Greenhouse identification method and device, electronic equipment and computer readable storage medium

Abstract

The invention provides a greenhouse identification method and device, electronic equipment and a computer readable storage medium, and relates to the technical field of remote sensing image processing.The greenhouse identification method comprises the steps that during greenhouse identification, original remote sensing image data of a target area are obtained firstly, and the original remote sensing image data comprise reflectivity data of a preset wave band; preprocessing the original remote sensing image data to obtain target remote sensing image data; and performing greenhouse identification on the target remote sensing image data based on a preset modified greenhouse index MPGI to obtain a target greenhouse identification result of the target area. Therefore, the modified greenhouse index MPGI is used for carrying out greenhouse identification on the preprocessed target remote sensing image data, the difference between the greenhouse and the non-greenhouse in the preset wave band is considered, and the initial discrimination parameters are used for enhancing the identification of the greenhouse and inhibiting the information of the non-greenhouse type, so that the greenhouse identification accuracy and the identification effect are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing image processing, and in particular to a greenhouse identification method, device, electronic equipment and computer-readable storage medium. Background Art

[0002] In the past few decades, plastic greenhouses have rapidly changed the agricultural model around the world. Compared with traditional agriculture, greenhouses are an effective modern model that can protect crops from unstable environments and provide off-season crops, and are widely used in the agricultural field. Although plastic greenhouses can produce high-quality crops, their widespread use has also caused some environmental problems. First, greenhouses produce plastic waste, leading to soil degradation and water pollution. Second, intensive agricultural models change land cover, further affecting landscapes and ecosystem services. Finally, greenhouses affect the aesthetics of the landscape, such as reduced visual clarity and reduced regional biodiversity. Therefore, accurately mapping the spatiotemporal distribution of greenhouses is necessary for environmental monitoring and the implementation of agricultural policies. The continuous increase in remote sensing data and the maturity of analysis methods have made it common to use remote sensing technology to monitor land cover, which provides the possibility of extracting greenhouses.

[0003] There are two types of greenhouse identification methods using remote sensing: one is the feature extraction method using land cover type characteristics, such as using object-based image analysis and decision tree classification to detect greenhouses; the other is the index-based extraction method, such as using the vegetable index (VegetationIndex, abbreviated as VI, or V I ) and other greenhouse indices to extract greenhouses. However, the feature extraction method requires high spatial resolution image support (such as greenhouse recognition based on images with a spatial resolution of 0.5m). This kind of image data is difficult to obtain, has weak spatiotemporal continuity, and requires a large number of training samples as a basis. For large-scale recognition, the feature extraction effect is poor. For the index-based extraction method, the problem of different objects with the same spectrum and different spectra of the same object is particularly prominent, resulting in poor recognition effect. Summary of the invention

[0004] The object of the present invention is to provide a greenhouse identification method, device, electronic equipment and computer-readable storage medium to improve the greenhouse identification accuracy and identification effect.

[0005] In a first aspect, an embodiment of the present invention provides a greenhouse identification method, comprising:

[0006] Acquire original remote sensing image data of the target area, wherein the original remote sensing image data includes reflectance data of preset bands, wherein the preset bands include coastal aerosol band, blue band, green band, near infrared band, shortwave infrared 1 and shortwave infrared 2;

[0007] Preprocessing the original remote sensing image data to obtain target remote sensing image data;

[0008] Based on the preset modified greenhouse index MPGI, greenhouse identification is performed on the target remote sensing image data to obtain the target greenhouse identification result of the target area; wherein the MPGI is defined as follows:

[0009]

[0010] v=(R Green -R Blue )×(R NIR -R SWIR1 -0.02);

[0011] Among them, R coastal is the reflectivity of the coastal aerosol band, R Blue is the reflectivity of the blue band, R Green is the reflectance of the green band, R NIR is the reflectivity in the near-infrared band, R SWIR1 is the reflectivity of short-wave infrared 1, R SWIR2 is the reflectivity of shortwave infrared 2, and v is the preset initial discrimination parameter.

[0012] Furthermore, the preprocessing of the original remote sensing image data to obtain target remote sensing image data includes:

[0013] Performing radiation calibration, atmospheric correction and cropping of the target area on the original remote sensing image data to obtain processed data;

[0014] The processed data is subjected to removal of typical land object types to obtain target remote sensing image data; wherein the typical land object types include one or more of forests, water bodies, snow and bare soil.

[0015] Furthermore, the removing of typical ground object types from the processed data to obtain target remote sensing image data includes:

[0016] Based on the normalized vegetation index, the improved water body index and the normalized difference bare soil index, the processed data is used to identify the typical land object type to obtain a non-greenhouse identification result;

[0017] The non-greenhouse area data corresponding to the non-greenhouse identification result is eliminated from the processed data to obtain target remote sensing image data.

[0018] Furthermore, the greenhouse identification is performed on the target remote sensing image data based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area, including:

[0019] Traversing each pixel in the target remote sensing image data;

[0020] For the traversed current pixel, based on the reflectivity data corresponding to the current pixel in the target remote sensing image data, the MPGI data corresponding to the current pixel is calculated, and the greenhouse recognition result of the current pixel is determined based on the MPGI data; wherein the MPGI data includes an initial discrimination parameter value and an MPGI value;

[0021] Until all pixels in the target remote sensing image data are traversed, the target greenhouse recognition result of the target area is obtained.

