CN112966548B - Soybean plot identification method and system - Google Patents
Soybean plot identification method and system Download PDFInfo
- Publication number
- CN112966548B CN112966548B CN202110060578.6A CN202110060578A CN112966548B CN 112966548 B CN112966548 B CN 112966548B CN 202110060578 A CN202110060578 A CN 202110060578A CN 112966548 B CN112966548 B CN 112966548B
- Authority
- CN
- China
- Prior art keywords
- soybean
- block
- identification
- data
- land
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 244000068988 Glycine max Species 0.000 title claims abstract description 182
- 235000010469 Glycine max Nutrition 0.000 title claims abstract description 182
- 238000000034 method Methods 0.000 title claims abstract description 56
- 238000005070 sampling Methods 0.000 claims abstract description 46
- 238000012549 training Methods 0.000 claims abstract description 37
- 230000011218 segmentation Effects 0.000 claims description 60
- 238000004458 analytical method Methods 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 8
- 238000003860 storage Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000013179 statistical model Methods 0.000 claims description 7
- 238000012795 verification Methods 0.000 claims description 6
- 238000013136 deep learning model Methods 0.000 claims description 3
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 238000011160 research Methods 0.000 description 8
- 240000008042 Zea mays Species 0.000 description 5
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 5
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 235000005822 corn Nutrition 0.000 description 5
- 238000010586 diagram Methods 0.000 description 5
- 238000011835 investigation Methods 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/188—Vegetation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Abstract
The invention provides a soybean plot identification method and a soybean plot identification system, wherein the method comprises the following steps: acquiring a remote sensing image picture of a region to be identified; inputting the remote sensing image picture into a recognition model, and outputting a soybean land block recognition result of the region; the identification model is obtained after training based on a sample image of the soybean sampling land block and corresponding identification labels, and the identification labels are predetermined according to the soybean sampling land block and correspond to the sample image one by one. The method realizes the high identification precision and the high accuracy of space positioning of the soybean planting land block based on satellite remote sensing data, and meets the actual application requirements.
Description
Technical Field
The invention relates to the technical field of agricultural image recognition, in particular to a soybean plot recognition method and system.
Background
The innovative regional income insurance which is mainly developed in China in recent years is introduced into satellite remote sensing technology to assist in the verification, investigation, damage assessment and claim settlement of agricultural insurance, wherein the verification of the marked crop insurance coverage is an important ring in the implementation process of agricultural insurance, and the extraction of crops in the current application is concentrated on pixel-level identification, and is not combined with land data, so that the estimation precision of the crop planting coverage is not high, the accuracy of space positioning is poor, inconvenience is brought to the field investigation work, and the application requirements of the verification, damage settlement and damage assessment work of insurance companies cannot be met.
Disclosure of Invention
The invention provides a soybean plot identification method and a soybean plot identification system, which are used for solving the problem that the identification precision and the space positioning of soybean planting plots obtained based on satellite remote sensing data in the prior art cannot meet the application requirements, and realizing higher identification precision and space positioning accuracy.
The invention provides a soybean plot identification method, which comprises the following steps:
acquiring a remote sensing image picture of a region to be identified;
inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified;
the identification model is obtained after training based on a sample image of the soybean sampling land block and corresponding identification labels, and the identification labels are predetermined according to the soybean sampling land block and correspond to the sample image one by one.
According to the soybean plot recognition method provided by the invention, the recognition model comprises a plot segmentation model, a soybean recognition model and a space statistics model;
inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified, wherein the remote sensing image picture comprises the following steps:
inputting the remote sensing image picture into the land block segmentation model, and outputting a segmented land block image and a segmented land block vector layer in the whole area;
inputting the segmented land block images into the soybean recognition model, and outputting crop class data layers of all pixel points on the land block in the whole area;
and inputting the segmented land block vector image layer in the whole area and the crop category data image layer of each pixel point on the land block in the whole area into the space statistical model for superposition analysis, and judging the land block and the statistical soybean planting area according to the superposition analysis result.
According to the soybean plot identification method provided by the invention, the partitioned plot vector image layer in the whole area and the pixel point crop category data image layer on the plots in the whole area are input into the space statistical model for superposition analysis, and the method comprises the following steps:
and carrying out one-to-one superposition on the map layer data of the corresponding coordinate points on the map vector layer of the block after the division in the whole area and the crop category data map layer of each pixel point on the block in the whole area according to the geographic coordinate information.
