Rapid remote sensing monitoring method and device for straw types based on comprehensive CHM characteristics

Abstract

The present invention provides a method and device for rapid remote sensing monitoring of straw types with integrated CHM features, the method comprising: obtaining a multispectral image and a CHM image of a monitoring area; generating a PNISI image according to the near infrared band and the red light band in the multispectral image; performing secondary band combination according to the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image, performing initial segmentation according to a height threshold to obtain a first straw image and a second straw image, and segmenting to obtain multiple first-class objects and second-class objects respectively; classifying the first-class objects according to the straw shape threshold and the PNISI threshold to obtain classification results of bundled, piled and upright straws; classifying the second-class objects based on the band threshold to obtain classification results of broken slag straw and bare land. The method can effectively avoid the problem of straw object fragmentation after conventional segmentation, and improve the accuracy and efficiency of remote sensing monitoring of farmland crop straw types.

Description

Technical Field

[0001] The invention relates to the field of crop remote sensing information processing, and in particular to a method and a device for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics. Background Art

[0002] After the harvest season of crops such as corn, wheat, and soybeans, some of the straw is harvested and processed for reuse, most of the straw is crushed by agricultural machinery and returned to the field and spread on the soil surface or plowed into the soil, some of the straw is harvested manually and gathered into piles or stacks, and some of the straw is not processed and stands upright in the field. Returning straw to the field can achieve protective tillage and soil carbon fixation. Piled straw will be removed from the field for reuse in the later stage, and upright straw is easily burned to pollute the atmosphere. Relevant departments such as agriculture and environmental protection pay great attention to the types and distribution status of crop straw in farmland. Timely, accurate and rapid acquisition of straw type, distribution area and location information can help them achieve efficient management, comprehensive utilization, and burning prevention and control of straw resources, which has great practical significance and application value.

[0003] In recent years, the management and utilization of straw resources and the prevention and control of burning have mainly relied on field surveys, radio publicity and on-site law enforcement by management and law enforcement personnel. The level of informatization is low and the cost is high. The monitoring accuracy and efficiency need to be improved urgently. Low-altitude remote sensing technology has developed rapidly, is flexible and low-cost, and can quickly obtain high-spatial resolution remote sensing images, providing a new data source and technical means for rapid monitoring of straw types. At present, there is a lack of research on remote sensing monitoring of straw types, and straw-related research has mainly focused on straw resource quantity, residue coverage and burning fire point monitoring. Some scholars have also conducted remote sensing monitoring of upright straw, but it is mainly based on conventional remote sensing monitoring parameters such as spectrum and texture, and the accuracy is not high. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a method and device for rapid remote sensing monitoring of straw types with integrated CHM characteristics.

[0005] The present invention provides a rapid remote sensing monitoring method for straw types based on CHM characteristics, comprising: obtaining a multispectral image of a monitoring area and crop canopy height (Canopy Height A multispectral image is prepared according to the present invention; wherein the multispectral image includes spectral information of a near-infrared band, a red band and a blue band; a stacked near-infrared straw index PNISI image is generated according to the spectral information of the near-infrared band and the red band in the multispectral image; a secondary band combination is performed according to the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image; according to a preset straw height threshold, the CHM-PNISI-multispectral image is initially segmented to obtain a first straw image including bales, piles and upright straws, and a second straw image including slag straw and bare land; the first straw image is segmented to obtain a plurality of first-class objects, and the second straw image is segmented to obtain a plurality of second-class objects; according to a preset straw shape threshold and a PNISI threshold, the plurality of first-class objects are classified to obtain classification results of bales, piles and upright straws; based on a preset band threshold, the plurality of second-class objects are classified to obtain classification results of slag straw and bare land.

[0006] According to a rapid remote sensing monitoring method for straw types integrating CHM features provided by the present invention, the shape threshold includes an area threshold and an aspect ratio threshold, the height threshold includes a CHM threshold, and the band threshold includes a blue light band threshold.

[0007] According to a straw type rapid remote sensing monitoring method based on CHM features provided by the present invention, the CHM-PNISI-multispectral image is initially segmented according to a preset straw height threshold to obtain a first straw image including bales, piles and upright straws, and a second straw image including broken straw and bare land, including: according to CHM>0.16m, the CHM-PNISI-multispectral image is segmented to obtain a first straw image including bales, piles and upright straws; according to CHM≤0.16m, the CHM-PNISI-multispectral image is segmented to obtain a second straw image including broken straw and bare land; accordingly, according to a preset straw shape threshold and a PNISI threshold, the multiple first-class objects are classified to obtain classification results of bales, piles and upright straws; based on a preset band threshold, the multiple second-class objects are classified to obtain classification results of broken straw and bare land, respectively including: according to area>0.3332m 2, PNISI>0.177, the first type of object is determined to be a bundled straw type; according to 0.1<PNISI<0.13, aspect ratio γ>2.1, the first type of object is determined to be a piled straw type; according to the first type of objects other than the bundled straw type and the piled straw type, the upright straw type is determined; according to 0.035<B1<0.21, the second type of object is determined to be a bare land type; according to the second type of objects other than the bare land type, the broken straw type is determined. Among them, B1 is the reflectance value of the blue light band.