[0022] Furthermore, the MPGI data corresponding to the current pixel is calculated based on the reflectivity data corresponding to the current pixel in the target remote sensing image data, and the greenhouse recognition result of the current pixel is determined based on the MPGI data, including:

[0023] Based on the first reflectivity data corresponding to the current pixel in the target remote sensing image data, an initial discrimination parameter value corresponding to the current pixel is calculated; wherein the first reflectivity data includes a reflectivity value of a blue band, a reflectivity value of a green band, a reflectivity value of a near infrared band, and a reflectivity value of a short-wave infrared 1;

[0024] Determine whether the initial discrimination parameter value corresponding to the current pixel is greater than 0, and obtain a first judgment result;

[0025] When the first judgment result is no, determining that the current pixel belongs to a non-greenhouse area;

[0026] When the first judgment result is yes, the MPGI value corresponding to the current pixel is calculated based on the second reflectivity data corresponding to the current pixel in the target remote sensing image data; wherein the second reflectivity data includes the reflectivity value of the coastal aerosol band, the reflectivity value of the blue band, the reflectivity value of the near infrared band, the reflectivity value of the shortwave infrared 1 and the reflectivity value of the shortwave infrared 2;

[0027] Determine whether the MPGI value corresponding to the current pixel is greater than or equal to a preset MPGI threshold, and obtain a second determination result;

[0028] When the second judgment result is yes, determining that the current pixel belongs to the greenhouse area;

[0029] When the second judgment result is no, it is determined that the current pixel belongs to a non-greenhouse area.

[0030] Furthermore, after greenhouse identification is performed on the target remote sensing image data based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area, the greenhouse identification method further includes:

[0031] Based on the target greenhouse identification result, a greenhouse distribution map of the target area is generated.

[0032] Furthermore, after greenhouse identification is performed on the target remote sensing image data based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area, the greenhouse identification method further includes:

[0033] Based on the target greenhouse identification result, the greenhouse ratio value of the target area is calculated.

[0034] In a second aspect, an embodiment of the present invention further provides a greenhouse identification device, comprising:

[0035] A data acquisition module is used to acquire original remote sensing image data of a target area, wherein the original remote sensing image data includes reflectance data of preset bands, wherein the preset bands include a coastal aerosol band, a blue band, a green band, a near infrared band, a short-wave infrared 1, and a short-wave infrared 2;

[0036] A preprocessing module, used for preprocessing the original remote sensing image data to obtain target remote sensing image data;

[0037] The greenhouse identification module is used to perform greenhouse identification on the target remote sensing image data based on a preset modified greenhouse index MPGI to obtain a target greenhouse identification result of the target area; wherein the MPGI is defined as follows:

[0038]

[0039] v=(R Green -R Blue )×(R NIR -R SWIR1 -0.02);

[0040] Among them, R coastal is the reflectivity of the coastal aerosol band, R Blue is the reflectivity of the blue band, R Green is the reflectance of the green band, R NIR is the reflectivity in the near-infrared band, R SWIR1 is the reflectivity of short-wave infrared 1, R SWIR2 is the reflectivity of shortwave infrared 2, and v is the preset initial discrimination parameter.

[0041] In a third aspect, an embodiment of the present invention further provides an electronic device, including a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the greenhouse identification method described in the first aspect is implemented.

[0042] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the greenhouse identification method described in the first aspect is executed.

[0043] The greenhouse identification method, device, electronic device and computer-readable storage medium provided by the embodiment of the present invention, when performing greenhouse identification, first obtain the original remote sensing image data of the target area, the original remote sensing image data includes the reflectivity data of the preset band, the preset band includes the coastal aerosol band, the blue band, the green band, the near infrared band, the short wave infrared 1 and the short wave infrared 2; then pre-process the original remote sensing image data to obtain the target remote sensing image data; then based on the preset modified greenhouse index MPGI, greenhouse identification is performed on the target remote sensing image data to obtain the target greenhouse identification result of the target area. In this way, the modified greenhouse index MPGI is used to perform greenhouse identification on the pre-processed target remote sensing image data, taking into account the difference between greenhouses and non-greenhouses in the preset bands, and using the initial discrimination parameters to strengthen the identification of greenhouses and suppress the information of non-greenhouse types, thereby improving the greenhouse identification accuracy and identification effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 A schematic diagram of a flow chart of a greenhouse identification method provided by an embodiment of the present invention;

[0046] Figure 2 A schematic flow chart of another greenhouse identification method provided by an embodiment of the present invention;

[0047] Figure 3 Reflectivity curves of various types of ground objects in different bands provided by the embodiment of the present invention;

[0048] Figure 4 Identification results corresponding to various greenhouse indexes provided in the embodiments of the present invention;

[0049] Figure 5The F1 values ​​of various greenhouse indexes provided by the embodiment of the present invention at different thresholds;

[0050] Figure 6 A schematic diagram of the structure of a greenhouse identification device provided by an embodiment of the present invention;

[0051] Figure 7 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] For greenhouse identification, unlike feature extraction methods, index-based extraction methods have low computational cost, fast extraction speed, and no data training is required. However, due to the similar reflectivity between mixed pixels and greenhouse pixels, mixed pixels of land cover are one of the challenges of land use classification. Among them, mixed pixels refer to the fact that due to the spatial resolution limitation of the sensor and the spatial heterogeneity of the surface cover, a single pixel often cannot completely correspond to a single type of land object, but contains mixed information of multiple types of land objects.