According to the soybean block identification method provided by the invention, the sample image of the soybean sampling block comprises a remote sensing image picture, radar image data and soybean sampling block vector data;
the identification model is obtained after training based on a sample image of a soybean sampling land block and a corresponding identification label, and comprises the following steps:
selecting a remote sensing image picture of a partial region to perform label processing and then training the land parcel segmentation model as a training sample image;
inputting the remote sensing image picture of the whole area as a verification sample image into the trained land block segmentation model to obtain segmented land block data;
training the soybean recognition model by using the radar image data, the soybean sampling block vector data and the segmented block data;
the soybean sampling block vector data is obtained through in-field sampling.
According to the soybean plot identification method provided by the invention, the remote sensing image picture of the selected partial region is subjected to label processing and then is used as a training sample image to train the plot segmentation model, and the method comprises the following steps:
selecting a target remote sensing image picture from the remote sensing image pictures of the whole area, editing and manufacturing an identification tag of a land block sample and a land block segmentation remote sensing image, and creating an identification tag data set according to the identification tag of the land block sample; the remote sensing images of the land block segmentation are in one-to-one correspondence with the identification tags of the land block samples;
and inputting the identification tag of the block sample and the corresponding block segmentation remote sensing image into a block segmentation model to realize semantic segmentation training of the block segmentation model.
According to the soybean block recognition method provided by the invention, the soybean recognition model is trained by utilizing radar image data, soybean sampling block vector data and the segmented block data, and the soybean recognition model comprises the following steps:
rasterizing the soybean sampling block vector data and the segmented block data according to the resolution of the radar image data to obtain a raster data set;
extracting time sequence radar data corresponding to each pixel point from each raster data in the raster data sets to obtain a soybean identification sample radar data set and a radar data set of each pixel point on a land block; the time sequence radar data corresponding to each pixel point comprises geographic coordinate information of a radar for spatial superposition analysis;
and training the time sequence deep learning model by using the soybean recognition sample radar data set to obtain the soybean recognition model, and recognizing and predicting the radar data set of each pixel point on the land.
According to the soybean plot identification method provided by the invention, the segmented plot vector layer in the whole region is obtained by mapping the segmented plot data with the remote sensing image picture of the whole region according to pixels based on the plot segmentation model and adding geographic coordinate information;
and the crop category data layer of each pixel point on the land in the whole area is obtained by identifying and predicting a radar data set of each pixel point on the land based on the soybean identification model.
The invention also provides a soybean plot recognition system, which comprises a data input module and a recognition module;
the data input module is used for acquiring remote sensing image pictures of the area to be identified;
the identification module is used for inputting the remote sensing image picture and outputting the identification result of the soybean land block of the area to be identified;
the identification module is obtained after training based on a sample image of the soybean sampling land block and corresponding identification tags, and the identification tags are predetermined according to the soybean sampling land block and correspond to the sample image one by one.
The invention also provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the soybean plot identification method as described above when executing the program.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the steps of the soybean plot identification method as described in any one of the above.
According to the soybean plot identification method and system provided by the invention, the problem that the existing identification of soybean planting plots obtained based on satellite remote sensing data cannot meet application requirements in precision and space positioning can be effectively solved through the plot segmentation method based on high-resolution remote sensing images and the soybean identification method based on time sequence radar data.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a soybean plot identification method provided by the invention;
FIG. 2 is a second flow chart of the soybean block identification method according to the present invention;
FIG. 3 is a third flow chart of the soybean block identification method according to the present invention;
FIG. 4 is a schematic diagram of a plot segmentation tag dataset provided by the present invention;
FIG. 5 is a schematic diagram of a remote sensing image dataset for block segmentation provided by the invention;
FIG. 6 is a partial sample view of a remote sensing image provided by the present invention;
FIG. 7 is a partial sample view of a remote sensing image picture matching geographic coordinate information provided by the invention;
FIG. 8 is a schematic diagram of a soybean block identification system provided by the invention;
FIG. 9 is a schematic view of a crop category planting area provided by the present invention;
fig. 10 is a schematic structural diagram of an electronic device provided by the present invention;
reference numerals:
1: a soybean planting area; 2: a soybean planting area; 3: corn planting area;
4: corn planting area; 5: corn planting area.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the present invention more apparent, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is apparent that the described embodiments are some embodiments of the present invention, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The following describes a soybean plot recognition method provided by the invention with reference to fig. 1 to 7.