[0008] According to a rapid remote sensing monitoring method for straw types with integrated CHM features provided by the present invention, the method of obtaining a multispectral image of a monitoring area and a CHM image of crop canopy height comprises: obtaining a high-resolution remote sensing image of the monitoring area, performing preprocessing to obtain the multispectral image and a digital surface model (DSM) image; generating a digital elevation model according to the DSM image and making a digital elevation model (DEM) image; generating a crop canopy height model according to the DEM image through difference calculation and making a CHM image; wherein the preprocessing comprises image stitching, DSM generation, band combination and cropping.

[0009] According to a rapid remote sensing monitoring method for straw types with integrated CHM features provided by the present invention, the stacked near-infrared straw index PNISI image is generated based on the spectral information of the near-infrared band and the red light band in the multispectral image, including generating the PNISI image after calculating the PNISI value according to the following formula:

[0010]

[0011] Among them, B3 is the reflectance value of the red band in the multispectral image, and B5 is the reflectance value of the near-infrared band in the multispectral image.

[0012] According to a rapid remote sensing monitoring method for straw types with integrated CHM features provided by the present invention, the first straw image is segmented to obtain multiple first-class objects, and the second straw image is segmented to obtain multiple second-class objects, including: according to a contrast separation segmentation method, the first straw image is segmented to obtain multiple first-class objects; according to a multi-scale segmentation method, the second straw image is segmented to obtain multiple second-class objects.

[0013] According to a rapid remote sensing monitoring method for straw types integrating CHM features provided by the present invention, after determining different straw types, the method further comprises: generating a distribution map of farmland crop straw types according to different classification results.

[0014] The present invention also provides a rapid remote sensing monitoring device for straw types with integrated CHM features, comprising: a first acquisition module, used to acquire a multispectral image of a monitoring area and a CHM image of crop canopy height, wherein the multispectral image includes spectral information of a near-infrared band, a red light band and a blue light band; a second acquisition module, used to generate a stacked near-infrared straw index PNISI image according to the spectral information of the near-infrared band and the red light band in the multispectral image; a band combination module, used to perform secondary band combination according to the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image; a first segmentation module, used to generate a CHM-PNISI-multispectral image according to a preset The CHM-PNISI-multispectral image is segmented initially based on a straw height threshold to obtain a first straw image including bales, piles and upright straws, and a second straw image including broken straw and bare land; a second segmentation module is used to segment the first straw image to obtain a plurality of first-class objects, and to segment the second straw image to obtain a plurality of second-class objects; a type classification module is used to classify the plurality of first-class objects according to a preset straw shape threshold and a PNISI threshold to obtain classification results of bales, piles and upright straws; and based on a preset band threshold, the plurality of second-class objects are classified to obtain classification results of broken straw and bare land.

[0015] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a rapid remote sensing monitoring method for straw types based on integrated CHM features as described in any one of the above is implemented.

[0016] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-described methods for rapid remote sensing monitoring of straw types based on integrated CHM features.

[0017] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for rapid remote sensing monitoring of straw types based on integrated CHM features as described in any one of the above is implemented.

[0018] The straw type rapid remote sensing monitoring method and device with integrated CHM features provided by the present invention integrates spatial (height, shape) and spectral multi-dimensional information to classify straw types, thereby improving the accuracy and efficiency of remote sensing monitoring of farmland crop straw types. In addition, the straw height threshold is used for initial segmentation, and the two types of images after segmentation are respectively subjected to secondary segmentation using, for example, a contrast separation segmentation algorithm and a multi-scale segmentation algorithm, which can effectively avoid the problem that the conventional single multi-scale segmentation method causes piles and bundles of straw objects to be fragmented and unable to form a complete shape. The method can meet the needs of relevant departments such as agriculture and environmental protection to obtain straw types, distribution areas and location information in a timely, accurate and rapid manner, and can provide information and method support for realizing efficient management, comprehensive utilization, and burning prevention and control of straw resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 This is one of the flow diagrams of the rapid remote sensing monitoring method of straw types based on comprehensive CHM characteristics provided by the present invention;

[0021] Figure 2 This is the second flow chart of the rapid remote sensing monitoring method of straw types based on comprehensive CHM characteristics provided by the present invention;

[0022] Figure 3 It is a structural schematic diagram of a rapid remote sensing monitoring device for straw types with integrated CHM characteristics provided by the present invention;

[0023] Figure 4 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0024] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not 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.