[0054] At present, the existing greenhouse identification methods have little research on fragmented areas, and the problems of different objects with the same spectrum and different spectra with the same object are particularly prominent. The existing greenhouse index has low recognition accuracy in fragmented areas. Among them, the fragmented area refers to an area where a large area of ​​natural landscape is divided into many small pieces due to human activities or other reasons. The fragmented terrain has higher requirements for recognition accuracy. Based on this, a greenhouse identification method, device, electronic device and computer-readable storage medium provided by the embodiment of the present invention propose a modified greenhouse index MPGI suitable for fragmented areas based on the spectrum of the object type (i.e., the reflectivity of multiple object types in different bands) in different bands. Compared with the existing greenhouse index, it has higher accuracy and better extraction results; it has strong adaptability and higher accuracy in different times and images; before using the MPGI index, a separation judgment method (i.e., using v>0 for preliminary judgment of greenhouses) is proposed to strengthen the recognition of greenhouses and suppress non-greenhouse type information.

[0055] It should be noted that the embodiments of the present invention are particularly suitable for fragmented terrain, and are more suitable for other terrains such as continuous terrain.

[0056] To facilitate understanding of this embodiment, a greenhouse identification method disclosed in an embodiment of the present invention is first introduced in detail.

[0057] The embodiment of the present invention provides a greenhouse identification method, which can be executed by an electronic device based on data processing capabilities. Figure 1 The flowchart of a greenhouse identification method shown in FIG. 1 mainly includes the following steps S110 to S130:

[0058] Step S110, obtaining original remote sensing image data of the target area, where the original remote sensing image data includes reflectivity data of a preset band.

[0059] Among them, the preset bands include coastal aerosol band B1, blue band B2, green band B3, near infrared band B5, shortwave infrared 1 (SWIR1) B6 and shortwave infrared 2 (SWIR2) B7, and may also include red band B4 and the like.

[0060] The original remote sensing image data may be, but is not limited to, image data covering the target area collected by the Operational Land Imager (OLI) of the Landsat 8 satellite, and the spatial resolution of the image data is 30 meters. Different bands (such as B1, B2, B3, B4, B5, B6, B7) represent signals in different electromagnetic spectrum intervals received by the OLI sensor, and the specific bands are as follows:

[0061] B1: Coastal aerosol band, with a wavelength range of about 0.43 to 0.45 μm, is used to detect coastal waters and atmospheric aerosols.

[0062] B2: Blue band, with a wavelength range of about 0.45 to 0.51 μm, is suitable for water body detection and distinguishing water bodies from ground objects.

[0063] B3: Green band, wavelength range is about 0.53 to 0.59 μm, used for vegetation monitoring.

[0064] B4: Red band, with a wavelength range of approximately 0.64 to 0.67 μm, is commonly used in agriculture and vegetation health assessment.

[0065] B5: Near-infrared band, with a wavelength range of approximately 0.85 to 0.88 μm, used to measure vegetation coverage.

[0066] B6: Shortwave infrared 1 (SWIR1), with a wavelength range of approximately 1.57 to 1.65 μm, is suitable for detecting soil moisture and mineral types.

[0067] B7: Shortwave infrared 2 (SWIR2), with a wavelength range of approximately 2.11 to 2.29 μm, is used for the identification of rocks and minerals.

[0068] Step S120, preprocessing the original remote sensing image data to obtain target remote sensing image data.

[0069] In order to accurately extract greenhouse information, the original remote sensing image data may be preprocessed before greenhouse identification. In some possible embodiments, the above step S120 may be implemented in the following manner: performing radiation calibration, atmospheric correction and cropping of the target area on the original remote sensing image data to obtain processed data; removing typical land object types from the processed data to obtain target remote sensing image data; wherein the typical land object types include one or more of forests, water bodies, snow and bare soil.

[0070] In one possible implementation, in order to achieve the removal of forests, water bodies, snow and bare soil, the typical land object types are removed from the processed data to obtain the target remote sensing image data. This step can be achieved through the following process: typical land object types are identified on the processed data based on the normalized vegetation index, the improved water body index and the normalized difference bare soil index to obtain non-greenhouse identification results; non-greenhouse area data corresponding to the non-greenhouse identification results are eliminated from the processed data to obtain the target remote sensing image data.

[0071] The above-mentioned normalized difference vegetation index (NDVI) can be expressed as: NDVI = (B5-B4) / (B5+B4), the modified water body index (MNDWI) can be expressed as: MNDWI = (B3-B6) / (B3+B6), and the normalized difference bare soil index (NDBSI) can be expressed as: NDBSI = (B6-B2) / (B6+B2), where B2, B3, B4, B5 and B6 are the reflectances corresponding to the blue band, green band, red band, near infrared band and shortwave infrared 1 (SWIR1), respectively.

