CN114049571A - Method and device for extracting water body area of hyperspectral image and electronic equipment - Google Patents
Method and device for extracting water body area of hyperspectral image and electronic equipment Download PDFInfo
- Publication number
- CN114049571A CN114049571A CN202210034953.4A CN202210034953A CN114049571A CN 114049571 A CN114049571 A CN 114049571A CN 202210034953 A CN202210034953 A CN 202210034953A CN 114049571 A CN114049571 A CN 114049571A
- Authority
- CN
- China
- Prior art keywords
- water body
- hyperspectral image
- body region
- reflectivity
- band
- 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.)
- Pending
Links
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
- Image Processing (AREA)
Abstract
The embodiment of the application provides a method and a device for extracting a water body region of a hyperspectral image and electronic equipment, and relates to the technical field of image processing. Firstly, extracting a water body region from a target hyperspectral image through a reflectivity threshold value of a near-infrared band to obtain a first water body region; then, performing water body region extraction on the first water body region through a water body index model established based on spectral characteristics of visible light and near-infrared bands to obtain a second water body region; and finally, denoising the second water body region to obtain a water body region in the target hyperspectral image. According to the method, the spectral information in the hyperspectral image is fully utilized to extract the water body region twice, so that the extraction precision of the water body region in the hyperspectral image can be improved.
Description
Technical Field
The application relates to the technical field of image processing, in particular to a method and a device for extracting a water body region of a hyperspectral image and electronic equipment.
Background
Surface water can include various solid and liquid bodies of water on the surface of land that are closely associated with climate change, human welfare, and biodiversity protection. In recent years, nationwide water resource investigation, evaluation and monitoring work is carried out in China, the technical advantages of hydrogeological investigation, satellite remote sensing and the like are required to be fully exerted, and the quantity, quality, spatial distribution, development and utilization, ecological conditions and dynamic changes of nationwide water resources are comprehensively mastered. The remote sensing technology is widely applied to the fields of water body extraction and water quality monitoring by the characteristics of large scale, high precision and near real-time imaging. With a series of technological innovations, remote sensing can be further divided into multispectral remote sensing and hyperspectral remote sensing. Compared with multispectral images, hyperspectral images have richer spectral information. Therefore, how to accurately extract the region of the water body from the spectral information of the hyperspectral image is a technical problem that needs to be solved by those skilled in the art.
Disclosure of Invention
In order to overcome at least the defects in the prior art, the application provides a method and a device for extracting a hyperspectral image water body area and electronic equipment.
In a first aspect, an embodiment of the present application provides a method for extracting a water body region of a hyperspectral image, where the method includes:
acquiring a target hyperspectral image containing a water body area to be extracted;
performing water body area extraction on the target hyperspectral image based on a reflectivity threshold of a near-infrared band to obtain a first water body area, wherein the first water body area consists of a water body and a suspected water body, and the reflectivity of the suspected water body in the near-infrared band is similar to the reflectivity of the water body in the near-infrared band;
performing water body area extraction on the first water body area based on a water body index model to obtain a second water body area consisting of the water body, wherein the water body index model is determined based on the spectral characteristics of the water body and the suspected water body in visible light and near infrared bands, and the water body index model is used for distinguishing the water body and the suspected water body in the first water body area;
and denoising the second water body region, and taking the denoised second water body region as a water body region in the target hyperspectral image.
In the scheme, firstly, a water body area is extracted from a target hyperspectral image through a reflectivity threshold value of a near-infrared band, so that a first water body area is obtained; then, performing water body region extraction on the first water body region through a water body index model established based on spectral characteristics of visible light and near-infrared bands to obtain a second water body region; and finally, denoising the second water body region to obtain a water body region in the target hyperspectral image. According to the method, the spectral information in the hyperspectral image is fully utilized to extract the water body area twice, so that the extraction precision of the water body area in the hyperspectral image can be improved, and the non-water body area and the water body area in the hyperspectral image can be distinguished.
In a possible implementation manner, the step of obtaining a target hyperspectral image containing a water body area to be extracted includes:
taking a satellite hyperspectral image or an aviation hyperspectral image containing a water body area as the target hyperspectral image, wherein the resolution of the satellite hyperspectral image and the aviation hyperspectral image is meter-level or sub-meter-level;
before the step of performing water body region extraction on the target hyperspectral image based on the reflectivity threshold value of the near-infrared band to obtain a first water body region, the method further comprises the following steps of:
and preprocessing the target hyperspectral image, wherein the preprocessing comprises at least one of geometric correction processing, atmospheric correction processing, radiation correction processing and normalization processing.
In a possible implementation manner, the step of performing water body region extraction on the target hyperspectral image based on the reflectivity threshold of the near-infrared band to obtain a first water body region includes:
acquiring spectral component images of the target hyperspectral image on a plurality of sub near-infrared wave bands of the near-infrared wave band;
calculating to obtain a spectrum mean image of the target hyperspectral image on the near-infrared wave band based on the spectrum component images on the plurality of sub-near-infrared spectrum sub-wave bands;
and extracting the water body area from the spectrum mean value image by adopting the reflectivity threshold value of the near-infrared band to obtain the first water body area.