[0025] The height and shape of different types of straw, such as upright straw, piled straw, bundled straw, and slag straw, are more significantly different than the spectrum and texture. Fully exploring the spatial features such as height and shape and combining them with spectral features are very helpful for improving the accuracy of straw monitoring. The present invention uses high spatial resolution images to fully explore the multi-dimensional feature differences of different types of straw, and integrates spatial and spectral features to achieve rapid remote sensing monitoring of straw types. It can provide high-resolution information support such as the spatial distribution type, area, and high-precision location of straw to relevant departments, and assist them in scientific management and prevention and control decisions.

[0026] Combine the following Figure 1-Figure 4 The invention describes a method and device for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics. Figure 1 FIG. 1 is one of the flow diagrams of the method for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics provided by the present invention, such as Figure 1 As shown, the present invention provides a rapid remote sensing monitoring method for straw types based on comprehensive CHM characteristics, comprising:

[0027] 101. Obtain multispectral images of the monitoring area and crop canopy height CHM images, where the multispectral images include spectral information of near-infrared bands, red light bands, and blue light bands.

[0028] For example, a DJI Phantom 4 drone was used to take aerial photos of corn stalks, and multispectral remote sensing images of the monitoring area with five bands, including blue light, green light, red light, red edge, and near infrared, were obtained. Specifically, it was selected to arrive at the monitoring area in clear weather with no wind or light wind, and to select a take-off and landing location without obstacles; according to the actual situation of the monitoring area, the aerial photography height (30m), the lateral overlap of the heading (80%), the flight speed (20m / s), the photo mode (timed shooting), etc. were determined, and the route planning software of the drone was used to set the route planning; to ensure good lighting conditions, the aerial photography mission was performed when the solar altitude angle was high between 10:00 am and 2:00 pm, and multispectral images with high spatial resolution (such as a pixel size of 1.4cm) were obtained.

[0029] Then the images are preprocessed, specifically: first check whether the obtained photos, coordinates and other data are complete; then check the photos one by one, remove the take-off and landing photos and the blurred photos with poor quality during the flight; import all the good quality photos into the PhotoScan professional image processing software; set the parameters such as the input and output coordinate system (WGS84); perform image stitching, DSM generation and other operations to generate stitched single-band images and DSM images. In order to facilitate subsequent processing, the ENVI remote sensing image professional processing software is used to combine the single-band images into multi-band images, and then the multi-band images and DSM images are cropped according to the scope of the monitoring area to obtain the original multi-band images and DSM images of the study area. The DSM images can be used to generate CHM images.

[0030] 102. Generate a stacked near-infrared straw index (PNISI) image based on spectral information of near-infrared bands and red light bands in the multispectral image.

[0031] The stacked near-infrared straw index PNISI image in the present invention integrates the information of the red light and near-infrared bands, and enhances the near-infrared information that is sensitive to straw information, which is helpful for extracting upright straws. By performing band operations on the original multi-band image, the PNISI image of the monitoring area is generated.

[0032] 103. Perform secondary band combination according to the multispectral image, the CHM image, and the PNISI image to obtain a CHM-PNISI-multispectral image.

[0033] The spectral information of the near-infrared band and the red band of the image is used to calculate and generate the stacked near-infrared straw index (Product Near-Infrared Straw Index, PNISI) image, and then the original multi-band image, PNISI image, and CHM image are combined for the second time to generate the altitude-PNISI-multispectral image. In the altitude-PNISI-multispectral image, each pixel has information of multiple dimensions, including altitude, PNISI value, and multiple band characteristics of the spectrum. If the spectral characteristics are 5 bands of blue light, green light, red light, red edge, and near infrared, then each pixel in the altitude-PNISI-multispectral image has information of seven dimensions.

[0034] 104. According to the preset straw height threshold, the CHM-PNISI-multispectral image is initially segmented to obtain a first straw image including bundles, piles and upright straw, and a second straw image including broken slag straw and bare land. Different types of straw have different utilization methods and supervision strategies. Rapid and accurate monitoring of straw types can provide information support for efficient management, comprehensive utilization, and burning prevention and control of straw resources, which has great practical significance and application value. However, there is a lack of monitoring of straw types at present, and related research mainly focuses on on-site straw monitoring, residue coverage monitoring, straw resource monitoring, and fire point monitoring. The present invention studies this practical problem that urgently needs to be solved in scientific research and production, aiming to find a method for rapid and accurate monitoring of straw types.