[0072] Forests, water bodies, snow and bare soil can be identified using the normalized vegetation index, improved water body index and normalized difference bare soil index. The non-greenhouse identification results can include each pixel belonging to the identified typical landform type, and the reflectance data corresponding to these pixels are the non-greenhouse area data. The non-greenhouse area data is removed from the processed data, and only the remaining pixels in the target area are used for greenhouse identification, which can eliminate the influence of forests, water bodies, snow and bare soil on greenhouse identification.

[0073] It should be noted that the specific process of performing radiation calibration, atmospheric correction and cropping of the target area on the original remote sensing image data can refer to the relevant existing technology and will not be repeated here.

[0074] Step S130, greenhouse recognition is performed on the target remote sensing image data based on the preset MPGI to obtain a target greenhouse recognition result of the target area.

[0075] Among them, the definition of MPGI is as follows:

[0076]

[0077] v=(R Green -R Blue )×(R NIR -R SWIR1 -0.02);

[0078] Among them, R coastal is the reflectivity of the coastal aerosol band, R Blue is the reflectivity of the blue band, R Green is the reflectance of the green band, R NIR is the reflectivity in the near-infrared band, R SWIR1 is the reflectivity of short-wave infrared 1, R SWIR2 is the reflectivity of shortwave infrared 2, and v is the preset initial discrimination parameter.

[0079] In some possible embodiments, the v value is first used for preliminary separation and judgment to strengthen the identification of greenhouses and suppress information of non-greenhouse types, and then the MPGI value is used for final greenhouse judgment to improve the accuracy and recognition effect of the recognition result. Based on this, the above step S130 can be implemented through the following sub-steps 131 to 133:

[0080] Sub-step 131, traverse each pixel in the target remote sensing image data.

[0081] The pixels in the target remote sensing image data are the pixels remaining after removing the pixels belonging to the typical land object type from the pixels corresponding to the target area.

[0082] Sub-step 132, for the current pixel traversed, based on the reflectivity data corresponding to the current pixel in the target remote sensing image data, calculate the MPGI data corresponding to the current pixel, and determine the greenhouse identification result of the current pixel based on the MPGI data; wherein the MPGI data includes the initial discrimination parameter value (i.e., the v value) and the MPGI value.

[0083] In specific implementation, the initial discrimination parameter value corresponding to the current pixel can be calculated based on the first reflectivity data corresponding to the current pixel in the target remote sensing image data; wherein the first reflectivity data includes the reflectivity value of the blue band, the reflectivity value of the green band, the reflectivity value of the near infrared band and the reflectivity value of the short-wave infrared 1; it is judged whether the initial discrimination parameter value corresponding to the current pixel is greater than 0 to obtain a first judgment result; when the first judgment result is no, it is determined that the current pixel belongs to a non-greenhouse area; when the first judgment result is yes, based on the current pixel in the target remote sensing image data The second reflectivity data corresponding to the pixel is used to calculate the MPGI value corresponding to the current pixel; wherein the second reflectivity data includes the reflectivity value of the coastal aerosol band, the reflectivity value of the blue band, the reflectivity value of the near-infrared band, the reflectivity value of the short-wave infrared 1 and the reflectivity value of the short-wave infrared 2; it is determined whether the MPGI value corresponding to the current pixel is greater than or equal to a preset MPGI threshold to obtain a second judgment result; when the second judgment result is yes, it is determined that the current pixel belongs to the greenhouse area; when the second judgment result is no, it is determined that the current pixel belongs to the non-greenhouse area.

[0084] The above MPGI threshold is a preset empirical value, which can be set specifically according to experimental results, and the embodiment of the present invention does not limit this.

[0085] Sub-step 133 is to traverse all pixels in the target remote sensing image data to obtain the target greenhouse recognition result in the target area.

[0086] The target greenhouse recognition result of the target area may include all pixels corresponding to the greenhouse area and all pixels corresponding to the non-greenhouse area in the target area.

[0087] Furthermore, in order to facilitate the user to view the target greenhouse identification result, the above greenhouse identification method also includes: generating a greenhouse distribution map of the target area based on the target greenhouse identification result. When generating the greenhouse distribution map, all pixels corresponding to the greenhouse area in the target area and all pixels corresponding to the non-greenhouse area can be filled with different colors or different grayscales, so that the user can intuitively see the distribution of greenhouses in the target area.

[0088] Furthermore, the greenhouse identification method further includes: based on the target greenhouse identification result, calculating the greenhouse ratio value of the target area. When calculating the greenhouse ratio value, the first number of all pixels corresponding to the greenhouse area in the target area and the second number of all pixels corresponding to the non-greenhouse area can be counted, and the greenhouse ratio value is calculated by the following formula: greenhouse ratio value = first number / (first number + second number). This makes it convenient for users to view the greenhouse ratio of the target area.

[0089] The greenhouse identification method provided by the embodiment of the present invention first obtains the original remote sensing image data of the target area when performing greenhouse identification, and the original remote sensing image data includes the reflectivity data of the preset bands, and the preset bands include the coastal aerosol band, the blue band, the green band, the near infrared band, the short-wave infrared 1 and the short-wave infrared 2; then the original remote sensing image data is preprocessed to obtain the target remote sensing image data; and then the target remote sensing image data is subjected to greenhouse identification based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area. In this way, the modified greenhouse index MPGI is used to perform greenhouse identification on the preprocessed target remote sensing image data, taking into account the difference between greenhouses and non-greenhouses in the preset bands, and using the initial discrimination parameters to strengthen the identification of greenhouses and suppress the information of non-greenhouse types, thereby improving the greenhouse identification accuracy and identification effect.