In a possible implementation manner, before the step of performing water body region extraction on the spectrum mean image by using the reflectance threshold of the near-infrared band to obtain the first water body region, the method further includes a step of determining the reflectance threshold of the near-infrared band, where the step includes:
acquiring a hyperspectral image of a sample, wherein the hyperspectral image of the sample comprises land surface objects of different types, and the land surface objects comprise asphalt, bright buildings, dark buildings, urban bare land, non-urban bare land, vegetation, shadows, shadow water bodies, urban water areas and non-urban water areas;
determining target reflectivity capable of distinguishing a water body and a suspected water body from other types of land surface objects according to the reflectivity of the land surface objects in the hyperspectral image of the sample in the near infrared band, and determining the target reflectivity as a reflectivity threshold of the near infrared band, wherein the water body comprises a shadow water body, an urban water body and a non-urban water body, and the suspected water body comprises a shadow.
In a possible implementation manner, before the step of performing water body region extraction on the first water body region based on the water body index model to obtain a second water body region composed of the water body, the method further includes a step of creating the water body index model, where the step includes:
determining a target position where a reflectivity curve corresponding to the water body and a reflectivity curve corresponding to the suspected water body have curve trend difference through analyzing reflectivity curves of the water body and the suspected water body in visible light and near infrared bands, and acquiring a first sub-band where the target position is located and a second sub-band located behind the target position;
and establishing a water body index model based on the wave band reflectivity corresponding to the first sub-wave band and the wave band reflectivity corresponding to the second sub-wave band.
In a possible implementation manner, the step of performing water body region extraction on the first water body region based on the water body index model to obtain a second water body region composed of the water body includes:
calculating a wave band reflectivity difference value between the wave band reflectivity of any pixel point in the first sub-wave band and the wave band reflectivity of the pixel point in the second sub-wave band in the first water body area;
taking the pixel points in the first water body area, of which the wave band reflectivity difference is smaller than or equal to a preset threshold value, as first pixel points, taking the pixel points in the first water body area, of which the wave band reflectivity difference is larger than the preset threshold value, as second pixel points, and taking the area formed by the first pixel points as the second water body area.
In a possible implementation manner, the step of performing denoising processing on the second water body region, and taking the denoised second water body region as the water body region in the target hyperspectral image includes:
and denoising the second water body region by adopting a median filtering mode, and taking the denoised second water body region as a water body region in the target hyperspectral image, wherein the median filtering adopts a filtering mask with the size of 7 x 7.
In one possible implementation manner, the range of the reflectivity threshold of the near-infrared band is 7% -9%;
the wavelength range of the first sub-band is 695 nm-705 nm, and the wavelength range of the second sub-band is 725 nm-735 nm.
In a second aspect, an embodiment of the present application further provides a water body region extraction device for hyperspectral image, the device includes:
the acquisition module is used for acquiring a target hyperspectral image containing a water body area to be extracted;
the first extraction module is used for extracting a water body area of the target hyperspectral image based on a reflectivity threshold of a near-infrared band to obtain a first water body area;
the second extraction module is used for establishing a water body index model according to the spectral characteristics of the water body in visible light and near infrared bands, and extracting the water body region from the first water body region based on the water body index model to obtain a second water body region;
and the processing module is used for denoising the second water body region, and taking the denoised second water body region as the water body region in the target hyperspectral image.
In a third aspect, an embodiment of the present application further provides an electronic device, where the electronic device includes a processor and a computer-readable storage medium, where the processor and the computer-readable storage medium are connected by a bus system, the computer-readable storage medium is used to store a program, an instruction, or a code, and the processor is used to execute the program, the instruction, or the code in the computer-readable storage medium, so as to implement the method for extracting a water body region of a hyperspectral image in any one possible implementation manner in the first aspect.
In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, where instructions are stored in the computer-readable storage medium, and when the instructions are executed, the computer is caused to execute the method for extracting a water body area of a hyperspectral image in the first aspect or any one of the possible implementation manners of the first aspect.
Based on any one of the above aspects, in the method, the device and the electronic device for extracting the water body region of the hyperspectral image, provided by the embodiment of the application, firstly, the water body region is extracted from the target hyperspectral image through a reflectivity threshold value of a near-infrared band, so that a first water body region is obtained; then, performing water body region extraction on the first water body region through a water body index model established based on spectral characteristics of visible light and near-infrared bands to obtain a second water body region; and finally, denoising the second water body region to obtain a water body region in the target hyperspectral image. According to the method, the spectral information in the hyperspectral image is fully utilized to extract the water body area twice, so that the extraction precision of the water body area can be improved, and the non-water body area and the water body area in the hyperspectral image can be distinguished.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that need to be called in the embodiments are briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and for those skilled in the art, other related drawings can be obtained according to the drawings without inventive efforts.
