CN110276797A - A method for extracting lake area - Google Patents

Abstract

The invention discloses a method for extracting the area of a lake. The method comprises the following steps: using a FROM-GLC data set to obtain a TIFF remote sensing image of a lake area, converting the TIFF remote sensing image of a lake area, and obtaining a lake area element file; using the lake area element The raster file of the lake area is obtained from the file; the mask file of the lake area is exported according to the raster file of the lake area; according to the mask file of the lake area, the long-sequence TIFF remote sensing images of the lake area are batch-processed, and the lake area is calculated in batches. The lake area extraction method provided by the invention has high extraction precision, can extract and calculate the lake area in batches, has high work efficiency, and has practicability and universality.

Description

A method for extracting lake area

technical field

The invention belongs to the technical field of lake research, and in particular relates to a method for extracting lake area.

Background technique

Lakes are key components of climate, water, and biogeochemical cycles, and changes in lake surface area have a major impact on environmental changes and human production activities. As the area of the lake decreases, the water storage capacity will deteriorate. Floods will often break out in the flood season, and there will be no water available in the dry season. This will have a negative impact on the surrounding ecological balance and biological diversity. Some species will disappear accordingly, and it will affect the local The climate makes the ecological environment more monotonous, which is not conducive to the survival and development of human beings. At present, exploring the interannual variation trend and driving force of lake area is a research hotspot at home and abroad. The accuracy of the analysis of lake area change depends on the accuracy of lake area extraction, and the efficiency of lake area age change trend analysis depends on the batch extraction of lake area. However, the existing lake area extraction methods have low extraction accuracy, low efficiency and poor practicability.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies in the prior art and provide a lake area extraction method with high extraction accuracy, capable of batch extraction and calculation of lake area, high work efficiency, practicability and universal applicability.

The present invention provides following technical scheme:

A method for extracting lake area, said method comprising the steps of:

Use the FROM-GLC dataset to obtain TIFF remote sensing images of the lake area, convert the TIFF remote sensing images of the lake area, and obtain lake surface element files;

Use the lake surface element file to obtain the raster file of the lake area;

Export the mask file of the lake area according to the raster file of the lake area;

According to the mask file of the lake area, the long-sequence TIFF remote sensing images of the lake area are processed in batches, and the lake area is extracted and calculated in batches.

Preferably, the acquisition method of the lake surface element file includes:

Under the ARCGIS10.2 software platform, make the lake line element shape file according to the TIFF remote sensing image of the lake area, and then use the data management tools in the ArcToolbox toolbox to select feature to polygon under features, and convert the lake line element shape file to Convert it to a lake surface feature shape file.

Preferably, the acquisition method of the lake area raster file comprises:

Use the lake surface element file to preliminarily crop the TIFF remote sensing image of the lake area mentioned above, and obtain the raster file of the lake area that is initially cropped;

According to the TIFF remote sensing image of the lake area and the characteristics of the corresponding lake area waters, the mask file is corrected by threshold setting, so that the initially cropped raster file of the lake area is gradually adjusted to the large range of the TIFF remote sensing image of the lake area, and the final cropped lake is obtained. area raster file.

Preferably, the method for obtaining the preliminary cropped lake area raster file includes:

Use the spatial analyst tools in the ArcToolbox toolbox of the ARCGIS10.2 software, select extraction, and then select extract by mask to generate a preliminary cropped lake raster file.

Preferably, the method for correcting the mask file includes:

Use the spatial analyst tools in the ArcToolbox toolbox of the ARCGIS10.2 software, select mapalgebra, and then select the raster calculator, set all areas greater than or equal to 0 to 300, and areas greater than or equal to 0 to be lake areas, and output the processed lake area raster again File, by merging the raster files several times, the raster file of the lake area within a large area affected by the TIFF remote sensing image of the lake area is obtained.

Preferably, the method for obtaining the finally cropped lake area raster file includes:

For the raster files of the lake area within a large area affected by TIFF remote sensing images of the lake area, use the data management of the ArcToolBox toolbox of the ARCGIS10.2 software to select raster, then select mosaic under rasterdataset, and assign the area within the lake range to 1 , other areas are assigned a value of -9999, and the lake area raster file is output, which is the final cropped lake raster file.

Preferably, the revisit period of the FROM-GLC data set is one day.

Preferably, the derivation method of the lake area mask file comprises:

Export the final cropped lake area raster file as a txt file;

The txt file is imported into MATLAB software, and the output lake area mask file is saved.