[0090] For ease of understanding, the above-mentioned greenhouse identification method is introduced in detail below.

[0091] The embodiment of the present invention proposes an improved plastic greenhouse index (i.e., modified greenhouse index, MPGI) for identifying greenhouses in fragmented areas based on the OLI image of Landsat 8. Through the land cover type spectrum curve (i.e., the above-mentioned land feature type spectrum), separability analysis and accuracy evaluation, the greenhouse information in the fragmented area can be effectively enhanced.

[0092] See also Figure 2 The flowchart of another greenhouse identification method shown in FIG. Figure 2 The leftmost side shows the true color images of two regions, with color bands B4, B3 and B2) for spectral analysis to obtain the ground feature type spectrum (i.e. Figure 2 The spectrum of the land object type includes the reflectivity curves of 11 common land object types from B1 to B7. The land object type spectrum can be divided into the first spectrum (i.e. Figure 2 The spectrum at the top of the graph) and the second spectrum (i.e. Figure 2The first spectrum includes the reflectance curves of 4 typical land object types from B1 to B7, and the second spectrum includes the reflectance curves of 7 remaining land object types from B1 to B7. By analyzing the first spectrum, NDVI (i.e., normalized difference vegetation index), mNDWI (i.e., improved water body index) and NDBSI (i.e., normalized difference bare soil index) can be used to perform typical land object analysis of 4 typical land object types; among which, the 4 typical land object types include forest (Forest), water (Water), snow (Snow) and bare soil (Rare Soil). By analyzing the second spectrum, MPGI can be used to perform residual land object analysis of 7 remaining land object types, and these 7 remaining land object types include artificial buildings (AS), blue buildings (BBuilding), mixed vegetation types (MVAS), plastic greenhouse (PG), reflective buildings (RAS), reflective blue buildings (RBB) and roads (RSoil); MPGI can be expressed as follows:

[0093]

[0094] r=R Green -R Blue

[0095] p=R NIR -R SWIR1 -0.02

[0096] v=r×p.

[0097] Combining the recognition results of the first spectrum and the second spectrum, we can get Figure 2 The recognition results corresponding to the MPGI in the far right, i.e. (a1) corresponding to the true color image (f1), and (a2) corresponding to the true color image (f2); Figure 2 The rightmost side also shows the use of APGI, PGI, V I , PMLI and other greenhouse index recognition results, namely (b1), (c1), (d1), (e1) corresponding to the true color image (f1), and (b2), (c2), (d2), (e2) corresponding to the true color image (f2). Among them, the Advanced Plastic Greenhouse Index (APGI), the Plastic Greenhouse Index (PGI), the Vegetable Index (V I ) and the Plastic Mulch Film Index (PMLI) are both designed to identify plastic greenhouses. Figure 2 The vegetation in the identification result shown refers to forest.

[0098] The following is a detailed introduction to the MPGI proposal process.

[0099] Based on Landsat remote sensing image data, common landform types were selected for spectral analysis to find out the differences between greenhouse and non-greenhouse types in different bands. Through preprocessing and separation judgment, greenhouse information was strengthened and non-greenhouse types were suppressed. In addition, a new greenhouse spectral index (MPGI) was proposed based on the differences in B1, B2, B5, B6, and B7. Through visual interpretation and accuracy evaluation, compared with the previous greenhouse index, the new index MPGI has better extraction effect and higher accuracy in fragmented areas.

[0100] 1. Preliminary preparation:

[0101] The OLI images of Landsat 8 with a long time series are used as research data. The OLI images of Landsat 8 can be downloaded from the relevant cloud platform. There are more clouds in summer, and most crops are harvested in summer. Therefore, winter images (January 18, 2020 and March 30, 2020) are selected, and the cloud percentage is less than 0.3%. These images have 7 spectral bands, from the coastal aerosol band to SWIR2, with a spatial resolution of 30 meters. The influence of clouds was eliminated by radiation calibration and atmospheric correction in the ENVI (Environmental Visualization System) software, and the true surface spectral reflectance was obtained. Combined with Google Earth high-definition images, 7 common types of land objects in fragmented areas (excluding the four typical types of land objects: forests, water bodies, snow and bare soil) were selected to analyze the reflectivity of each type in different bands, such as Figure 3 shown.

[0102] 2. Spectral analysis:

[0103] The revised Plastic Greenhouse Index (MPGI) will enhance information on various types of plastic greenhouses. Figure 3 The reflectance spectrum curves of PG, AS, RSoil, BBuilding, RBB, MVAS and RAS. Figure 3 It can be seen that PG, RBB, and MVAS have the highest reflectivity in B5, and the other types have higher reflectivity in B6. Similarly, the reflectivity of PG drops sharply from B5 to B7, while other types of landforms increase or decrease slowly. Therefore, increasing the difference between B5 and B7 can separate greenhouses from non-greenhouses, and the spectral difference characteristics of greenhouses can be captured by the normalized difference index (corresponding to the difference between B5 and B7). In addition, other bands are selected to enhance the information in the greenhouse. Figure 3As shown, in the B1 and B2 bands, the reflectivity of PG is higher than that of AS and MVAS, but lower than that of RSoil, BBuilding, RBB and RAS. Therefore, these two bands are reliable indicators for distinguishing greenhouses from non-greenhouses.