[0035] Specifically, the strategy of "secondary segmentation and different algorithms" can be adopted. First, according to the height difference of the ground objects in the CHM image, the appropriate height threshold is determined, and the initial segmentation is performed to generate the first straw image of piles and bundles of upright straw and the second straw image of bare ground with broken straw.

[0036] 105. Segment the first straw image to obtain a plurality of first-category objects, and segment the second straw image to obtain a plurality of second-category objects.

[0037] Then, methods such as contrast separation segmentation algorithm and multi-scale segmentation algorithm are used to segment the first straw image and the second straw image respectively to obtain multiple different objects, one of which is bundles, piles or upright straws, and the other is broken straws and bare land.

[0038] 106. Classify the multiple first-class objects according to the preset straw shape threshold and PNISI threshold to obtain classification results of bundled, piled and upright straw; classify the multiple second-class objects based on the preset band threshold to obtain classification results of broken straw and bare land.

[0039] The appropriate threshold is selected and the images of piled and bundled upright straw and the images of broken straw bare land are segmented and the threshold is debugged. Then, the shape parameter images are generated by optimizing and extracting the shape parameters of the images of piled and bundled upright straw. Finally, the high spectral image and the shape parameter image are integrated to construct a classification rule set, and multi-space multispectral image classification and mapping are carried out to realize the rapid remote sensing monitoring of crop straw types in farmland.

[0040] Specifically, shape threshold, PNISI threshold and band threshold can be divided after averaging all pixels of each object.

[0041] Based on the characteristics that the height of upright straw, piled straw and bundled straw is significantly higher than that of broken straw and bare land, and the obvious round and strip-shaped characteristics of piled straw and bundled straw, the present invention fully exploits the spatial characteristics of height, shape and other types of straw, and uses spectral characteristics to perform remote sensing monitoring of crop straw types in farmland. Compared with previous related studies that only used traditional parameters such as spectrum and texture for straw monitoring, the straw monitoring accuracy has been significantly improved.

[0042] The straw type rapid remote sensing monitoring method with integrated CHM features provided in this embodiment integrates spatial (height, shape) and spectral multi-dimensional information to classify straw types, thereby improving the accuracy and efficiency of remote sensing monitoring of farmland crop straw types. In addition, the straw height threshold is used for initial segmentation, and the two types of images after segmentation are respectively subjected to secondary segmentation using, for example, a contrast separation segmentation algorithm and a multi-scale segmentation algorithm, which can effectively avoid the problem that the conventional single multi-scale segmentation method causes the piles and bundles of straw objects to be fragmented and unable to form a complete shape. This method can meet the needs of relevant departments such as agriculture and environmental protection to obtain straw types, distribution areas and location information in a timely, accurate and rapid manner, and can provide information and method support for their efficient management, comprehensive utilization, and burning prevention and control of straw resources.

[0043] In one embodiment, the shape threshold includes an area threshold and an aspect ratio threshold, the height threshold includes a CHM threshold, and the band threshold includes a blue light band threshold.

[0044] Specifically, shape rule sets of height, area and aspect ratio can be set, and then classification can be performed based on the measured threshold. At the same time, the embodiment of the present invention provides a rule set for the blue light band as a spectral orientation to improve the accuracy of the classification results.

[0045] In one embodiment, the CHM-PNISI-multispectral image is initially segmented according to a preset straw height threshold to obtain a first straw image including bales, piles and upright straw, and a second straw image including broken straw and bare ground, including: according to CHM>0.16m, the CHM-PNISI-multispectral image is segmented to obtain a first straw image including bales, piles and upright straw; according to CHM≤0.16m, the CHM-PNISI-multispectral image is segmented to obtain a second straw image including broken straw and bare ground.

[0046] Accordingly, according to the preset straw shape threshold and PNISI threshold, the multiple first-class objects are classified to obtain the classification results of bundled, piled and upright straws. Based on the preset band threshold, the multiple second-class objects are classified to obtain the classification results of broken straws and bare land, which respectively include: according to Aera>0.3332m 2, PNISI>0.177, the first type of object is determined to be the straw bundle type; according to 0.1<PNISI<0.13, the aspect ratio γ>2.1, the first type of object is determined to be the straw pile type; according to the first type of objects other than the straw bundle type and the straw pile type, the upright straw type is determined; according to 0.035<B1<0.21, the second type of object is determined to be the bare land type; according to the second type of objects other than the bare land type, the crushed straw type is determined; where B1 is the reflectance value of the blue light band.