Preferably, the export method of the txt file comprises:

Utilize the conversion tool of the ArcToolBox tool box of the ARCGIS10.2 software, select the raster to ascii under from raster to generate the txt file.

Preferably, the calculation method of the lake area includes:

Edit the tiffread program with MATLAB software, crop the long-sequence TIFF remote sensing images of the lake area in batches according to the lake area mask file, obtain the daily number of water pixels in the lake range, and then calculate the long-sequence daily lake area by formula (1) area:

s=watercount×463×463 (1)

In formula (1), s is the daily lake area, and watercount is the number of water pixels.

Compared with prior art, the beneficial effect of the present invention is:

(1) The present invention uses the FROM-GLC data set to obtain TIFF remote sensing images of the lake area, which solves the misclassification of mountain, cloud and water areas, and the problem of spectral water and land mixed areas, and improves the extraction accuracy of the lake area.

(2) The method for extracting lake area provided by the present invention is simple to operate, can crop original TIFF remote sensing images in batches, extract and calculate lake area in batches, improves work efficiency, can be used to extract the area of lakes in various regions, and has strong practicality and universality.

Description of drawings

Fig. 1 is the regional map of Poyang Lake line elements in the embodiment;

Fig. 2 is a regional map of Poyang Lake surface elements in the embodiment;

Fig. 3 is the TIFF remote sensing image of the original lake area of Poyang Lake in the embodiment;

Fig. 4 is the preliminary cropped Poyang Lake raster figure in the embodiment;

Fig. 5 is the raster map of Poyang Lake in the secondary cropping in the embodiment;

Fig. 6 is the final cropped Poyang Lake raster diagram in the embodiment;

Fig. 7 is the area change trend diagram of Poyang Lake in 2001-2015 in the embodiment;

Fig. 8 is a comparison chart of the interannual variation of Poyang Lake area in the embodiment and Document 1;

Fig. 9 is a comparison diagram of the area change of Poyang Lake in the embodiment and Document 2, wherein (a) is a diagram of the area change of Poyang Lake in the embodiment, and (b) is a diagram of the area change of Poyang Lake in Document 2.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings. The following examples are only used to illustrate the technical solution of the present invention more clearly, but not to limit the protection scope of the present invention.

Example

A method for extracting lake area, taking Poyang Lake as an example, the method includes the following steps:

Step 1. Creation of lake surface element files

Using the FROM-GLC data set to obtain the h28v06 image of Poyang Lake on the second day of 2001 (TIFF remote sensing image of the lake area), under the ARCGIS10.2 software platform, create a shape file of Poyang Lake line elements based on the TIFF remote sensing image of the lake area, and then Use the data management tools in the ArcToolbox toolbox, select featureto polygon under features, and convert the Poyang Lake line feature shape file into the Poyang Lake surface feature shape file, as shown in Figure 1 for the Poyang Lake line feature area, and Figure 2 for the Poyang Lake surface feature area.

Step 2. Create a raster file in the lake area

Use the shape file of Poyang Lake surface elements to preliminarily crop the TIFF remote sensing image of the lake area. Specifically, use the spatial analyst tools in the ArcToolbox toolbox of the ARCGIS10.2 software, select extraction, and then select extract by mask to generate a preliminary cropped Poyang Lake raster The file, as shown in Figure 3 is the original TIFF remote sensing image of the lake area, and Figure 4 is the pre-cut raster file of Poyang Lake.

It can be seen from Figure 4 that the range of the pre-cut Poyang Lake raster file is smaller than the original TIFF remote sensing image range of the lake area, so the next threshold setting is performed, that is, the pre-cut Poyang Lake raster file is recalculated. In order to use the spatial analyst tools in the ArcToolbox toolbox of the ARCGIS10.2 software, select mapalgebra, then select the raster calculator, set all areas greater than or equal to 0 to 300, and the area greater than or equal to 0 is the range of Poyang Lake, and output the processed The raster file of Poyang Lake is shown in Figure 5.

It can be seen from Figure 5 that the range of the processed raster file of Poyang Lake is still smaller than the range of the original TIFF remote sensing image of the lake area. By merging the raster files again, the processed raster file of Poyang Lake is adjusted to the range of the original TIFF remote sensing image of the lake area. Within a large range, use the data management of the ArcToolBox toolbox of the ARCGIS10.2 software, select raster, and then select mosaic under the raster dataset to adjust the value of the raster file of Poyang Lake after the merger is completed, and the area within the lake Assign a value of 1, other areas assign a value of -9999, and output the lake area raster file, which is the final cropped Poyang Lake raster file, as shown in Figure 6.