[0104] However, the complex land cover types (i.e., land object types) and the diversity of plastic greenhouses will cause the spectral characteristics of some land covers to be similar to those of greenhouses. Under the combination of NIR-SWIR1 and NIR-SWIR2 bands, the separation degree of PG-RBB and PG-MVAS is poor. Therefore, this embodiment adopts separation judgment. Figure 3 It can be seen that the reflectivity changes of RBB and PG in each band are almost the same, but the reflectivity of RBB decreases between B2 and B3, and we use r to represent the difference. The r values ​​in RBB and BBuilding are negative, and the r values ​​of other types of objects are positive. In addition, MVAS, PG, RBB, and MVAS show different downward trends in the range of B5 to B6. The PG reflectivity curve decreases significantly, while the RAS curve is flat. This amplitude difference is represented by p. Here, we choose the reflectivity change of RBB as the standard, so that the p value of RBB is positive. The p values ​​of PG, RBB and MVAS are positive, while the p values ​​of other types of objects are negative. Based on the above analysis, the r and p products of PG, BBuilding and MVAS are all positive (among them, the r and p of greenhouse and MVAS are all positive; the r and p of BBuilding are all negative), and the r and p of the other types of objects are one positive and one negative (the r of RBB is negative and the p is positive; the r of other types of objects is positive and the p is negative). Let v represent the product of r and p. When v is greater than 0, we propose to use MPGI to further separate greenhouses; otherwise (i.e., v is not greater than 0), it is classified as non-greenhouse. The final form of MPGI is:

[0105] r=R Green -R Blue

[0106] p=R NIR -R SWIR1 -0.02

[0107] v=r×p.

[0108] 3. Index applicable test:

[0109] In order to evaluate the applicability of the index, six greenhouse indices (MPGI, APGI, PGI, PMLI, V Iand PGHI) to extract greenhouses in the study area. It is worth noting that when using MGPI, APGI, PGI, and VI, those greater than their thresholds are classified as PG, while PMLI is the opposite (i.e., those less than its threshold are classified as PG). Figure 4 The true color image (a) and the local enlarged image (b) of a certain area show the greenhouse recognition results. The color bands are B4, B3 and B2, among which (a1) to (a6) are MPGI, APGI, PGI, PMLI, V I The recognition results of the true color image (a) corresponding to PGHI, (b1) to (b6) are MPGI, APGI, PGI, PMLI, V I The recognition result of the local enlarged image (b) corresponding to PGHI. From the visual evaluation, the recognition result of MPGI is consistent with the real image, and MPGI can accurately extract the diversity and density of greenhouses. Among them, the plastic greenhouse index (PGHI) can be expressed as: B2 / B7, and the greenhouse is classified as a greenhouse if it is greater than its threshold. B2 and B7 are the reflectance corresponding to the blue band and short-wave infrared 2 (SWIR2), respectively.

[0110] MPGI, APGI, PGI, PMLI and V I The F1 values ​​at different thresholds are as follows Figure 5 As shown. First, the overall accuracy trends of all greenhouse indices are basically similar, with only one peak and a gradual decrease on both sides of the peak. Second, among the five greenhouse indices, MPGI has the highest mapping accuracy, with an F1 value of 88.6%. APGI is also effective in identifying greenhouses, with an F1 score of 86.7%. PGI and PMLI achieved an accuracy of 70%-80%. The accuracy of VI is less than 40%. It should be noted that the optimal thresholds of the greenhouse indices are different due to the different formulas and data sets corresponding to the greenhouse indices.

[0111] The embodiments of the present invention have the following beneficial effects:

[0112] 1. The separation and judgment step effectively suppresses non-greenhouse type information and enhances greenhouse identification.

[0113] 2. The MPGI index has strong adaptability to time and images, and has higher extraction accuracy in fragmented areas than the existing greenhouse index.

[0114] Corresponding to the above-mentioned greenhouse identification method, the embodiment of the present invention also provides a greenhouse identification device. Figure 6 The structural diagram of a greenhouse identification device shown in FIG. 1 includes:

[0115] The data acquisition module 601 is used to acquire the original remote sensing image data of the target area, wherein the original remote sensing image data includes the reflectivity data of the preset bands, and the preset bands include the coastal aerosol band, the blue band, the green band, the near infrared band, the short wave infrared 1 and the short wave infrared 2;

[0116] A preprocessing module 602 is used to preprocess the original remote sensing image data to obtain target remote sensing image data;

[0117] The greenhouse identification module 603 is used to perform greenhouse identification on the target remote sensing image data based on a preset modified greenhouse index MPGI to obtain a target greenhouse identification result of the target area; wherein the MPGI is defined as follows:

[0118]

[0119] v=(R Green -R Blue )×(R NIR -R SWIR1 -0.02);

[0120] Among them, R coastal is the reflectivity of the coastal aerosol band, R Blue is the reflectivity of the blue band, R Green is the reflectance of the green band, R NIR is the reflectivity in the near-infrared band, R SWIR1 is the reflectivity of short-wave infrared 1, R SWIR2 is the reflectivity of shortwave infrared 2, and v is the preset initial discrimination parameter.