The invention provides a soybean plot identification method, as shown in figure 1, comprising the following steps:
s1, acquiring a remote sensing image picture of an area to be identified;
specifically, the remote sensing image picture of the soybean land block of the area to be identified is obtained from high-quality and high-resolution remote sensing image data in the soybean growing period.
S2, inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified;
the identification model is obtained after training based on a sample image of the soybean sampling land block and corresponding identification labels, and the identification labels are predetermined according to the soybean sampling land block and correspond to the sample image one by one.
Compared with the prior art, the method and the device for identifying the soybean plots based on the recognition model after deep learning have the advantages that the soybean plots are judged and the soybean planting areas are counted, and the statistical accuracy of the soybean plot planting areas and the accuracy of plot space positioning are effectively improved.
According to the soybean plot recognition method provided by the invention, the recognition model comprises a plot segmentation model, a soybean recognition model and a space statistics model;
as shown in fig. 2, inputting the remote sensing image picture into the recognition model, and outputting the recognition result of the soybean land block of the region to be recognized, including:
s21, inputting the remote sensing image picture into the land block segmentation model, and outputting a segmented land block image and a segmented land block vector layer in the whole area to the land block vector layer;
s22, inputting the segmented land block images into the soybean recognition model, and outputting crop class data layers of all pixel points on the land block in the whole area;
s23, inputting the segmented land block vector image layer in the whole area and the crop category data image layer of each pixel point on the land block in the whole area into the space statistical model for superposition analysis, and judging the land block and the statistical soybean planting area according to the superposition analysis result.
Specifically, the identification model comprises a plot segmentation model, a soybean identification model and a space statistics model; and sequentially adopting a block segmentation method and a soybean recognition method for the input remote sensing image picture of the soybean block of the region to be recognized, and carrying out space superposition on the obtained block vector image layer after segmentation in the whole region and the crop class data image layer of each pixel point on the block in the whole region.
According to the soybean plot identification method provided by the invention, the partitioned plot vector image layer in the whole area and the pixel point crop category data image layer on the plots in the whole area are input into the space statistical model for superposition analysis, and the method comprises the following steps:
and carrying out one-to-one superposition on the map layer data of the corresponding coordinate points on the map vector layer of the block after the division in the whole area and the crop category data map layer of each pixel point on the block in the whole area according to the geographic coordinate information.
Specifically, superposition analysis of the block vector image layer after the division in the whole area and the crop class data image layer of each pixel point on the block in the whole area is realized by adopting a space analysis module (spatial analyst tools) in arcgis software, and the block is judged to be the soybean block by setting the soybean pixel point occupation ratio to be more than 80 percent in the application.
According to the soybean block identification method provided by the invention, the sample image of the soybean sampling block comprises a remote sensing image picture, radar image data and soybean sampling block vector data;
specifically, in the early data preparation stage, firstly, field investigation is carried out to obtain soybean planting land block samples in a research area, namely soybean sampling land block vector data, and secondly, high-quality high-resolution remote sensing image data in a soybean growing period and time sequence radar image data in the soybean growing period are obtained.
As shown in fig. 3, the recognition model is obtained after training based on a sample image of a soybean sampling land block and a corresponding recognition tag, and includes:
s201, selecting a remote sensing image picture of a partial region for label processing, and then training the land parcel segmentation model as a training sample image;
s202, inputting a remote sensing image picture of the whole area as a verification sample image into the trained land block segmentation model to obtain segmented land block data;
s203, training the soybean recognition model by using the radar image data, the soybean sampling block vector data and the segmented block data;
the soybean sampling block vector data is obtained through in-field sampling.
According to the soybean plot identification method provided by the invention, the remote sensing image picture of the selected partial region is subjected to label processing and then is used as a training sample image to train the plot segmentation model, and the method comprises the following steps:
selecting a target remote sensing image picture from the remote sensing image pictures of the whole area, editing and manufacturing an identification tag of a land block sample and a land block segmentation remote sensing image, and creating an identification tag data set according to the identification tag of the land block sample; the remote sensing images of the land block segmentation are in one-to-one correspondence with the identification tags of the land block samples;
specifically, the label processing mode is to manually and visually interpret and delineate the land parcel in part of the research area by combining the field investigation data, and the land parcel dividing label data set shown in fig. 4 and the corresponding land parcel dividing remote sensing image data set shown in fig. 5 are manufactured.