[0047] Specifically, the present invention determines the appropriate thresholds of various parameters and constructs a classification rule set suitable for straw type monitoring by analyzing and debugging multi-space multi-spectral images that integrate spatial (height, shape) and spectral information.

[0048] According to the feature that the height of piled, bundled and upright straw is obviously higher than that of broken straw and bare land, the strategy of "secondary segmentation, different algorithms and different rules" is adopted to classify piled, bundled and upright straw into one category and broken straw and bare land into another category. According to the differences in their spatial characteristics, different segmentation algorithms are used for image segmentation processing to improve the classification accuracy. Therefore, the CHM band in the CHM-PNISI-multispectral image is used to initially segment and generate images of piled and bundled upright straw (CHM>0.16m) and images of broken straw and bare land (CHM≤0.16m).

[0049] After detailed analysis and comparison debugging, it was found that the shape differences of piled, bundled and upright straw were the most prominent, so the contrast separation segmentation algorithm was used to re-segment the images of piled and bundled upright straw. After segmentation, the images can more completely outline the circular characteristics of piled straw, the strip-shaped characteristics of bundled straw and the discrete characteristics of upright straw. Taking into account the image resolution of the monitoring area, the spatial morphological characteristics of the main target objects and the image processing timeliness, the contrast separation segmentation parameters finally selected through debugging are: chessboard size is 1000, step size is 5, minimum dark area is 0.1, maximum dark area is 0.1, and minimum object size is 1.

[0050] The spectral characteristics of bare ground with broken stalks are more prominent, and the shape and height characteristics are similar. Therefore, the bare ground with broken stalks is segmented again using a multi-scale segmentation algorithm based on multi-band information. Taking into account the image resolution of the monitoring area, the spatial morphological characteristics of the main target objects, and the image processing timeliness, the multi-scale segmentation parameters finally selected through debugging are: the segmentation scale parameter is 100, the shape factor parameter is 0.3, and the compactness parameter is 0.5. This operation can be completed in eCognition, a professional high-resolution image processing software.

[0051] The shape of upright straw is scattered and rough, the bundled straw is distributed in strips, and the piled straw is distributed in cones or long cones. According to the differences in straw shape characteristics, through parameter screening, the shape parameters suitable for extracting straw types in the monitoring area are selected - aspect ratio (γ) and area (A).

[0052]

[0053] Among them, γ is the aspect ratio of the segmented object, l is the length value of the segmented object, and w is the width value of the segmented object.

[0054] A=N*Pix 2

[0055] Among them, A is the area of ​​the segmented object, N is the number of pixels of the segmented object, and Pix is ​​the pixel size of a single pixel.

[0056] After multiple debugging, it is determined that the shape parameters of straw type monitoring are the shape information of aspect ratio and area based on CHM image. Through data calculation, aspect ratio and area shape images are generated. PNISI and blue light B1 bands are selected as the spectral parameters of the monitoring area. This operation can be achieved through eCognition software.

[0057] In one embodiment, the method of obtaining a multispectral image of the monitoring area and a crop canopy height CHM image includes: obtaining a high-resolution remote sensing image of the monitoring area, and obtaining the multispectral image and a digital surface model (DSM) image after preprocessing; generating a digital elevation model based on the DSM image and making a digital elevation model (DEM) image; generating a crop canopy height model through difference calculation based on the DEM image and making a CHM image; wherein the preprocessing includes image stitching, DSM generation, band combination and cropping.

[0058] For example, the high spatial resolution multispectral image obtained above is preprocessed to obtain the original multiband image and DSM image. Then, the original multiband image and DSM image are combined to evenly puncture the exposed soil area in the monitoring area to produce a point shp file. By extracting the value to the point, the height information of the puncture point on the DSM image is assigned to the puncture point to generate a height shp file. The Kriging interpolation algorithm is used to interpolate the height shp file to produce a DEM model image of the monitoring area. Generate a DEM image. Then perform a difference operation between the DSM image and the DEM image to produce a crop canopy height model and generate a CHM image of the monitoring area. This operation can be achieved through ArcGIS software.

[0059] In one embodiment, generating a stacked near-infrared straw index (PNISI) image based on spectral information of a near-infrared band and a red band in the multispectral image includes:

[0060]

[0061] Among them, B3 is the reflectance value of the red band in the multispectral remote sensing image, and B5 is the reflectance value of the near-infrared band in the multispectral remote sensing image. This operation can be achieved through ENVI software.