[0121] The greenhouse identification device provided by the embodiment of the present invention first obtains the original remote sensing image data of the target area when performing greenhouse identification, and the original remote sensing image data includes the reflectivity data of the preset bands, and the preset bands include the coastal aerosol band, the blue band, the green band, the near infrared band, the short-wave infrared 1 and the short-wave infrared 2; then the original remote sensing image data is preprocessed to obtain the target remote sensing image data; and then the target remote sensing image data is subjected to greenhouse identification based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area. In this way, the modified greenhouse index MPGI is used to perform greenhouse identification on the preprocessed target remote sensing image data, taking into account the difference between greenhouses and non-greenhouses in the preset bands, and using the initial discrimination parameters to strengthen the identification of greenhouses and suppress the information of non-greenhouse types, thereby improving the greenhouse identification accuracy and identification effect.

[0122] Furthermore, the above-mentioned preprocessing module 602 is specifically used for: performing radiation calibration, atmospheric correction and cropping of the target area on the original remote sensing image data to obtain processed data; removing typical land object types from the processed data to obtain target remote sensing image data; wherein the typical land object types include one or more of forests, water bodies, snow and bare soil.

[0123] Furthermore, the above-mentioned preprocessing module 602 is also used to: identify the typical land feature type on the processed data based on the normalized vegetation index, the improved water body index and the normalized difference bare soil index to obtain a non-greenhouse identification result; and remove the non-greenhouse area data corresponding to the non-greenhouse identification result from the processed data to obtain the target remote sensing image data.

[0124] Furthermore, the above-mentioned greenhouse identification module 603 is specifically used to: traverse each pixel in the target remote sensing image data; for the traversed current pixel, based on the reflectivity data corresponding to the current pixel in the target remote sensing image data, calculate the MPGI data corresponding to the current pixel, and determine the greenhouse identification result of the current pixel based on the MPGI data; wherein the MPGI data includes an initial discrimination parameter value and an MPGI value; until all pixels in the target remote sensing image data are traversed, the target greenhouse identification result of the target area is obtained.

[0125] Furthermore, the greenhouse recognition module 603 is also used to: calculate the initial discrimination parameter value corresponding to the current pixel based on the first reflectivity data corresponding to the current pixel in the target remote sensing image data; wherein the first reflectivity data includes the reflectivity value of the blue band, the reflectivity value of the green band, the reflectivity value of the near infrared band and the reflectivity value of the short-wave infrared 1; determine whether the initial discrimination parameter value corresponding to the current pixel is greater than 0, and obtain a first judgment result; when the first judgment result is no, determine that the current pixel belongs to a non-greenhouse area; when the first judgment result is yes, based on the target remote sensing image data, The second reflectivity data corresponding to the current pixel in the image data is used to calculate the MPGI value corresponding to the current pixel; wherein the second reflectivity data includes the reflectivity value of the coastal aerosol band, the reflectivity value of the blue band, the reflectivity value of the near-infrared band, the reflectivity value of the short-wave infrared 1 and the reflectivity value of the short-wave infrared 2; whether the MPGI value corresponding to the current pixel is greater than or equal to a preset MPGI threshold is determined to obtain a second judgment result; when the second judgment result is yes, it is determined that the current pixel belongs to the greenhouse area; when the second judgment result is no, it is determined that the current pixel belongs to the non-greenhouse area.

[0126] Furthermore, the above device also includes:

[0127] A distribution map generating module is used to generate a greenhouse distribution map of the target area based on the target greenhouse identification result.

[0128] Furthermore, the above device also includes:

[0129] The proportion calculation module is used to calculate the greenhouse proportion value of the target area based on the target greenhouse identification result.

[0130] The implementation principle and technical effects of the device provided in this embodiment are the same as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0131] like Figure 7 As shown, an electronic device 700 provided by an embodiment of the present invention includes: a processor 701, a memory 702 and a bus, the memory 702 stores a computer program that can be run on the processor 701, when the electronic device 700 is running, the processor 701 and the memory 702 communicate through the bus, and the processor 701 executes the computer program to implement the above-mentioned greenhouse identification method.

[0132] Specifically, the memory 702 and the processor 701 can be general-purpose memories and processors, which are not specifically limited here.

[0133] The embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the greenhouse identification method described in the above method embodiment is executed. The computer-readable storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a RAM, a magnetic disk or an optical disk, and other media that can store program codes.

[0134] The term "and / or" herein is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the term "at least one" herein represents any combination of at least two of any one or more of a plurality of. For example, including at least one of A, B, and C can represent including any one or more elements selected from the set consisting of A, B, and C.

[0135] In all examples shown and described herein, any specific values ​​should be interpreted as merely exemplary and not as limiting, and thus other examples of the exemplary embodiments may have different values.