Step 3. Make the mask file of the lake area

Export the final cropped Poyang Lake raster file as a txt file, specifically, use the conversion tool of the ArcToolBox toolbox of the ARCGIS10.2 software to select the raster to ascii under from raster to generate the txt file; then import the txt file into MATLAB The software saves and outputs the Poyang Lake mask file.

Step 4. Calculation of the long-term area of the lake

Edit the tiffread program with MATLAB software, crop the 2001-2015 Poyang Lake range TIFF remote sensing images in batches according to the Poyang Lake mask file, and obtain the daily water pixel number in the Poyang Lake range from 2001 to 2015, and then use the formula (1 ) to calculate the daily lake surface area of Poyang Lake from 2001 to 2015:

s=watercount×463×463 (1)

In formula (1), s is the daily lake surface area, and watercount is the daily number of water pixels.

Statistics on the surface area of Poyang Lake

Use Excel software to process the long-term Poyang Lake surface area obtained in the above examples, calculate the monthly average area of Poyang Lake from 2001 to 2015, and then further calculate the yearly average area of Poyang Lake from 2001 to 2015, and make Poyang Lake 2001 -The trend chart of area change in 2015, as shown in Figure 7, the area data of Poyang Lake from 2001 to 2015 are shown in Table 1:

Table 1 Lake area series of Poyang Lake from 2001 to 2015

Verification of the surface area of Poyang Lake

In this example, literature 1 (Zhang Kexiang. MODIS monitoring of typical lake area changes in the middle and lower reaches of the Yangtze River [D]. East China University of Science and Technology, 2015) and literature 2 (Xu Xiaohua, Zhang Weiqi, Hu Qiang. Remote sensing monitoring of Poyang Lake surface area based on MODIS Research [J]. Jiangxi Water Conservancy Science and Technology, 2008, 34(4): 256-258) to verify the lake area.

In Document 1, the 250m vegetation index product MOD13Q1 synthesized by Terra/MODIS data in 16 days was used as the basic data, and its overall classification accuracy was 89.59%. -GLC data are compared, and the deviation percentage is calculated according to the formula (2):

r=(Area _MOD13Q1 -Area _FROM-GLC )/Area _FROM-GLC ×100% (2)

In formula (2): r is the deviation percentage, Area _MOD13Q1 is the area of Poyang Lake obtained based on MOD13Q1, and Area _FROM-GLC is the area of Poyang Lake obtained based on FROM-GLC. The calculation results are shown in Table 3.

Table 2 Average annual lake surface area of Poyang Lake from 2001 to 2013 in Document 1

Table 3 The percentage of deviation between Document 1 and the area of Poyang Lake extracted in this example

See Figure 8 for the comparison of the interannual change law of the area of Poyang Lake in Document 1 and this example.

Document 2 uses MODIS data to extract the lake surface area of Poyang Lake in the first half of 2007 and the second half of 2008. Now select the MODIS/FROM-GLC data of the same date to compare with the area of Poyang Lake in this example, as shown in Figure 9(a) and (b).

It can be seen from Figure 8 and Figure 9 that the overall trend of the lake surface area of Poyang Lake extracted in this example is consistent with that in Document 1 and Document 2, thus verifying the time series of the lake surface area of Poyang Lake obtained in this example is correct and reasonable. It can be seen from Table 3 that the difference between the surface area of Poyang Lake obtained in this example and Document 1 is 6%-27%. The reason for these differences is the time resolution, spatial resolution and spectral resolution of MOD13Q1 data and FROM-GLC data Rates are different.

The lake area extraction method provided in this example is based on the FROM-GLC dataset to obtain TIFF remote sensing images of the lake area, the revisit period is one day, and the ground object resolution is 463m.

The FROM-GLC data set is a global multi-category land use data product based on the landsat8TM/OLI satellite. It is the first product in the world with a spectral resolution of 30 meters, a fine resolution observer and monitoring of global land cover, and can Extracts a global water film. Comparing this data product with other products, it is found that producer accuracy (PA) and user accuracy (UA) have increased by 4.62% and 0.51% respectively, and the total area of global inland water bodies has decreased by 15.83%.

The FROM-GLC data set adopts an object-based method, which is to calculate the topographic features, spectral features and geometric relations of the cloud for each water object in the water film from the FROM-GLC data product, and set specific rules to determine the water Whether the object is misclassified, use spectral information and terrain data to automatically identify mountain shadows, and use the geometric relationship between the sun-sensor-cloud to automatically identify cloud shadows, thereby eliminating misclassified water bodies. This data product solves two common problems in image water classification results, that is, shadow (including mountain and cloud shadow misclassification) and water-land spectral mixed area misclassification, so it can improve the accuracy of lake area extraction.

In addition, by editing the tiffread program with MATLAB software, the original TIFF remote sensing images can be cropped in batches, and the lake area can be extracted and calculated in batches, which improves the work efficiency and can be used to extract the area of lakes in various regions, which has strong practicability and universal applicability .

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the technical principle of the present invention, some improvements and modifications can also be made. It should also be regarded as the protection scope of the present invention.

Claims (10)

1. a lake area extraction method, is characterized in that, described method comprises the steps: Use the FROM-GLC dataset to obtain TIFF remote sensing images of the lake area, convert the TIFF remote sensing images of the lake area, and obtain lake surface element files; Use the lake surface element file to obtain the raster file of the lake area; Export the mask file of the lake area according to the raster file of the lake area; According to the mask file of the lake area, the long-sequence TIFF remote sensing images of the lake area are processed in batches, and the lake area is extracted and calculated in batches. 2. the lake area extraction method according to claim 1, is characterized in that, the acquisition method of described lake surface element file comprises: Under the ARCGIS10.2 software platform, the lake line element shape file was made according to the TIFF remote sensing image of the lake area, and then the data management tools in the ArcToolbox toolbox were used to select the feature topolygon under features to convert the lake line element shape file It is the shape file of the lake surface feature. 3. The lake area extraction method according to claim 1, wherein the method for obtaining the lake area raster file comprises: Use the lake surface element file to preliminarily crop the TIFF remote sensing image of the lake area mentioned above, and obtain the raster file of the lake area that is initially cropped; According to the TIFF remote sensing image of the lake area and the characteristics of the corresponding lake area waters, the mask file is corrected by threshold setting, so that the initially cropped raster file of the lake area is gradually adjusted to the large range of the TIFF remote sensing image of the lake area, and the final cropped lake is obtained. area raster file. 4. the lake area extraction method according to claim 3, is characterized in that, the acquisition method of the lake area raster file of preliminary cutting comprises: Use the spatial analyst tools in the ArcToolbox toolbox of the ARCGIS10.2 software, select extraction, and then select extract by mask to generate a preliminary cropped lake raster file. 5. The lake area extraction method according to claim 3, wherein the method for correcting the mask file comprises: Use the spatial analyst tools in the ArcToolbox toolbox of the ARCGIS10.2 software, select mapalgebra, and then select the raster calculator, set all areas greater than or equal to 0 to 300, and areas greater than or equal to 0 to be lake areas, and output the processed lake area raster again File, by merging the raster files several times, the raster file of the lake area within a large range of the TIFF remote sensing image of the lake area is obtained. 6. the lake area extraction method according to claim 5, is characterized in that, the acquisition method of the lake area raster file that finally cuts out comprises: For the raster files of the lake area within a large area affected by TIFF remote sensing images of the lake area, use the data management of the ArcToolBox toolbox of the ARCGIS10.2 software to select raster, then select mosaic under rasterdataset, and assign the area within the lake range to 1 , other areas are assigned a value of -9999, and the lake area raster file is output, which is the final cropped lake raster file. 7. The lake area extraction method according to claim 1, wherein the revisit period of the FROM-GLC data set is one day. 8. the lake area extraction method according to claim 1, is characterized in that, the derivation method of described lake area mask file comprises: Export the final cropped lake area raster file as a txt file; The txt file is imported into MATLAB software, and the output lake area mask file is saved. 9. the lake area extraction method according to claim 8, is characterized in that, the derivation method of described txt file comprises: Utilize the conversion tool of the ArcToolBox tool box of the ARCGIS10.2 software, select the raster to ascii under from raster to generate the txt file. 10. The lake area extraction method according to claim 1, wherein the calculation method of the lake area comprises: Edit the tiffread program with MATLAB software, crop the long-sequence TIFF remote sensing images of the lake area in batches according to the lake area mask file, obtain the daily number of water pixels in the lake range, and then calculate the long-sequence daily lake area by formula (1) area: s=watercount×463×463 (1) In formula (1), s is the daily lake area, and watercount is the number of water pixels.