Fig. 1 is a schematic flow chart of a method for extracting a water body region of a hyperspectral image according to an embodiment of the present application;
fig. 2 is a comparison graph of an extraction result of a water body region and an extraction reference result provided in the embodiment of the present application;
FIG. 3 is a flowchart illustrating a sub-step of step S102 in FIG. 1;
FIG. 4 is a graph of reflectivity profiles in the blue band for different types of land surface objects provided by embodiments of the present application;
FIG. 5 is a graph of reflectivity profiles in the green band for different types of land surface objects provided by embodiments of the present application;
FIG. 6 is a graph of reflectivity profiles in the red band for different types of land surface objects provided by embodiments of the present application;
FIG. 7 is a graph of reflectance profiles in the near infrared band for different types of land surface objects provided by embodiments of the present application;
FIG. 8 is a graph of the reflectance of a body of water and shadows in the visible and near infrared bands provided in an embodiment of the present application;
fig. 9 is an AISA hyperspectral image of a city area in jiaxing city provided in an embodiment of the present application;
FIG. 10 is an OHS-2D hyperspectral image of urban and suburban areas of Changshan according to an embodiment of the present application;
FIG. 11 is a hyperspectral image of a visible and NIR imaging spectrometer of Mitsuba Bauhura, provided by an embodiment of the present application;
fig. 12 is a diagram showing evaluation indexes of the three high-spectrum images in fig. 9 to 11 after water extraction is performed by using different schemes according to the embodiment of the present application;
fig. 13 is a schematic view of functional modules of a water body region extraction device of a hyperspectral image according to an embodiment of the application;
fig. 14 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it should be understood that the drawings in the present application are for illustrative and descriptive purposes only and are not used to limit the scope of protection of the present application. Additionally, it should be understood that the schematic drawings are not necessarily drawn to scale. The flowcharts used in this application illustrate operations implemented according to some of the embodiments of the present application. It should be understood that the operations of the flow diagrams may be performed out of order, and steps without logical context may be performed in reverse order or simultaneously. One skilled in the art, under the guidance of this application, may add one or more other operations to, or remove one or more operations from, the flowchart.
In addition, the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, generally described and illustrated in the figures herein, can be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the present application, presented in the accompanying drawings, is not intended to limit the scope of the claimed application, but is merely representative of selected embodiments of the application. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present application without making any creative effort, shall fall within the protection scope of the present application.
Referring to fig. 1, fig. 1 illustrates a schematic flow chart of a method for extracting a water body region of a hyperspectral image according to an embodiment of the present application, where the method for extracting a water body region of a hyperspectral image may be executed by an electronic device (e.g., a computer device), and orders of partial steps in the method for extracting a water body region of a hyperspectral image according to the embodiment of the present application may be interchanged according to actual needs, or partial steps may be omitted or deleted. The detailed steps of the method for extracting the hyperspectral image from the water body area are described below with reference to fig. 1.
And S101, acquiring a target hyperspectral image containing a water body area to be extracted.
In this step, a hyperspectral image shot by a satellite or an aerial (such as shot by an unmanned aerial vehicle) can be used as a candidate hyperspectral image, and a hyperspectral image including a water body area is selected from the candidate hyperspectral images to be used as a target hyperspectral image. The target hyperspectral image may include hyperspectral images of different areas, such as a hyperspectral image of an urban area, a hyperspectral image of a suburban area, a hyperspectral image of a rural area, and the like; the water body area can correspond to rivers, lakes and other water areas.
After the target hyperspectral image is acquired, preprocessing can be performed on the target hyperspectral image, wherein the preprocessing can include at least one of geometric correction processing, atmospheric correction processing, radiation correction processing and normalization processing.
For example, the preprocessing of the AISA hyperspectral image includes geometric correction, radiation correction, atmospheric correction and normalization, the preprocessing of the OHS-2D hyperspectral image includes radiation correction, atmospheric correction and normalization, and the preprocessing of the hyperspectral image of the visible and NIR imaging spectrometer includes normalization, wherein the normalization normalizes the reflectivity of the hyperspectral image to a uniform range (e.g., 0-1). In addition, the target hyperspectral images from different image sources may also be subjected to different software when the same preprocessing is performed, for example, the AISA hyperspectral images can be subjected to radiation correction and atmospheric correction by adopting MODTRAN 5.3.2 software, and the OHS-2D hyperspectral images can be subjected to radiation correction and atmospheric correction by adopting ENVI software.
Step S102, performing water body area extraction on the target hyperspectral image based on a reflectivity threshold value of a near-infrared band to obtain a first water body area.
In this step, because the reflectivity of the near-infrared band has better distinguishing performance for the non-aqueous objects on the water body and most of the land surface than the reflectivity of other bands (for example, blue light band, green light band and red light band), in this application embodiment, most of the non-aqueous objects in the target hyperspectral image can be distinguished from the water body by the reflectivity threshold of the near-infrared band, so as to extract the first water body region from the target hyperspectral image.
After the processing of step S102, the obtained first water body area is composed of a water body and a suspected water body, where the suspected water body may include a shadow in the target hyperspectral image. The reflectivity of the suspected water body in the near infrared band is similar to the reflectivity of the water body in the near infrared band, so that the suspected water body and the water body cannot be distinguished by the reflectivity threshold value of the near infrared band.
Step S103, performing water body region extraction on the first water body region based on the water body index model to obtain a second water body region composed of water bodies.
In this step, the water index model may be determined according to spectral characteristics of the water body and the suspected water body in visible light and near infrared bands, and the water index model may be used to distinguish the water body and the suspected water body in the first water body region.
And S104, denoising the second water body region, and taking the denoised second water body region as a water body region in the target hyperspectral image.
After the processing of step S103, since the noise interference may include some single water body pixels in the extraction result, and these single water body pixels may affect the extraction accuracy, in this step, the single water body pixels existing in the extracted second water body region are removed.