And inputting the identification tag of the block sample and the corresponding block segmentation remote sensing image into a block segmentation model to realize semantic segmentation training of the block segmentation model.
Specifically, implementing semantic segmentation training on the parcel segmentation model includes: the plot label and the corresponding remote sensing image are input to be trained by adopting a plot segmentation model of deeplab v < 3+ >, and the mIoU of the model in the application of the embodiment reaches 89.8, and the OA reaches 94.8.
According to the soybean block recognition method provided by the invention, the soybean recognition model is trained by utilizing radar image data, soybean sampling block vector data and the segmented block data, and the soybean recognition model comprises the following steps:
rasterizing the soybean sampling block vector data and the segmented block data according to the resolution of the radar image data to obtain a raster data set;
extracting time sequence radar data corresponding to each pixel point from each raster data in the raster data sets to obtain a soybean identification sample radar data set and a radar data set of each pixel point on a land block; the time sequence radar data corresponding to each pixel point comprises geographic coordinate information of a radar for spatial superposition analysis;
specifically, preparation of a crop identification sample is performed before a soybean identification model is trained, a soybean block and a vector block after segmentation are resampled according to resolution (for example, ten meters) of radar image data, time sequence radar data corresponding to each grid pixel point are extracted, and a crop identification sample data set is manufactured, wherein the crop identification sample data set comprises a soybean identification sample radar data set and radar data sets of all pixel points on the block.
And training the time sequence deep learning model by using the soybean recognition sample radar data set to obtain the soybean recognition model, and recognizing and predicting the radar data set of each pixel point on the land.
Specifically, the time series deep learning module is based on a long-short-term memory network LSTM; and training a soybean recognition model based on LSTM, predicting pixel points on the land block in the research area, and outputting prediction results (soybean, corn or other) of each pixel point on the land block.
According to the soybean block identification method provided by the invention, the block vector layer after the block is segmented in the whole region is obtained by mapping the segmented block data with the remote sensing image picture of the whole region based on the block segmentation model, and part of samples of the block vector layer are shown in fig. 6 and are obtained by adding geographic coordinate information according to pixels;
specifically, based on the deeplab v3+ training block segmentation model, the whole research area is subjected to block segmentation and is matched with geographic coordinate information, and part of samples of the block segmentation model are shown in fig. 7.
And the crop category data layer of each pixel point on the land in the whole area is obtained by identifying and predicting a radar data set of each pixel point on the land based on the soybean identification model.
The following describes a soybean block recognition system provided by the present invention, and the following description and the above-described soybean block recognition method can be referred to correspondingly.
The invention provides a soybean plot identification system 800, as shown in fig. 8, comprising a data input module 810 and an identification module 820;
the data input module 810 is configured to obtain a remote sensing image picture of the area to be identified;
the recognition module 820 is configured to input the remote sensing image picture, and output a recognition result of the soybean land block of the region to be recognized;
the recognition module 820 is obtained after training based on a sample image of the soybean sampling land and a corresponding recognition tag, and the recognition tag is predetermined according to the soybean sampling sample and corresponds to the sample image one by one.
Specifically, while training the recognition module 820, the data input module 810 receives sample images of a soybean sample plot including a field sampled derived research area, remote sensing image data and radar image data of the research area; the recognition modules comprise a land parcel segmentation module 821, a soybean recognition module 822 and a space statistics analysis module 823; wherein the parcel segmentation module 821 comprises a parcel segmentation model and the soybean recognition module 822 comprises a soybean recognition model; transmitting the remote sensing image data to the block segmentation module 821, simultaneously transmitting the radar image data and the soybean sampling block vector data to the soybean recognition module 822, and simultaneously receiving the segmented block data output by the block segmentation module 821 by the soybean recognition module 822; the space statistics analysis module 823 receives the segmented block vector image layer in the whole area output by the block segmentation module 821, and simultaneously receives the crop class data image layer of each pixel point on the block in the whole area output by the soybean recognition module 822.
The block segmentation module 821 is configured to train a block segmentation model by using the remote sensing image picture to obtain segmented block data, and obtain a segmented block vector layer in the whole area according to the segmented block data;
the soybean recognition module 822 is configured to train a soybean recognition model according to the radar image data, the soybean sampling block vector data and the segmented block data, so as to obtain a crop class data layer of each pixel point on the block in the whole area;
the space statistics analysis module 823 is configured to perform superposition analysis on the block vector layer after the division in the whole area and the pixel point crop category data layer on the block in the whole area, and determine the soybean block and count the soybean planting area according to the result of superposition analysis.