[0062] In one embodiment, segmenting the first straw image to obtain multiple first-class objects and segmenting the second straw image to obtain multiple second-class objects includes: segmenting the first straw image to obtain multiple first-class objects according to a contrast separation segmentation method; segmenting the second straw image to obtain multiple second-class objects according to a multi-scale segmentation method.

[0063] Contrast split segmentation is to segment the whole scene image into dark image objects and bright image objects according to the set contrast threshold. The algorithm evaluates the optimal separation parameters for each object. When selecting the pixel layer as the image object area, chessboard segmentation is initially performed, and then each square object is split.

[0064] Multiresolution Segmentation refers to an image segmentation method that uses a bottom-up region merging technique to continuously merge pixels to generate image objects. The segmentation calculation is modified by changing the scale parameter. The higher the scale parameter, the larger the image object after segmentation, and the smaller the scale parameter, the smaller the image object after segmentation.

[0065] In one embodiment, after determining different straw types, the method further includes: generating a distribution map of crop straw types in the farmland according to different classification results.

[0066] Based on multi-spatial multispectral images, the types of crop straw in farmland are classified according to the straw type monitoring classification rule set of images of upright straw in piles and bundles and images of straw residue and bare land, and then the classification results are merged to generate a distribution map of crop straw types in farmland with five types of objects, namely upright straw, piled straw, bundled straw, straw residue and bare land. The distribution map of crop straw types in farmland can be edited, and the map name, legend, compass, scale, etc. can be added, and the export parameters can be set to produce the remote sensing monitoring thematic map of crop straw type distribution in farmland, which can be completed through ArcGIS software.

[0067] As an embodiment of the present invention, drone aerial photography was carried out in a corn stalk area in a certain city to obtain a multispectral remote sensing image containing 5 bands with a high spatial resolution of centimeters in the monitoring area. Figure 2 As shown in the figure, rapid remote sensing monitoring of corn stalk types was carried out in the monitoring area, and a special map of corn stalk type monitoring in a certain city was generated. The total accuracy of stalk type monitoring is 93%, which can meet the needs of relevant departments such as agriculture and environmental protection to obtain stalk type, distribution area and location information in a timely, accurate and rapid manner. It can provide information and method support for the efficient management, comprehensive utilization and burning prevention and control of stalk resources, which has great practical significance and application value.

[0068] The rapid remote sensing monitoring device for straw types based on CHM characteristics provided by the present invention is described below. The rapid remote sensing monitoring device for straw types based on CHM characteristics described below and the rapid remote sensing monitoring method for straw types based on CHM characteristics described above can be referenced to each other.

[0069] Figure 3 : is a schematic diagram of the structure of the straw type rapid remote sensing monitoring device with comprehensive CHM characteristics provided by the present invention, such as Figure 3 As shown, the rapid remote sensing monitoring device for straw types with integrated CHM features includes: a first acquisition module 301, a second acquisition module 302, a band combination module 303, a first segmentation module 304, a second segmentation module 305 and a type classification module 306. Among them, the first acquisition module 301 is used to acquire the multispectral image of the monitoring area and the crop canopy height CHM image, wherein the multispectral image includes spectral information of the near-infrared band, the red band and the blue band; the second acquisition module 302 is used to generate a stacked near-infrared straw index PNISI image based on the spectral information of the near-infrared band and the red band in the multispectral image; the band combination module 303 is used to perform secondary band combination based on the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image; the first segmentation module 304 is used to classify the CHM according to a preset straw height threshold. The M-PNISI-multispectral image is segmented initially to obtain a first straw image including bales, piles and upright straws, and a second straw image including broken straw and bare land; the second segmentation module 305 is used to segment the first straw image to obtain a plurality of first-class objects, and segment the second straw image to obtain a plurality of second-class objects; the type classification module 306 is used to classify the plurality of first-class objects according to a preset straw shape threshold and a PNISI threshold, and obtain classification results of bales, piles and upright straws; based on a preset band threshold, the plurality of second-class objects are classified to obtain classification results of broken straw and bare land.

[0070] The device embodiments provided in the embodiments of the present invention are for implementing the above-mentioned method embodiments. For specific processes and detailed contents, please refer to the above-mentioned method embodiments, which will not be repeated here.

[0071] The implementation principle and technical effects of the rapid remote sensing monitoring device for straw types based on comprehensive CHM characteristics provided in the embodiment of the present invention are the same as those of the aforementioned embodiment of the rapid remote sensing monitoring method for straw types based on comprehensive CHM characteristics. For the sake of brief description, for matters not mentioned in the embodiment of the rapid remote sensing monitoring device for straw types, reference may be made to the corresponding contents in the aforementioned embodiment of the rapid remote sensing monitoring method for straw types based on comprehensive CHM characteristics.

