote sensing interpretation: manual visual interpretation and computer-generated classification and interpretation. Manual visual interpretation is simple in principle but places high demands on the personnel involved. The extraction results from different operators often exhibit significant errors, resulting in uncertain accuracy. With the development of computer technology, automated coastline interpretation methods have become mainstream research, effectively avoiding errors caused by different operators and enabling automated coastline data acquisition. However, due to the constant evolution of coastlines, existing automated coastline acquisition methods are unable to perform corresponding corrections based on these changes. This ultimately results in large errors in the coastline data collected after the changes, making it impossible to accurately and dynamically monitor the coastline as it continues to evolve. Summary of the Invention

[0004] In response to the problems existing in the above-mentioned prior art, the present invention provides a method for monitoring and analyzing changes in muddy coastlines based on remote sensing technology, which uses tidal creek tide level correction to make corresponding corrections according to coastline changes, thereby realizing dynamic and accurate monitoring of the coastline as it continues to change.

[0005] In order to achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a method for monitoring and analyzing muddy coastline changes based on remote sensing technology, characterized by the following specific steps:

[0006] Step 1: Acquire remote sensing images of the target coastal area;

[0007] Step 2: Preprocessing the remote sensing image obtained in step 1 to obtain an initial remote sensing image of the target coastal area;

[0008] Step 3: constructing a water body index based on the remote sensing image obtained in step 2, thereby obtaining a water body index image;

[0009] Step 4: First, select different segmentation thresholds, and use each segmentation threshold to divide the target coastal area and background area of the water index image obtained in step 3. Calculate the inter-class variance between the target coastal area and the background area corresponding to each segmentation threshold, and select the threshold corresponding to the maximum inter-class variance from multiple inter-class variances, which is the target segmentation threshold. Finally, divide the water index image according to the target segmentation threshold to obtain a binary image of the water index image;

[0010] Step 5: Convert the binary image obtained in step 4 into a vector image, extract the boundary of the surface vector object in the vector image to obtain the instantaneous water edge of the target coastal area, and use the visual interpretation method to determine the target coastline category corresponding to the water edge;

[0011] Step 6: Using tidal gully-based tide level correction, the instantaneous water edge in the binary image of the target coastal area is corrected to obtain the target coastline;

[0012] Step 7: Input the target coastline obtained in step 6 into the digital coastline analysis system, and then repeat steps 1 to 6 after a period of time to obtain the current target coastline again and input it into the digital coastline analysis system. Repeat this cycle. The digital coastline analysis system will analyze the end point change rate and linear regression change rate of the multiple target coastlines collected, so as to dynamically monitor the target coastline and obtain the changes in the target coastline.

[0013] Furthermore, the step 2 is specifically as follows: performing orthorectification and registration, radiometric calibration, atmospheric correction, and image cropping and splicing on the remote sensing image in sequence to obtain an initial remote sensing image of the target coastal area.

[0014] Furthermore, the water index in step 3 is the modified normalized water index (MNDWI). This index is more effective than the normalized water index (NDWI) in extracting water bodies within urban areas. The NDWI index image is often mixed with urban construction land information, which expands the range and area of the extracted water bodies, resulting in reduced accuracy. Therefore, the modified normalized water index (MNDWI) is selected. The specific expression is:

[0015] MNDWI=(p(Green)-p(MIR)) / (p(Green)+p(MIR))

[0016] MIR stands for mid-infrared band and Green stands for green light band.

[0017] Furthermore, the target coastline categories in step five include natural coastlines and artificial coastlines.

[0018] Furthermore, the step six is specifically as follows: adopting tide level correction based on tidal creeks, first determining the tidal creeks corresponding to the target coastline, the tidal creeks are formed by the local strong scouring caused by the late ebb tide flow or the surface escape flow after the ebb tide on the tidal flat surface, and gradually developing and evolving; the inventors of the present application have found that: this type of tidal creek usually originates near the mean high tide line, and its overall flow direction is roughly perpendicular to the coastline, forming a dendritic water system on the plane, wherein the main trunk of the tidal creek is in the middle and lower part of the intertidal area, and as the tidal creek moves upstream, it branches out in the middle and upper part of the intertidal area in a dendritic manner; based on this, let the end position of the tidal creek be a, and the boundary position between the artificial building and the tidal flat be b, and calculate the coastline distance L to be corrected in the target coastal area according to the following formula:

[0019] L=(ba) / 2

[0020] The instantaneous water edge in the binary image of the target coastal area is moved vertically by a distance L away from the water area to obtain the tide-corrected coastline, namely the target coastline.

[0021] Furthermore, in step seven, the end-point change rate and linear regression change rate analysis are performed on the target coastline for multiple times. Specifically, the coastline change end-point rate (EPR) is the ratio of the coastline distance between two time phases to their time interval. The linear regression change rate (LRR) is determined by fitting the least squares regression line to all coastline points in the transect to minimize the sum of squared residuals and fit the coastline changes over many years.

[0022] The calculation formulas for EPR and LRR are as follows:

[0023]

[0024] LRR=ax+b

[0025]

[0026]

[0027] Where D is the distance between the coastlines of the two time phases, T is the time interval between the two time phases; a and b are the slope and intercept of the coastline position series fitting line, respectively, and x i is the X-axis coordinate position of the coastline at period i, y i is the distance interval between the shoreline point and the baseline point at period i on a certain section perpendicular to the true coastline, and n is the number of coastline phases;

[0028] In addition, the distance between the most recently collected coastline and the earliest collected coastline can be calculated using the net coastline movement calculation formula. The specific formula is:

[0029] NSM=D new -D old

[0030] Where NSM represents the distance between the latest coastline and the earliest coastline during the study period. new Indicates the most recent coastline, D old Indicates the earliest coastline.

[0031] Compared with the existing technology, the present invention first pre-processes the remote sensing image to obtain the initial remote sensing image of the target coastal area; constructs the water body index based on the remote sensing image to obtain the water body index image; then determines the target segmentation threshold to divide the water body index image to obtain a binary image of the water body index image; then obtains the instantaneous water edge of the target coastal area and the target coastline category; adopts the tide level correction based on the tidal ditch to correct the instantaneous water edge in the binary image of the target coastal area to obtain the target coastline; the tidal ditch usually originates near the average high tide line, and its overall flow direction is roughly perpendicular to the coastline, forming a dendritic water system on the plane, in which the tidal ditch The main trunk is in the middle and lower part of the intertidal zone, and branches out in a tree-like manner in the middle and upper part of the intertidal zone as the tidal creek moves upstream. As the coastline changes, the tidal creek will also change. Therefore, the tidal creek is used as a reference for tide level correction, so that the target coastline in each remote sensing image can be obtained. After the coastline changes, the correction method can still accurately determine the target coastline after the change. Finally, the target coastline collected at different times is input into the digital coastline analysis system, and the digital coastline analysis system performs end point change rate and linear regression change rate analysis on the multiple target coastlines collected, so as to dynamically monitor the target coastline and obtain the change of the target coastline. The present invention first uses a specific target segmentation threshold to obtain a binary image, and then, based on the corresponding change relationship between the tidal creek and the coastline discovered by the inventor's research, uses the tidal level correction based on the tidal creek to correct the target coastline. It can still have an accurate correction effect after the coastline changes, thereby realizing dynamic and accurate monitoring of the coastline as the coastline continues to change. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 This is a schematic diagram of the overall process of the present invention

[0033] Figure 2 This is a remote sensing image after radiation calibration according to an embodiment of the present invention.

[0034] Figure 3 This is a remote sensing image after atmospheric correction according to an embodiment of the present invention.

[0035] Figure 4 It is a remote sensing image processed by MNDWI in an embodiment of the present invention.

[0036] Figure 5 This is a binary image segmented according to the target segmentation threshold in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described below.

[0038] Taking the coastline monitoring in Yancheng as an example, Figure 1 As shown, the specific steps of this embodiment are:

[0039] Step 1: Obtain Landsat series remote sensing images covering the entire coastline of Yancheng on the geospatial data cloud as remote sensing images of the target coastal area. To improve accuracy, the requirement for remote sensing images is: cloud cover is less than or equal to 5%.

[0040] Step 2: The remote sensing images obtained in step 1 are processed using ENVI5.3 software, which performs orthorectification and registration, radiometric calibration, atmospheric correction, and image cropping and splicing in sequence to obtain the initial remote sensing image of the target coastal area. The specific process is as follows:

[0041] like Figure 2 As shown, the remote sensing image is first subjected to radiometric calibration. Radiometric calibration is the process of converting the digital quantization (DN) of an image into physical quantities such as radiance, reflectivity, or surface temperature. Radiometric calibration parameters are typically stored in metadata files. The General Radiometric Calibration tool in ENVI can automatically read these parameters from metadata files to complete radiometric calibration.

[0042] like Figure 3 As shown in the figure, the radiometrically calibrated remote sensing image is subjected to Flaash atmospheric correction. The Flaash atmospheric correction uses the code of the MODTRAN 4+ radiation transfer model. It is based on pixel-level correction to correct the cascade effects caused by diffuse reflection, including classification maps of cirrus and opaque clouds, and can adjust the spectral smoothing caused by artificial suppression.

[0043] The resolution of the panchromatic band is 15m, and the resolution of the multispectral band is 30m. In order to make the remote sensing image more prominent in the features of the ground objects and facilitate the subsequent interpretation work, this embodiment adopts the band fusion method of Brovey (color normalization) transformation to fuse the panchromatic and multispectral bands of the atmospherically corrected panchromatic image. This makes the image have both high spatial resolution and texture characteristics and rich spectral information, thereby achieving the purpose of rich image map information, good visual effects, and high quality, and ultimately obtaining the initial remote sensing image of the target coastal area.

[0044] Step 3: Construct a water index based on the remote sensing image obtained in step 2. The water index is the modified normalized water index (MNDWI). This index is more effective than the normalized water index (NDWI) in extracting water bodies within urban areas. NDWI index images are often mixed with urban construction land information, which expands the range and area of extracted water bodies, resulting in reduced accuracy. Therefore, the modified normalized water index (MNDWI) is selected. The specific expression is:

[0045] MNDWI=(p(Green)-p(MIR)) / (p(Green)+p(MIR))

[0046] Among them, MIR stands for mid-infrared band, and Green stands for green light band.

[0047] like Figure 4 As shown, the water body index image is obtained through the above water body index.

[0048] Step 4: First, select different segmentation thresholds. Use each segmentation threshold to divide the target coastal area and background area from the water index image obtained in step 3. Calculate the inter-class variance between the target coastal area and the background area corresponding to each segmentation threshold, and select the threshold corresponding to the maximum inter-class variance from multiple inter-class variances, which is the target segmentation threshold, as shown in the following example: Figure 5 As shown, finally, the water index image is divided according to the target segmentation threshold to obtain a binary image of the water index image;

[0049] Step 5: Convert the binary image obtained in step 4 into a vector image, extract the boundaries of the surface vector objects in the vector image to obtain the instantaneous water edge of the target coastal area, and use a visual interpretation method to determine the target coastline category corresponding to the water edge; the target coastline category includes natural coastlines and artificial coastlines; the tide level correction in this embodiment is targeted at muddy coastlines within natural coastlines;

[0050] Step 6: Use tidal gully-based tide level correction to correct the instantaneous water edge in the binary image of the target coastal area to obtain the target coastline. Specifically:

[0051] Tidal level correction based on tidal creeks is first determined by determining the tidal creek corresponding to the target coastline. Tidal creeks are formed by localized strong scouring caused by late ebb tide flow or surface escape flow after ebb tide on the tidal flat surface. The inventors of this application have found that such tidal creeks usually originate near the mean high tide line, and their overall flow direction is roughly perpendicular to the coastline, forming a dendritic water system on the plane. The main trunk of the tidal creek is in the middle and lower part of the intertidal zone, and as the tidal creek moves upstream, it branches out like a dendrite in the middle and upper part of the intertidal zone. Based on this, the position of the end of the tidal creek is set as a, and the position of the boundary between the artificial structure and the tidal flat is set as b. The coastline distance L to be corrected in the target coastal area is calculated according to the following formula:

[0052] L=(ba) / 2

[0053] The instantaneous water edge in the binary image of the target coastal area is moved vertically by a distance L away from the water area to obtain the tide-corrected coastline, namely the target coastline.

[0054] Step 7: Input the target coastline obtained in step 6 into the digital coastline analysis system, and then repeat steps 1 to 6 after a period of time to obtain the current target coastline again and input it into the digital coastline analysis system. Repeat this cycle, and use the digital coastline analysis system to analyze the end-point change rate and linear regression change rate of the collected multiple target coastlines. Specifically, the coastline change end-point rate (EPR) is the ratio of the coastline distance between two time phases to their time interval, and the linear regression change rate (LRR) is determined by fitting the least squares regression line to all coastline points in the transect to minimize the sum of squared residuals, thus fitting the changes in the coastline over many years.

[0055] The calculation formulas for EPR and LRR are as follows:

[0056]

[0057] LRR=ax+b

[0058]

[0059]

[0060] Where D is the distance between the coastlines of the two time phases, T is the time interval between the two time phases; a and b are the slope and intercept of the coastline position series fitting line, respectively, and x i is the X-axis coordinate position of the coastline at period i, y i is the distance interval between the shoreline point and the baseline point at period i on a certain section perpendicular to the true coastline, and n is the number of coastline phases;

[0061] In addition, the distance between the most recently collected coastline and the earliest collected coastline can be calculated using the net coastline movement calculation formula. The specific formula is:

[0062] NSM=D new -D old

[0063] Where NSM represents the distance between the latest coastline and the earliest coastline during the study period. new Indicates the most recent coastline, D old Indicates the earliest phase of the coastline;

[0064] This enables dynamic monitoring of the target coastline to reveal changes in the target coastline.

[0065] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for monitoring and analyzing muddy coastline changes based on remote sensing technology, characterized in that: The specific steps are: Step 1: Acquire remote sensing images of the target coastal area; Step 2: Preprocessing the remote sensing image obtained in step 1 to obtain an initial remote sensing image of the target coastal area; Step 3: constructing a water body index based on the remote sensing image obtained in step 2, thereby obtaining a water body index image; Step 4: First, select different segmentation thresholds, and use each segmentation threshold to divide the target coastal area and background area of the water index image obtained in step 3. Calculate the inter-class variance between the target coastal area and the background area corresponding to each segmentation threshold, and select the threshold corresponding to the maximum inter-class variance from multiple inter-class variances, which is the target segmentation threshold. Finally, divide the water index image according to the target segmentation threshold to obtain a binary image of the water index image; Step 5: Convert the binary image obtained in step 4 into a vector image, extract the boundary of the surface vector object in the vector image to obtain the instantaneous water edge of the target coastal area, and use the visual interpretation method to determine the target coastline category corresponding to the water edge; Step 6: Using tidal gully-based tide level correction, the instantaneous water edge in the binary image of the target coastal area is corrected to obtain the target coastline; Step 7: Input the target coastline obtained in step 6 into the digital coastline analysis system, and then repeat steps 1 to 6 after a period of time to obtain the current target coastline again and input it into the digital coastline analysis system. Repeat this cycle, and use the digital coastline analysis system to analyze the end-point change rate and linear regression change rate of the collected multiple target coastlines. Specifically, the coastline change end-point rate (EPR) is the ratio of the coastline distance between two time phases to their time interval, and the linear regression change rate (LRR) is determined by fitting the least squares regression line to all coastline points in the transect to minimize the sum of squared residuals, thus fitting the changes in the coastline over many years. The calculation formulas for EPR and LRR are as follows: ； Where D is the distance between the coastlines of the two time phases, T is the time interval between the two time phases; a and b are the slope and intercept of the coastline position series fitting line, respectively, and x i is the X-axis coordinate position of the coastline at period i, y i It is the distance between the coastline point and the baseline point in period i on a certain section perpendicular to the real coastline, and n is the number of coastline phases; thus, the target coastline can be dynamically monitored to obtain the changes in the target coastline.

2. The method for monitoring and analyzing muddy coastline changes based on remote sensing technology according to claim 1 is characterized in that: The step 2 specifically includes: performing orthorectification and registration, radiometric calibration, atmospheric correction, and image cropping and splicing on the remote sensing image in sequence to obtain an initial remote sensing image of the target coastal area.

3. The method for monitoring and analyzing muddy coastline changes based on remote sensing technology according to claim 1, characterized in that: The water index in step 3 is the improved normalized water index MNDWI, and the specific expression is: MNDWI =(p(Green)-p(MIR)) / (p(Green)+p(MIR)) MIR stands for mid-infrared band and Green stands for green light band.

4. The method for monitoring and analyzing muddy coastline changes based on remote sensing technology according to claim 1 is characterized in that: The target coastline categories in step 5 include natural coastlines and artificial coastlines.

5. The method for monitoring and analyzing muddy coastline changes based on remote sensing technology according to claim 1, characterized in that: The sixth step is specifically as follows: adopting tide correction based on tidal creeks, first determine the tidal creeks corresponding to the target coastline. Tidal creeks are formed by the late ebb current or post-ebb surface flow on the tidal flat surface, which gradually develops and evolves due to the strong local scouring. Assuming the end position of the tidal creek is a, and the boundary position between the artificial building and the tidal flat is b, calculate the coastline distance to be corrected L of the target coastal area according to the following formula: L=(ba) / 2 The instantaneous water edge in the binary image of the target coastal area is moved vertically by a distance L away from the water area to obtain the tide-corrected coastline, namely the target coastline.

Images