For example, as shown in the schematic view of the planting area of the crop category shown in fig. 9, the above-mentioned divided land images and the soybean pixel data are subjected to spatial superposition analysis, the area where the soybean pixel is highly concentrated and is an independent divided land is determined as the soybean planting land, and the soybean planting area is counted. The pixel points of the land block 1 and the land block 2 which are identified as soybeans are highly concentrated, and the land block is determined to be a soybean planting area; and similarly, plots 3, 4 and 5 are corn planting areas.
Compared with the prior art, the method and the device fully utilize the field sampling to obtain the soybean sampling land block vector data of the research area, and the obtained remote sensing image data and radar image data of the research area, and judge the soybean land block and count the soybean planting area through the trained land block segmentation module and the trained soybean identification module in sequence, so that the statistical accuracy of the soybean land block planting area and the problem of land block space positioning are effectively improved.
Fig. 10 illustrates a physical structure diagram of an electronic device, as shown in fig. 10, which may include: a processor 1010, a communication interface (Communications Interface) 1020, a memory 1030, and a communication bus 1040, wherein the processor 1010, the communication interface 1020, and the memory 1030 communicate with each other via the communication bus 1040. Processor 1010 may invoke logic instructions in memory 1030 to perform a soybean plot identification method comprising: acquiring a remote sensing image picture of a soybean land block of an area to be identified; inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified; the identification model is obtained after training based on a sample image of the soybean sampling land block and corresponding identification labels, and the identification labels are predetermined according to the soybean sampling land block and correspond to the sample image one by one.
Further, the logic instructions in the memory 1030 described above may be implemented in the form of software functional units and stored in a computer readable storage medium when sold or used as a stand alone product. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
In another aspect, the present invention also provides a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions which, when executed by a computer, enable the computer to perform the method of identifying soybean plots provided by the methods described above, the method comprising: acquiring a remote sensing image picture of a region to be identified; inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified; the identification model is obtained after training based on a sample image of the soybean sampling land block and corresponding identification labels, and the identification labels are predetermined according to the soybean sampling land block and correspond to the sample image one by one.
In yet another aspect, the present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, is implemented to perform the above-provided soybean plot identification methods, the method comprising: acquiring a remote sensing image picture of a region to be identified; inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified; the identification model is obtained after training based on a sample image of the soybean sampling land block and corresponding identification labels, and the identification labels are predetermined according to the soybean sampling land block and correspond to the sample image one by one.
The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.
From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solution of the present invention, and are not limiting; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims (9)
1. A soybean plot identification method, comprising:
acquiring a remote sensing image picture of a region to be identified;
inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified;
the identification model is obtained after training based on a sample image of a soybean sampling land block and a corresponding identification label, and the identification label is predetermined according to the soybean sampling land block and corresponds to the sample image one by one;
the identification model comprises a land parcel segmentation model, a soybean identification model and a space statistics model;
inputting the remote sensing image picture into an identification model, and outputting an identification result of the soybean land block of the area to be identified, wherein the remote sensing image picture comprises the following steps:
inputting the remote sensing image picture into the land block segmentation model, and outputting a segmented land block image and a segmented land block vector layer in the whole area;
inputting the segmented land block images into the soybean recognition model, and outputting crop class data layers of all pixel points on the land block in the whole area;
and inputting the segmented land block vector image layer in the whole area and the crop category data image layer of each pixel point on the land block in the whole area into the space statistical model for superposition analysis, and judging the land block and the statistical soybean planting area according to the superposition analysis result.
2. The soybean plot identification method of claim 1, wherein inputting the segmented plot vector layer in the whole area and the plot class data layer of each pixel point on the plot in the whole area into the spatial statistical model for superposition analysis comprises:
and carrying out one-to-one superposition on the map layer data of the corresponding coordinate points on the map vector layer of the block after the division in the whole area and the crop category data map layer of each pixel point on the block in the whole area according to the geographic coordinate information.