[0072] Figure 4 is a schematic diagram of the structure of the electronic device provided by the present invention, such as Figure 4 As shown, the electronic device may include: a processor 401, a communication interface 402, a memory 403 and a communication bus 404, wherein the processor 401, the communication interface 402 and the memory 403 communicate with each other through the communication bus 404. The processor 401 may call the logic instructions in the memory 403 to execute a rapid remote sensing monitoring method for straw types with integrated CHM features, the method comprising: obtaining a multispectral image and a CHM image of the monitoring area, wherein the multispectral image includes spectral information of a near-infrared band, a red band and a blue band; generating a stacked near-infrared straw index PNISI image based on the spectral information of the near-infrared band and the red band in the multispectral image; performing secondary band combination based on the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image; and obtaining a straw index PNISI-multispectral image based on a preset straw index. The CHM-PNISI-multispectral image is segmented initially based on a straw height threshold to obtain a first straw image including bales, piles and upright straws, and a second straw image including broken straws and bare land; the first straw image is segmented to obtain a plurality of first-class objects, and the second straw image is segmented to obtain a plurality of second-class objects; the plurality of first-class objects are classified according to a preset straw shape threshold and a PNISI threshold to obtain classification results of bales, piles and upright straws; the plurality of second-class objects are classified based on a preset band threshold to obtain classification results of broken straws and bare land.

[0073] In addition, the logic instructions in the above-mentioned memory 403 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0074] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the straw type rapid remote sensing monitoring method with integrated CHM characteristics provided by the above methods. The method includes: acquiring a multispectral image and a CHM image of the monitoring area, wherein the multispectral image includes spectral information of a near-infrared band, a red light band, and a blue light band; generating a stacked near-infrared straw index PNISI image based on the spectral information of the near-infrared band and the red light band in the multispectral image; and generating a near-infrared straw index PNISI image based on the multispectral image, the CHM image, and the PNISI image. A secondary band combination is performed to obtain a CHM-PNISI-multispectral image; according to a preset straw height threshold, the CHM-PNISI-multispectral image is initially segmented to obtain a first straw image including bales, piles and upright straws, and a second straw image including broken straw and bare land; the first straw image is segmented to obtain a plurality of first-class objects, and the second straw image is segmented to obtain a plurality of second-class objects; according to a preset straw shape threshold and a PNISI threshold, the plurality of first-class objects are classified to obtain classification results of bales, piles and upright straws; based on a preset band threshold, the plurality of second-class objects are classified to obtain classification results of broken straw and bare land.

[0075] In another aspect, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the method for rapid remote sensing monitoring of straw types based on the comprehensive CHM features provided by the above methods is implemented, the method comprising: obtaining a multispectral image and a CHM image of a monitoring area, wherein the multispectral image includes spectral information of a near-infrared band, a red light band, and a blue light band; generating a stacked near-infrared straw index PNISI image based on the spectral information of the near-infrared band and the red light band in the multispectral image; performing secondary band combination based on the multispectral image, the CHM image, and the PNISI image to obtain a CHM -PNISI-multispectral image; according to a preset straw height threshold, the CHM-PNISI-multispectral image is initially segmented to obtain a first straw image including bales, piles and upright straws, and a second straw image including broken straw and bare land; the first straw image is segmented to obtain a plurality of first-class objects, and the second straw image is segmented to obtain a plurality of second-class objects; according to a preset straw shape threshold and a PNISI threshold, the plurality of first-class objects are classified to obtain classification results of bales, piles and upright straws; based on a preset band threshold, the plurality of second-class objects are classified to obtain classification results of broken straw and bare land.

[0076] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0077] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0078] 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 make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A rapid remote sensing monitoring method for straw types based on CHM characteristics, characterized in that: include: Obtain multispectral images of the monitoring area and crop canopy height CHM images, where the multispectral images include spectral information of near-infrared bands, red bands, and blue bands; Generate a stacked near-infrared straw index (PNISI) image based on spectral information of near-infrared bands and red light bands in the multispectral image; Perform secondary band combination according to the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image; According to a preset straw height threshold, the CHM-PNISI-multispectral image is initially segmented to obtain a first straw image including bundles, piles and upright straws, and a second straw image including broken straws and bare land; Segmenting the first straw image to obtain a plurality of first-category objects, and segmenting the second straw image to obtain a plurality of second-category objects; Classifying the plurality of first-class objects according to a preset straw shape threshold and a PNISI threshold to obtain classification results of bundled, piled and upright straw; Based on the preset band threshold, the multiple second-category objects are classified to obtain classification results of broken slag straw and bare land.