According to the method provided by the embodiment of the application, firstly, a water body area is extracted from a target hyperspectral image through a reflectivity threshold value of a near-infrared band, and a first water body area is obtained; then, performing water body region extraction on the first water body region through a water body index model established based on spectral characteristics of visible light and near-infrared bands to obtain a second water body region; and finally, denoising the second water body region to obtain a water body region in the target hyperspectral image. Referring to fig. 2, the method performs water body region extraction twice by fully utilizing the spectral information in the hyperspectral image, so that the extraction precision of the water body region can be improved, and the non-water body region and the water body region in the hyperspectral image can be distinguished.
Further, in the embodiment of the present application, please refer to fig. 3, step S102 can be implemented by the following sub-steps.
And a substep S1021 of obtaining spectral component images of the target hyperspectral image on a plurality of sub-near infrared wave bands of the near infrared wave bands.
For example, if there are N sub-near-infrared bands in the near-infrared band with the wavelength of 760nm to 900nm, spectral component images on the N sub-near-infrared bands are obtained, and a pixel point in the spectral component image of each sub-near-infrared band corresponds to a gray scale value.
And a substep S1022, calculating to obtain a spectrum mean image of the target hyperspectral image on the near-infrared waveband based on the spectrum component images on the plurality of sub-near-infrared spectrum subbands.
And dividing the sum of the gray-scale values of the same pixel point in the images with different spectral components by the number N of the sub near-infrared bands to obtain the average value of the pixel point, so that the spectral average image of the target hyperspectral image in the near-infrared bands can be obtained.
And a substep S1023 of extracting the water body area from the spectrum mean value image by adopting a reflectivity threshold value of a near-infrared band to obtain a first water body area.
The reflectivity threshold of the near infrared band can be used for distinguishing the reflectivity of the water body and the suspected water body from other land surface objects.
Further, in this embodiment of the application, before the substep S1023, the method for extracting a water body region of a hyperspectral image provided in this embodiment of the application may further include a step of determining a reflectance threshold of a near-infrared band, and this step may be implemented in the following manner.
Firstly, a hyperspectral image of a sample is obtained.
In this embodiment, the sample hyperspectral image may include land surface objects of different types, where the land surface objects may include asphalt, light buildings, dark buildings, urban bare land, non-urban bare land, and vegetation that are easily distinguished from water by the reflectivity of the near-infrared band, where the light buildings refer to buildings that are not occluded, and the buildings with higher gray scale values (e.g., gray scale values greater than a gray scale threshold) that make up pixels in the hyperspectral image; the dark building refers to a building which is shielded and forms pixels in the hyperspectral image, and similarly, the non-urban bare land also refers to soil which forms pixels in the hyperspectral image and has a low gray scale value. The land surface object may further include a shadow that is not easily distinguishable from the water body by the reflectivity of the near infrared band, and in addition, the water body in the land surface object may exist in the form of a shadow water body, an urban water body, and a non-urban water body.
Then, determining a target reflectivity capable of distinguishing the water body and the suspected water body from other types of land surface objects according to the reflectivity of the land surface objects in the hyperspectral image of the sample in the near infrared band, and determining the target reflectivity as a reflectivity threshold of the near infrared band.
Referring to fig. 4-7, reflectivity profiles of different types of land surface objects in a sample hyperspectral image in different wavelength bands are illustrated, wherein fig. 4 shows the reflectivity profiles of different types of land surface objects in a blue wavelength band, fig. 5 shows the reflectivity profiles of different types of land surface objects in a green wavelength band, fig. 6 shows the reflectivity profiles of different types of land surface objects in a red wavelength band, and fig. 7 shows the reflectivity profiles of different types of land surface objects in a near infrared wavelength band. From the graph, the reflectivity of the near-infrared band can obviously distinguish the water body from other types of land surface objects, and the distinguishing effect is good. As shown in fig. 7, when the reflectivity threshold is 8%, the shadow water is only classified into one type, and other land surface objects and water bodies are not classified into one type, so that a good distinguishing effect can be achieved. In the embodiment of the application, the value range of the reflectivity threshold can be set to be 7% -9%.
Further, in this embodiment of the application, before step S103, the method for extracting a water body region of a hyperspectral image provided in this embodiment of the application may further include a step of creating the water body index model, and this step may be implemented in the following manner.
Firstly, determining a target position where curve trend differences exist between a reflectivity curve corresponding to a water body and a reflectivity curve corresponding to a suspected water body by analyzing reflectivity curves of the water body and the suspected water body in visible light and near infrared bands, and acquiring a first sub-band positioned in front of the target position and a second sub-band positioned behind the target position.
And then, establishing a water body index model based on the wave band reflectivity corresponding to the first sub-wave band and the wave band reflectivity corresponding to the second sub-wave band.
Referring to fig. 8, fig. 8 illustrates a graph of reflectance curve analysis of water and shadow in visible and near infrared bands, and as shown in the graph, the reflectance curves of the shadow water, urban water and non-urban water have a trend change at about 700nm, i.e. an upward trend changes into a downward trend. The reflectivity curve of the simple shadow is in an ascending trend in the range of 680nm to 750 nm. A water body index model can be constructed through the difference of the trend changes, specifically, a target position (namely a peak position) with the trend changes is determined firstly, and then a first sub-wave band (for example, a wave band of 695 nm-705 nm) where the target position is located and a second sub-wave band (for example, a wave band of 725 nm-735 nm) located behind the target position are obtainedWavelength band), for example, the reflectance of the wavelength band corresponding to the first sub-wavelength band is the reflectance corresponding to the wavelength 700nm
And the reflectivity of the wave band corresponding to the second sub-wave band is the reflectivity corresponding to the wavelength of 730nm
Then, the constructed water body index model can be as follows:
wherein the content of the first and second substances,
the threshold value is preset, and optionally,
the water index model can be set to be 0, and the water body and the suspected water body can be distinguished through the water index model.