3. The soybean block identification method of claim 1, wherein the sample image of the soybean sampled block comprises a remote sensing image picture, radar image data, and soybean sampled block vector data;
the identification model is obtained after training based on a sample image of a soybean sampling land block and a corresponding identification label, and comprises the following steps:
selecting a remote sensing image picture of a partial region to perform label processing and then training the land parcel segmentation model as a training sample image;
inputting the remote sensing image picture of the whole area as a verification sample image into the trained land block segmentation model to obtain segmented land block data;
training the soybean recognition model by using the radar image data, the soybean sampling block vector data and the segmented block data;
the soybean sampling block vector data is obtained through in-field sampling.
4. The method for identifying soybean plots according to claim 3, wherein the training of the plot segmentation model as the training sample image after the remote sensing image picture of the selected partial region is subjected to the label processing comprises:
selecting a target remote sensing image picture from the remote sensing image pictures of the whole area, editing and manufacturing an identification tag of a land block sample and a land block segmentation remote sensing image, and creating an identification tag data set according to the identification tag of the land block sample; the remote sensing images of the land block segmentation are in one-to-one correspondence with the identification tags of the land block samples;
and inputting the identification tag of the block sample and the corresponding block segmentation remote sensing image into a block segmentation model to realize semantic segmentation training of the block segmentation model.
5. A soybean plot recognition method according to claim 3, wherein training the soybean recognition model using the radar image data, soybean sampled plot vector data, and the segmented plot data comprises:
rasterizing the soybean sampling block vector data and the segmented block data according to the resolution of the radar image data to obtain a raster data set;
extracting time sequence radar data corresponding to each pixel point from each raster data in the raster data sets to obtain a soybean identification sample radar data set and a radar data set of each pixel point on a land block; the time sequence radar data corresponding to each pixel point comprises geographic coordinate information of a radar for spatial superposition analysis;
and training the time sequence deep learning model by using the soybean recognition sample radar data set to obtain the soybean recognition model, and recognizing and predicting the radar data set of each pixel point on the land.
6. The soybean block identification method according to claim 5, wherein the block vector layer after the block is segmented in the whole area is obtained by mapping the segmented block data with the remote sensing image picture of the whole area according to pixels based on the block segmentation model and adding geographic coordinate information;
and the crop category data layer of each pixel point on the land in the whole area is obtained by identifying and predicting a radar data set of each pixel point on the land based on the soybean identification model.
7. The soybean plot recognition system is characterized by comprising a data input module and a recognition module;
the data input module is used for acquiring remote sensing image pictures of the area to be identified;
the identification module is used for inputting the remote sensing image picture and outputting the identification result of the soybean land block of the area to be identified;
the identification module is obtained after training based on a sample image of the soybean sampling land block and a corresponding identification label, and the identification label is predetermined according to the soybean sampling land block and corresponds to the sample image one by one;
the identification module comprises a land parcel segmentation model, a soybean identification model and a space statistics model;
the identification module is specifically used for inputting the remote sensing image picture into the land parcel segmentation model, and outputting a segmented land parcel image and a segmented land parcel vector image layer in the whole area;
inputting the segmented land block images into the soybean recognition model, and outputting crop class data layers of all pixel points on the land block in the whole area;
and inputting the segmented land block vector image layer in the whole area and the crop category data image layer of each pixel point on the land block in the whole area into the space statistical model for superposition analysis, and judging the land block and the statistical soybean planting area according to the superposition analysis result.
8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor performs the steps of the soybean plot identification method of any one of claims 1 to 6 when the program is executed.