2. The rapid remote sensing monitoring method of straw types based on comprehensive CHM characteristics according to claim 1 is characterized in that: The shape threshold includes an area threshold and an aspect ratio threshold, the height threshold includes a CHM threshold, and the band threshold includes a blue light band threshold.

3. The rapid remote sensing monitoring method of straw types based on comprehensive CHM characteristics according to claim 2 is characterized in that: The CHM-PNISI-multispectral image is segmented initially according to a preset straw height threshold to obtain a first straw image including bundles, piles and upright straws, and a second straw image including broken straws and bare land, including: According to CHM>0.16, the CHM-PNISI-multispectral image is segmented to obtain a first straw image including bundles, piles and upright straws; According to CHM≤0.16m, the CHM-PNISI-multispectral image is segmented to obtain a second straw image including broken straw and bare land; Accordingly, according to the preset straw shape threshold and PNISI threshold, the multiple first-class objects are classified to obtain the classification results of bundled, piled and upright straws. Based on the preset band threshold, the multiple second-class objects are classified to obtain the classification results of broken straw and bare land, which respectively include: According to area>0.3332m 2 , PNISI>0.177, it is determined that one type of object is the straw bale type; According to 0.1<PNISI<0.13 and aspect ratio γ>2.1, the first type of objects is determined to be piled straw type; According to a class of objects other than the bale straw type and the pile straw type, the upright straw type is determined; According to 0.035<B1<0.21, the second type of object is determined to be a bare land type; According to the second type of objects other than the bare land type, the type of slag straw is determined; Among them, B1 is the reflectivity value of the blue light band.

4. The method for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics according to claim 1, characterized in that: The method of obtaining the multispectral image of the monitoring area and the crop canopy height CHM image includes: Acquire high-resolution remote sensing images of the monitoring area, and obtain the multispectral images and digital surface model (DSM) images after preprocessing; Generate a digital elevation model and produce a DEM image based on the DSM image; Based on the DEM image, a crop canopy height model is generated through difference calculation and a CHM image is produced; The preprocessing includes image stitching, DSM generation, band combination and cropping.

5. The method for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics according to claim 1, characterized in that: The generating of the stacked near-infrared straw index (PNISI) image according to the spectral information of the near-infrared band and the red band in the multispectral image comprises generating the PNISI image after calculating the PNISI value according to the following formula: Among them, B3 is the reflectance value of the red band in the multispectral image, and B5 is the reflectance value of the near-infrared band in the multispectral image.

6. The method for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics according to claim 1, characterized in that: The step of segmenting the first straw image to obtain a plurality of first-category objects and segmenting the second straw image to obtain a plurality of second-category objects includes: According to the contrast separation segmentation method, the first straw image is segmented to obtain a plurality of first-class objects; According to the multi-scale segmentation method, the second straw image is segmented to obtain a plurality of second-category objects.

7. The method for rapid remote sensing monitoring of straw types based on comprehensive CHM characteristics according to claim 1, characterized in that: After obtaining the classification results of the residue straw and bare land, the method further includes: Based on different classification results, a distribution map of farmland crop straw types is generated.

8. A rapid remote sensing monitoring device for straw types based on CHM characteristics, characterized in that: include: The first acquisition module is used to acquire multispectral images of the monitoring area and crop canopy height CHM images, wherein the multispectral images include spectral information of near infrared band, red band and blue band; The second acquisition module is used to generate a stacked near-infrared straw index PNISI image based on the spectral information of the near-infrared band and the red band in the multispectral image; A band combination module, used for performing secondary band combination according to the multispectral image, the CHM image and the PNISI image to obtain a CHM-PNISI-multispectral image; A first segmentation module is used to perform initial segmentation on the CHM-PNISI-multispectral image according to a preset straw height threshold, to obtain a first straw image including bundles, piles and upright straws, and a second straw image including broken straws and bare land; A second segmentation module is used to segment the first straw image to obtain a plurality of first-category objects, and segment the second straw image to obtain a plurality of second-category objects; A type classification module is used to classify the plurality of first-class objects according to a preset straw shape threshold and a PNISI threshold, and obtain classification results of bundled, piled and upright straw; Based on the preset band threshold, the multiple second-category objects are classified to obtain classification results of broken slag straw and bare land.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the method for rapid remote sensing monitoring of straw types based on integrated CHM features as described in any one of claims 1 to 7 is implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for rapid remote sensing monitoring of straw types based on integrated CHM features as claimed in any one of claims 1 to 7 is implemented.

Images