[0136] The flow charts and block diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0137] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0138] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0139] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0140] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A greenhouse identification method, characterized in that: include: Acquire original remote sensing image data of the target area, wherein the original remote sensing image data includes reflectance data of preset bands, wherein the preset bands include coastal aerosol band, blue band, green band, near infrared band, shortwave infrared 1 and shortwave infrared 2; Preprocessing the original remote sensing image data to obtain target remote sensing image data; Based on the preset modified greenhouse index MPGI, greenhouse identification is performed on the target remote sensing image data to obtain the target greenhouse identification result of the target area; wherein the MPGI is defined as follows: v=(R Green -R Blue )×(R NIR -R SWIR1 -0.02); Among them, R coastal is the reflectivity of the coastal aerosol band, R Blue is the reflectivity of the blue band, R Green is the reflectance of the green band, R NIR is the reflectivity in the near-infrared band, R SWIR1 is the reflectivity of short-wave infrared 1, R SWIR2 is the reflectivity of shortwave infrared 2, and v is the preset initial discrimination parameter.

2. The greenhouse identification method according to claim 1, characterized in that: The preprocessing of the original remote sensing image data to obtain target remote sensing image data includes: Performing radiation calibration, atmospheric correction and cropping of the target area on the original remote sensing image data to obtain processed data; The processed data is subjected to removal of typical land object types to obtain target remote sensing image data; wherein the typical land object types include one or more of forests, water bodies, snow and bare soil.

3. The greenhouse identification method according to claim 2, characterized in that: The removing of typical ground object types from the processed data to obtain target remote sensing image data includes: Based on the normalized vegetation index, the improved water body index and the normalized difference bare soil index, the processed data is used to identify the typical land object type to obtain a non-greenhouse identification result; The non-greenhouse area data corresponding to the non-greenhouse identification result is eliminated from the processed data to obtain target remote sensing image data.

4. The greenhouse identification method according to claim 1, characterized in that: The greenhouse identification is performed on the target remote sensing image data based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area, including: Traversing each pixel in the target remote sensing image data; For the traversed current pixel, based on the reflectivity data corresponding to the current pixel in the target remote sensing image data, the MPGI data corresponding to the current pixel is calculated, and the greenhouse recognition result of the current pixel is determined based on the MPGI data; wherein the MPGI data includes an initial discrimination parameter value and an MPGI value; Until all pixels in the target remote sensing image data are traversed, the target greenhouse recognition result of the target area is obtained.

5. The greenhouse identification method according to claim 4, characterized in that: The step of calculating the MPGI data corresponding to the current pixel based on the reflectivity data corresponding to the current pixel in the target remote sensing image data, and determining the greenhouse recognition result of the current pixel based on the MPGI data, includes: Based on the first reflectivity data corresponding to the current pixel in the target remote sensing image data, an initial discrimination parameter value corresponding to the current pixel is calculated; wherein the first reflectivity data includes a reflectivity value of a blue band, a reflectivity value of a green band, a reflectivity value of a near infrared band, and a reflectivity value of a short-wave infrared 1; Determine whether the initial discrimination parameter value corresponding to the current pixel is greater than 0, and obtain a first judgment result; When the first judgment result is no, determining that the current pixel belongs to a non-greenhouse area; When the first judgment result is yes, the MPGI value corresponding to the current pixel is calculated based on the second reflectivity data corresponding to the current pixel in the target remote sensing image data; wherein the second reflectivity data includes the reflectivity value of the coastal aerosol band, the reflectivity value of the blue band, the reflectivity value of the near infrared band, the reflectivity value of the shortwave infrared 1 and the reflectivity value of the shortwave infrared 2; Determine whether the MPGI value corresponding to the current pixel is greater than or equal to a preset MPGI threshold, and obtain a second determination result; When the second judgment result is yes, determining that the current pixel belongs to the greenhouse area; When the second judgment result is no, it is determined that the current pixel belongs to a non-greenhouse area.

6. The greenhouse identification method according to any one of claims 1 to 5, characterized in that: After greenhouse identification is performed on the target remote sensing image data based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area, the greenhouse identification method further includes: Based on the target greenhouse identification result, a greenhouse distribution map of the target area is generated.

7. The greenhouse identification method according to any one of claims 1 to 5, characterized in that: After greenhouse identification is performed on the target remote sensing image data based on the preset modified greenhouse index MPGI to obtain the target greenhouse identification result of the target area, the greenhouse identification method further includes: Based on the target greenhouse identification result, the greenhouse ratio value of the target area is calculated.

8. A greenhouse identification device, characterized in that: include: A data acquisition module is used to acquire original remote sensing image data of a target area, wherein the original remote sensing image data includes reflectance data of preset bands, wherein the preset bands include a coastal aerosol band, a blue band, a green band, a near infrared band, a short-wave infrared 1, and a short-wave infrared 2; A preprocessing module, used for preprocessing the original remote sensing image data to obtain target remote sensing image data; The greenhouse identification module is used to perform greenhouse identification on the target remote sensing image data based on a preset modified greenhouse index MPGI to obtain a target greenhouse identification result of the target area; wherein the MPGI is defined as follows: v=(R Green -R Blue )×(R NIR -R SWIR1 -0.02); Among them, R coastal is the reflectivity of the coastal aerosol band, R Blue is the reflectivity of the blue band, R Green is the reflectance of the green band, R NIR is the reflectivity in the near-infrared band, R SWIR1 is the reflectivity of short-wave infrared 1, R SWIR2 is the reflectivity of shortwave infrared 2, and v is the preset initial discrimination parameter.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the greenhouse identification method described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the greenhouse identification method according to any one of claims 1 to 7 is executed.

Images