9. A non-transitory computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of the soybean plot identification method of any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110060578.6A CN112966548B (en) | 2021-01-18 | 2021-01-18 | Soybean plot identification method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110060578.6A CN112966548B (en) | 2021-01-18 | 2021-01-18 | Soybean plot identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112966548A CN112966548A (en) | 2021-06-15 |
| CN112966548B true CN112966548B (en) | 2023-10-27 |
Family
ID=76271547
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110060578.6A Active CN112966548B (en) | 2021-01-18 | 2021-01-18 | Soybean plot identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112966548B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113657158A (en) * | 2021-07-13 | 2021-11-16 | 西安电子科技大学 | Google Earth Engine-based large-scale soybean planting region extraction algorithm |
| CN113780057A (en) * | 2021-07-23 | 2021-12-10 | 贵州图智信息技术有限公司 | Idle land identification method and device |
| CN113763396B (en) * | 2021-09-02 | 2023-07-25 | 中国农业科学院农业信息研究所 | Remote sensing image land parcel extraction method and system based on deep learning |
| CN114529097B (en) * | 2022-02-26 | 2023-01-24 | 黑龙江八一农垦大学 | Multi-scale crop phenological period remote sensing dimensionality reduction prediction method |
| CN115082808B (en) * | 2022-06-17 | 2023-05-09 | 安徽大学 | Soybean planting area extraction method based on high-resolution first data and U-Net model |
| CN115205691B (en) * | 2022-09-15 | 2023-01-13 | 江西农业大学 | Rice planting area identification method and device, storage medium and equipment |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102800113A (en) * | 2012-07-18 | 2012-11-28 | 合肥工业大学 | Digital image analysis method based on fractal dimension |
| CN103544502A (en) * | 2013-10-29 | 2014-01-29 | 上海市城市建设设计研究总院 | High-resolution remote-sensing image ship extraction method based on SVM |
| CN103714562A (en) * | 2013-12-26 | 2014-04-09 | 广东威创视讯科技股份有限公司 | Method and device for determining drawing point in complex area |
| CN104915636A (en) * | 2015-04-15 | 2015-09-16 | 北京工业大学 | Remote sensing image road identification method based on multistage frame significant characteristics |
| CN104931022A (en) * | 2015-04-21 | 2015-09-23 | 国家测绘地理信息局卫星测绘应用中心 | Satellite image three-dimensional area network adjustment method based on satellite-borne laser height measurement data |
| CN108335261A (en) * | 2018-02-08 | 2018-07-27 | 重庆市地理信息中心 | A kind of Optical remote satellite orthography garland region automatic testing method |
| CN109543630A (en) * | 2018-11-28 | 2019-03-29 | 苏州中科天启遥感科技有限公司 | Remote sensing image forest land extracting method and system, storage medium, electronic equipment based on deep learning |
| CN109840553A (en) * | 2019-01-17 | 2019-06-04 | 苏州中科天启遥感科技有限公司 | The extracting method and system, storage medium, electronic equipment for agrotype of ploughing |
| CN110008854A (en) * | 2019-03-18 | 2019-07-12 | 中交第二公路勘察设计研究院有限公司 | Unmanned plane image Highway Geological Disaster recognition methods based on pre-training DCNN |
| CN110598503A (en) * | 2018-06-12 | 2019-12-20 | 何雪梅 | Remote sensing image recognition algorithm of neural network |
| CN111414878A (en) * | 2020-03-26 | 2020-07-14 | 京东城市(北京)数字科技有限公司 | Method and device for social attribute analysis and image processing of land parcel |
| CN111476197A (en) * | 2020-04-24 | 2020-07-31 | 中科天盛卫星技术服务有限公司 | Oil palm identification and area extraction method and system based on multi-source satellite remote sensing image |
| CN111768337A (en) * | 2020-06-01 | 2020-10-13 | 中国科学院空天信息创新研究院 | Image processing method and device and electronic equipment |
| CN112101309A (en) * | 2020-11-12 | 2020-12-18 | 北京道达天际科技有限公司 | Ground object target identification method and device based on deep learning segmentation network |
-
2021
- 2021-01-18 CN CN202110060578.6A patent/CN112966548B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102800113A (en) * | 2012-07-18 | 2012-11-28 | 合肥工业大学 | Digital image analysis method based on fractal dimension |
| CN103544502A (en) * | 2013-10-29 | 2014-01-29 | 上海市城市建设设计研究总院 | High-resolution remote-sensing image ship extraction method based on SVM |
| CN103714562A (en) * | 2013-12-26 | 2014-04-09 | 广东威创视讯科技股份有限公司 | Method and device for determining drawing point in complex area |
| CN104915636A (en) * | 2015-04-15 | 2015-09-16 | 北京工业大学 | Remote sensing image road identification method based on multistage frame significant characteristics |
| CN104931022A (en) * | 2015-04-21 | 2015-09-23 | 国家测绘地理信息局卫星测绘应用中心 | Satellite image three-dimensional area network adjustment method based on satellite-borne laser height measurement data |