In the embodiment of the present application, step S103 may be implemented in the following manner.
Firstly, calculating the wave band reflectivity difference value between the wave band reflectivity of any pixel point in the first sub-wave band and the wave band reflectivity of the pixel point in the second sub-wave band in the first water body area.
Then, taking the pixel points in the first water body area, of which the wave band reflectivity difference is smaller than or equal to a preset threshold value, as first pixel points, taking the pixel points in the first water body area, of which the wave band reflectivity difference is larger than the preset threshold value, as second pixel points, and taking the area formed by the first pixel points as the second water body area.
In the embodiment of the present application, step S104 may be implemented in the following manner.
And denoising the second water body region by adopting a median filtering mode, and taking the denoised second water body region as the water body region in the target hyperspectral image. In particular, median filtering may be performed using a 7 x 7 size filter mask.
In order to verify the reliability of the water body region extraction method of the hyperspectral images, the water body region extraction results of the hyperspectral images of three different image sources are evaluated simultaneously through the prior art and the scheme of the application. Fig. 9-11 illustrate three hyperspectral images from different sources, fig. 9 is an AISA hyperspectral image of the urban area of Jiaxing city, fig. 10 is an OHS-2D hyperspectral image of the urban area of Changsha city and the suburban area, and fig. 11 is a hyperspectral image of a visible and NIR imaging spectrometer of the Harvest-Sanda Bay village. The following evaluation indexes can be adopted in the embodiments of the present application: user precision (UA), producer Precision (PA), overall precision (OA), Kappa coefficient (Kappa), metric (F1), and average cross-over ratio (miou). Their calculation formula is as follows:
wherein the content of the first and second substances,
is the true yang rate,
The false positive rate,
Is the false yin rate,
Is the true yin rate,
The number of pixels of the hyperspectral image,
As a process parameter, the closer the value of the evaluation index is to 1, the better the classification effect is, i.e., the better the extraction result on the water body region is. Referring to fig. 12, fig. 12 illustrates values of evaluation indexes after three high-spectrum images are extracted from a water body by using different schemes, and as can be seen from fig. 12, the overall performance of the evaluation indexes extracted from the three high-spectrum images by using the scheme provided by the present application is higher than that of a high-spectrum differential water index method (HDWI) and a normalized differential water index method (NDWI), and is equivalent to that of a support vector machine method (SVM), which indicates that the scheme provided by the present application has a better processing capability for hyperspectral images from different image sources and in different water body environments.
Further, referring to fig. 13, fig. 13 is a functional module schematic diagram of the water body region extraction device 100 of a hyperspectral image according to an embodiment of the present application, in which the water body region extraction device 100 of a hyperspectral image according to an embodiment of a method executed by an electronic device in the embodiment of the present application is divided into functional modules, that is, the following functional modules corresponding to the water body region extraction device 100 of a hyperspectral image may be used to execute the above method embodiments. The hyperspectral image-based water body region extraction device 100 may include an acquisition module 110, a first extraction module 120, a second extraction module 130, and a processing module 140, and the functions of the hyperspectral image-based water body region extraction device 100 are described in detail below.
The acquiring module 110 is configured to acquire a target hyperspectral image including a water body area to be extracted.
The obtaining module 110 may use hyperspectral images captured by a satellite or an aerial (e.g., captured by an unmanned aerial vehicle) as candidate hyperspectral images, and select a hyperspectral image including a water body area from the candidate hyperspectral images as a target hyperspectral image. The target hyperspectral image may include hyperspectral images of different areas, such as a hyperspectral image of an urban area, a hyperspectral image of a suburban area, a hyperspectral image of a rural area, and the like; the water body area can correspond to rivers, lakes and other water areas.
After the target hyperspectral image is acquired, preprocessing can be performed on the target hyperspectral image, wherein the preprocessing can include at least one of geometric correction processing, atmospheric correction processing, radiation correction processing and normalization processing.
For example, the preprocessing of the AISA hyperspectral image includes geometric correction, radiation correction, atmospheric correction and normalization, the preprocessing of the OHS-2D hyperspectral image includes radiation correction, atmospheric correction and normalization, and the preprocessing of the hyperspectral image of the visible and NIR imaging spectrometer includes normalization, wherein the normalization normalizes the reflectivity of the hyperspectral image to a uniform range (e.g., 0-1). In addition, the target hyperspectral images from different image sources may also be subjected to different software when the same preprocessing is performed, for example, the AISA hyperspectral images can be subjected to radiation correction and atmospheric correction by adopting MODTRAN 5.3.2 software, and the OHS-2D hyperspectral images can be subjected to radiation correction and atmospheric correction by adopting ENVI software.
In this embodiment, the obtaining module 110 may be configured to perform the step S101, and for the detailed implementation of the obtaining module 110, reference may be made to the detailed description of the step S101.
The first extraction module 120 is configured to perform water body region extraction on the target hyperspectral image based on a reflectivity threshold of a near-infrared band, so as to obtain a first water body region.
Because the reflectivity of the near-infrared band has better distinguishing performance for the water body and the non-water objects on most land surfaces than the reflectivity of other bands (such as a blue band, a green band, and a red band), in this embodiment of the application, the first extraction module 120 can distinguish most of the non-water objects in the target hyperspectral image from the water body by using the reflectivity threshold of the near-infrared band, so as to extract the first water body region from the target hyperspectral image. The first water body area is composed of a water body and a suspected water body, wherein the suspected water body can comprise a shadow in the target hyperspectral image. The reflectivity of the suspected water body in the near infrared band is similar to the reflectivity of the water body in the near infrared band, so that the suspected water body and the water body cannot be distinguished by the reflectivity threshold value of the near infrared band.
The first extraction module 120 may be configured to perform the step S102, and for a detailed implementation of the first extraction module 120, reference may be made to the detailed description of the step S102.
And a second extraction module 130, configured to perform water body region extraction on the first water body region based on the water body index model, so as to obtain a second water body region composed of water bodies.
The water body index model can be determined according to the spectral characteristics of the water body and the suspected water body in visible light and near infrared bands, and the water body index model can be used for distinguishing the water body and the suspected water body in the first water body area.
The second extraction module 130 may be configured to perform the step S103, and as to the detailed implementation of the second extraction module 130, reference may be made to the detailed description of the step S103.
And the processing module 140 is configured to perform denoising processing on the second water body region, and use the denoised second water body region as a water body region in the target hyperspectral image.
Since noise interference may include individual water body pixels in the extraction result, which may affect the extraction accuracy, the processing module 140 is configured to remove individual water body pixels existing in the extracted second water body region.
The processing module 140 may be configured to perform the step S104, and for the detailed implementation of the processing module 140, reference may be made to the detailed description of the step S104.
It should be noted that the division of the modules in the above apparatus or system is only a logical division, and the actual implementation may be wholly or partially integrated into one physical entity or may be physically separated. And these modules can be implemented in the form of software (e.g., open source software) that can be invoked by a processor; or may be implemented entirely in hardware; and part of the modules can be realized in the form of calling software by a processor, and part of the modules can be realized in the form of hardware. As an example, the first extraction module 120 may be implemented by a single processor, and may be stored in a memory of the apparatus or system in the form of program codes, and a certain processor of the apparatus or system calls and executes the functions of the first extraction module 120, and the implementation of other modules is similar, and thus will not be described herein again. In addition, the modules can be wholly or partially integrated together or can be independently realized. The processor described herein may be an integrated circuit with signal processing capability, and in the implementation process, each step or each module in the above technical solutions may be implemented in the form of an integrated logic circuit in the processor or a software program executed.
Referring to fig. 14, fig. 14 is a schematic diagram of a hardware structure of an electronic device 10 for implementing the method for extracting a water body region of a hyperspectral image according to an embodiment of the present disclosure. As shown in fig. 14, the electronic device 10 may include a processor 11, a computer-readable storage medium 12, and a bus 13.
In a specific implementation process, the processor 11 executes computer-executable instructions stored in the computer-readable storage medium 12 (for example, the water body region extraction device 100 of the hyperspectral image shown in fig. 13 includes the acquisition module 110, the first extraction module 120, the second extraction module 130, and the processing module 140), so that the processor 11 can execute the water body region extraction method of the hyperspectral image according to the above method embodiment, where the processor 11 and the computer-readable storage medium 12 may be connected through the bus 13.
For a specific implementation process of the processor 11, reference may be made to the above-mentioned method embodiments executed by the electronic device 10, and implementation principles and technical effects are similar, and details of the embodiment of the present application are not described herein again.
The bus 13 may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, the buses in the figures of the present application are not limited to only one bus or one type of bus.
In addition, an embodiment of the present application further provides a readable storage medium, where a computer executing instruction is stored in the readable storage medium, and when a processor executes the computer executing instruction, the method for extracting a water body region of a hyperspectral image as described above is implemented.
To sum up, according to the method, the device and the electronic device for extracting the water body region of the hyperspectral image provided by the embodiment of the application, firstly, the water body region is extracted from the target hyperspectral image through the reflectivity threshold value of the near-infrared band, so that a first water body region is obtained; then, performing water body region extraction on the first water body region through a water body index model established based on spectral characteristics of visible light and near-infrared bands to obtain a second water body region; and finally, denoising the second water body region to obtain a water body region in the target hyperspectral image. According to the method, the spectral information in the hyperspectral image is fully utilized to extract the water body area twice, so that the extraction precision of the water body area can be improved, and the non-water body area and the water body area in the hyperspectral image can be distinguished.
The embodiments described above are only a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, as generally described and illustrated in the figures herein, may be arranged and designed in a wide variety of different configurations. Thus, the detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the application, but is merely representative of selected embodiments of the application. Based on this, the protection scope of the present application shall be subject to the protection scope of the claims. Moreover, all other embodiments that can be made available by a person skilled in the art without making any inventive step based on the embodiments of the present application shall fall within the scope of protection of the present application.
Claims (10)
1. A method for extracting a water body region of a hyperspectral image is characterized by comprising the following steps:
acquiring a target hyperspectral image containing a water body area to be extracted;
performing water body area extraction on the target hyperspectral image based on a reflectivity threshold of a near-infrared band to obtain a first water body area, wherein the first water body area consists of a water body and a suspected water body, and the reflectivity of the suspected water body in the near-infrared band is similar to the reflectivity of the water body in the near-infrared band;
performing water body area extraction on the first water body area based on a water body index model to obtain a second water body area consisting of the water body, wherein the water body index model is determined based on the spectral characteristics of the water body and the suspected water body in visible light and near infrared bands, and the water body index model is used for distinguishing the water body and the suspected water body in the first water body area;
and denoising the second water body region, and taking the denoised second water body region as a water body region in the target hyperspectral image.
2. The method for extracting the hyperspectral image of the water body area according to claim 1, wherein the step of obtaining the target hyperspectral image containing the water body area to be extracted comprises:
taking a satellite hyperspectral image or an aviation hyperspectral image containing a water body area as the target hyperspectral image, wherein the resolution of the satellite hyperspectral image and the aviation hyperspectral image is meter-level or sub-meter-level;
before the step of performing water body region extraction on the target hyperspectral image based on the reflectivity threshold value of the near-infrared band to obtain a first water body region, the method further comprises the following steps of:
and preprocessing the target hyperspectral image, wherein the preprocessing comprises at least one of geometric correction processing, atmospheric correction processing, radiation correction processing and normalization processing.
3. The method for extracting the water body region of the hyperspectral image according to claim 2, wherein the step of extracting the water body region of the target hyperspectral image based on the reflectance threshold of the near-infrared band to obtain the first water body region comprises:
acquiring spectral component images of the target hyperspectral image on a plurality of sub near-infrared wave bands of the near-infrared wave band;
calculating to obtain a spectrum mean image of the target hyperspectral image on the near-infrared wave band based on the spectrum component images on the plurality of sub-near-infrared spectrum sub-wave bands;
and extracting the water body area from the spectrum mean value image by adopting the reflectivity threshold value of the near-infrared band to obtain the first water body area.
4. The method for extracting the water body region of the hyperspectral image according to claim 3, wherein before the step of extracting the water body region from the spectrum mean image by using the reflectance threshold of the near-infrared band to obtain the first water body region, the method further comprises a step of determining the reflectance threshold of the near-infrared band, and the step comprises:
acquiring a hyperspectral image of a sample, wherein the hyperspectral image of the sample comprises land surface objects of different types, and the land surface objects comprise asphalt, bright buildings, dark buildings, urban bare land, non-urban bare land, vegetation, shadows, shadow water bodies, urban water areas and non-urban water areas;
determining target reflectivity capable of distinguishing a water body and a suspected water body from other types of land surface objects according to the reflectivity of the land surface objects in the hyperspectral image of the sample in the near infrared band, and determining the target reflectivity as a reflectivity threshold of the near infrared band, wherein the water body comprises a shadow water body, an urban water body and a non-urban water body, and the suspected water body comprises a shadow.
5. The method for extracting the water body region according to the hyperspectral image of claim 4, wherein before the step of performing the water body region extraction on the first water body region based on the water body index model to obtain the second water body region consisting of the water body, the method further comprises the step of creating the water body index model, and the step comprises:
determining a target position where a reflectivity curve corresponding to the water body and a reflectivity curve corresponding to the suspected water body have curve trend difference through analyzing reflectivity curves of the water body and the suspected water body in visible light and near infrared bands, and acquiring a first sub-band where the target position is located and a second sub-band located behind the target position;
and establishing a water body index model based on the wave band reflectivity corresponding to the first sub-wave band and the wave band reflectivity corresponding to the second sub-wave band.
6. The method for extracting the water body region according to the hyperspectral image of claim 5, wherein the step of performing the water body region extraction on the first water body region based on the water body index model to obtain the second water body region composed of the water body comprises:
calculating a wave band reflectivity difference value between the wave band reflectivity of any pixel point in the first sub-wave band and the wave band reflectivity of the pixel point in the second sub-wave band in the first water body area;
taking the pixel points in the first water body area, of which the wave band reflectivity difference is smaller than or equal to a preset threshold value, as first pixel points, taking the pixel points in the first water body area, of which the wave band reflectivity difference is larger than the preset threshold value, as second pixel points, and taking the area formed by the first pixel points as the second water body area.
7. The method for extracting the water body region of the hyperspectral image according to claim 6, wherein the step of denoising the second water body region and using the denoised second water body region as the water body region in the target hyperspectral image comprises:
and denoising the second water body region by adopting a median filtering mode, and taking the denoised second water body region as a water body region in the target hyperspectral image, wherein the median filtering adopts a filtering mask with the size of 7 x 7.
8. The method for extracting a water body region from a hyperspectral image according to any of claims 5 to 7,
the range of the reflectivity threshold value of the near-infrared band is 7% -9%;
the wavelength range of the first sub-band is 695 nm-705 nm, and the wavelength range of the second sub-band is 725 nm-735 nm.
9. The utility model provides a water regional extraction element of high spectral image which characterized in that, the device includes:
the acquisition module is used for acquiring a target hyperspectral image containing a water body area to be extracted;
the first extraction module is used for extracting a water body area from the target hyperspectral image based on a reflectivity threshold of a near-infrared band to obtain a first water body area, wherein the first water body area is composed of a water body and a suspected water body, and the reflectivity of the suspected water body in the near-infrared band is similar to the reflectivity of the water body in the near-infrared band;
the second extraction module is used for extracting a water body area from the first water body area based on a water body index model to obtain a second water body area consisting of the water body, wherein the water body index model is determined based on the spectral characteristics of the water body and the suspected water body in visible light and near infrared bands, and the water body index model is used for distinguishing the water body and the suspected water body in the first water body area;
and the processing module is used for denoising the second water body region, and taking the denoised second water body region as the water body region in the target hyperspectral image.
10. An electronic device, comprising a processor and a computer-readable storage medium, wherein the processor and the computer-readable storage medium are connected through a bus system, the computer-readable storage medium is used for storing a program, an instruction or a code, and the processor is used for executing the program, the instruction or the code in the computer-readable storage medium to implement the method for extracting a water body region of a hyperspectral image according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210034953.4A CN114049571A (en) | 2022-01-13 | 2022-01-13 | Method and device for extracting water body area of hyperspectral image and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210034953.4A CN114049571A (en) | 2022-01-13 | 2022-01-13 | Method and device for extracting water body area of hyperspectral image and electronic equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114049571A true CN114049571A (en) | 2022-02-15 |
Family
ID=80196463
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210034953.4A Pending CN114049571A (en) | 2022-01-13 | 2022-01-13 | Method and device for extracting water body area of hyperspectral image and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114049571A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117649612A (en) * | 2023-10-19 | 2024-03-05 | 成都大学 | Satellite hyperspectral remote sensing data surface water body extraction method based on hybrid algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046087A (en) * | 2015-08-04 | 2015-11-11 | 中国资源卫星应用中心 | Water body information automatic extraction method for multi-spectral image of remote sensing satellite |
| CN107421892A (en) * | 2016-05-23 | 2017-12-01 | 核工业北京地质研究院 | A kind of hyperspectral data processing method for water body information |
| CN108734122A (en) * | 2018-05-17 | 2018-11-02 | 北京理工大学 | A kind of EO-1 hyperion city water body detection method based on adaptive samples selection |
-
2022
- 2022-01-13 CN CN202210034953.4A patent/CN114049571A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105046087A (en) * | 2015-08-04 | 2015-11-11 | 中国资源卫星应用中心 | Water body information automatic extraction method for multi-spectral image of remote sensing satellite |
| CN107421892A (en) * | 2016-05-23 | 2017-12-01 | 核工业北京地质研究院 | A kind of hyperspectral data processing method for water body information |
| CN108734122A (en) * | 2018-05-17 | 2018-11-02 | 北京理工大学 | A kind of EO-1 hyperion city water body detection method based on adaptive samples selection |
Non-Patent Citations (1)
| Title |
|---|
| JIAWEI YANG等: "Water extraction of hyperspectral imagery based on a fast and effective decision tree water index", 《JOURNAL OF APPLIED REMOTE SENSING》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117649612A (en) * | 2023-10-19 | 2024-03-05 | 成都大学 | Satellite hyperspectral remote sensing data surface water body extraction method based on hybrid algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yousif et al. | Improving urban change detection from multitemporal SAR images using PCA-NLM | |
| US9779503B2 (en) | Methods for measuring the efficacy of a stain/tissue combination for histological tissue image data | |
| CN108765465B (en) | Unsupervised SAR image change detection method | |
| CN110120046B (en) | Potential landslide identification method integrating DEM (digital elevation model), optical remote sensing and deformation information | |
| US20220129674A1 (en) | Method and device for determining extraction model of green tide coverage ratio based on mixed pixels | |
| CN115082451B (en) | Stainless steel soup ladle defect detection method based on image processing | |
| CN112307901B (en) | SAR and optical image fusion method and system for landslide detection | |
| Zhai | Inversion of organic matter content in wetland soil based on Landsat 8 remote sensing image | |
| van Zwanenberg et al. | Edge detection techniques for quantifying spatial imaging system performance and image quality | |
| Zhang et al. | Salient region detection in remote sensing images based on color information content | |
| CN114049571A (en) | Method and device for extracting water body area of hyperspectral image and electronic equipment | |
| CN117575953B (en) | Detail enhancement method for high-resolution forestry remote sensing image | |
| Shi et al. | Automatic shadow detection in high-resolution multispectral remote sensing images | |
| CN113962900A (en) | Method, device, equipment and medium for detecting infrared dim target under complex background | |
| CN114066862A (en) | Indicator identification method and system based on color gamut and contour characteristics | |
| CN116188510B (en) | Enterprise emission data acquisition system based on multiple sensors | |
| CN112651945A (en) | Multi-feature-based multi-exposure image perception quality evaluation method | |
| CN109241865B (en) | Vehicle detection segmentation algorithm under weak contrast traffic scene | |
| CN111738984A (en) | Skin image spot evaluation method and system based on watershed and seed filling | |
| CN113284066B (en) | Automatic cloud detection method and device for remote sensing image | |
| CN114170145B (en) | Heterogeneous remote sensing image change detection method based on multi-scale self-coding | |
| Shi et al. | Urban feature shadow extraction based on high-resolution satellite remote sensing images | |
| CN113822361B (en) | SAR image similarity measurement method and system based on Hamming distance | |
| CN117252789B (en) | Shadow reconstruction method and device for high-resolution remote sensing image and electronic equipment | |
| Afreen et al. | A method of shadow detection and shadow removal for high resolution 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220215 |
|
| RJ01 | Rejection of invention patent application after publication |