| CN108335261A (en) * | 2018-02-08 | 2018-07-27 | 重庆市地理信息中心 | A kind of Optical remote satellite orthography garland region automatic testing method |
| CN110598503A (en) * | 2018-06-12 | 2019-12-20 | 何雪梅 | Remote sensing image recognition algorithm of neural network |
| CN109543630A (en) * | 2018-11-28 | 2019-03-29 | 苏州中科天启遥感科技有限公司 | Remote sensing image forest land extracting method and system, storage medium, electronic equipment based on deep learning |
| CN109840553A (en) * | 2019-01-17 | 2019-06-04 | 苏州中科天启遥感科技有限公司 | The extracting method and system, storage medium, electronic equipment for agrotype of ploughing |
| CN110008854A (en) * | 2019-03-18 | 2019-07-12 | 中交第二公路勘察设计研究院有限公司 | Unmanned plane image Highway Geological Disaster recognition methods based on pre-training DCNN |
| CN111414878A (en) * | 2020-03-26 | 2020-07-14 | 京东城市(北京)数字科技有限公司 | Method and device for social attribute analysis and image processing of land parcel |
| CN111476197A (en) * | 2020-04-24 | 2020-07-31 | 中科天盛卫星技术服务有限公司 | Oil palm identification and area extraction method and system based on multi-source satellite remote sensing image |
| CN111768337A (en) * | 2020-06-01 | 2020-10-13 | 中国科学院空天信息创新研究院 | Image processing method and device and electronic equipment |
| CN112101309A (en) * | 2020-11-12 | 2020-12-18 | 北京道达天际科技有限公司 | Ground object target identification method and device based on deep learning segmentation network |
Non-Patent Citations (5)
| Title |
|---|
| Harvest date forecast for soybeans from maximum vegetative development using satellite images;Willyan Ronaldo Becker 等;《International Journal of Remote Sensing》;1121-1138 * |
| 卫星遥感估产技术在大豆区域收入保险中的应用;陈爱莲 等;《智慧农业》;第2卷(第03期);139-152，摘要，正文第3.2节 * |
| 基于植被指数时序谱类内差异特征的冬小麦遥感识别研究;杨闫君;《中国博士论文全文数据库 农业科技辑》;D043-3 * |
| 基于轻量和积网络的田间杂草识别;王顺;《中国优秀硕士学位论文全文数据库 农业科技辑》;D046-341 * |
| 基于遥感技术的作物病虫害监测研究进展;黄文江 等;《农业工程技术》;39-45 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112966548A (en) | 2021-06-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112966548B (en) | Soybean plot identification method and system | |
| CN111986099B (en) | Tillage monitoring method and system based on convolutional neural network with residual error correction fused | |
| CN108229490B (en) | Key point detection method, neural network training method, device and electronic equipment | |
| CN112949549A (en) | Super-resolution-based change detection method for multi-resolution remote sensing image | |
| US20220215656A1 (en) | Method, apparatus, device for image processing, and storage medium | |
| CN110363176B (en) | Image analysis method and device | |
| CN113192646B (en) | Target detection model construction method and device for monitoring distance between different targets | |
| CN112927234A (en) | Point cloud semantic segmentation method and device, electronic equipment and readable storage medium | |
| WO2024060684A1 (en) | Model training method, image processing method, device, and storage medium | |
| CN112464766A (en) | Farmland automatic identification method and system | |
| Farooq et al. | Transferable convolutional neural network for weed mapping with multisensor imagery | |
| CN111507119A (en) | Identification code identification method and device, electronic equipment and computer readable storage medium | |
| CN110751163B (en) | Target positioning method and device, computer readable storage medium and electronic equipment | |
| CN111626313B (en) | Feature extraction model training method, image processing method and device | |
| CN116704324A (en) | Target detection method, system, equipment and storage medium based on underwater image | |
| CN111104965A (en) | Vehicle target identification method and device | |
| CN116246161A (en) | Method and device for identifying target fine type of remote sensing image under guidance of domain knowledge | |
| CN116091913A (en) | Method, device, equipment and storage medium for identifying block boundary vector | |
| CN115471745A (en) | Network model and device for plant identification and electronic equipment | |
| CN115223181A (en) | Text detection-based method and device for recognizing characters of seal of report material | |
| Sartorio et al. | An adaptive approach for simultaneous classification of remote sensing scenes including rural and urban targets | |
| CN114419308A (en) | Color identification method, system, electronic device and medium based on region of interest | |
| Khan et al. | Crop Type Classification using Multi-temporal Sentinel-2 Satellite Imagery: A Deep Semantic Segmentation Approach | |
| CN110852145B (en) | Image detection method, device and system for unmanned aerial vehicle image | |
| Pang et al. | PTRSegNet: A Patch-to-Region Bottom-Up Pyramid Framework for the Semantic Segmentation of Large-Format Remote Sensing Images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |