A corn biomass inversion method based on UAV microwave radiation data

Abstract

A corn biomass inversion method based on drone microwave radiation data belongs to the technical field of agricultural biomass inversion prediction. The present invention solves the problem of low accuracy of existing corn biomass inversion methods. The present invention establishes a multi-layer penetration transformation attribute description module based on drone microwave radiation data, which can describe the characteristics of corn at various levels from a vertical scale, and then use a random forest model to learn these characteristics and regress corn biomass. The method of the present invention can obtain accurate pixel-level corn biomass estimation results of drone microwave images. The present invention effectively prevents the data at different positions of the corn plant from being mixed together for decision-making by layering the original data, and can effectively discover the correlation between drone microwave radiation data and corn biomass, thereby effectively improving the accuracy of corn biomass estimation results. The method of the present invention can be applied to corn biomass inversion.

Description

Technical Field

[0001] The present invention belongs to the technical field of agricultural biomass inversion prediction, and in particular relates to a corn biomass inversion method based on drone microwave radiation data. Background Art

[0002] As an important food crop and industrial raw material in the world, corn production is directly related to national food security, agricultural economic stability, and the realization of carbon sequestration ecological functions. Realizing automated and accurate inversion of corn biomass can provide a scientific basis for regional growth status assessment, disaster warning, and resource allocation, and help management departments formulate regulatory strategies in advance and optimize planting structures, thereby playing a significant role in ensuring food security and improving management efficiency. Especially in the context of climate change and tight arable land resources, the development of efficient inversion methods has become an urgent need for the development of smart agriculture, and has very important social and economic value. The current mainstream corn biomass inversion technologies can be divided into the following three categories:

[0003] (1) Vegetation index method based on optical remote sensing. For example, spectral characteristic parameters such as NDVI (Normalized Difference Vegetation Index) and EVI (Enhanced Vegetation Index) are used to establish an inversion model. However, this method relies on the optical reflectance characteristics of the vegetation canopy. Due to the height of corn plants, when the canopy density is high in the middle and late stages of corn growth, the penetration ability decreases significantly, making it difficult to capture the biomass information in the lower layers, resulting in a significant reduction in inversion accuracy.

[0004] (2) Satellite microwave remote sensing technology. Although it has the advantage of penetrating clouds and vegetation, its single pixel spatial resolution is extremely low (about 25 km × 25 km), which cannot meet the needs of accurate monitoring at the farmland scale.

[0005] (3) UAV microwave radiation technology. Machine learning methods are introduced for data fusion and decision-making. Due to the diversity of corn varieties, there are many differences in height and density within the field. At the same time, it is limited by the inability to collect a full set of field samples (collecting one sample requires harvesting corn in the corresponding area of ​​one pixel, drying it, and weighing it). This easily leads to the confusion of corn samples of multiple heights and densities, causing the model to overfit to a certain ineffective feature of the sample, making it difficult to stably support the dynamic inversion of corn biomass in a large-scale scenario.

[0006] In summary, the existing technology still has problems such as insufficient resolution, high corn plants making it difficult to capture the biomass information of the lower layers, and confusion in the inversion content caused by the variety and height of the corn plants. Therefore, the accuracy of the existing corn biomass inversion method is still low. There is an urgent need for a new type of automated inversion method for corn biomass to achieve stable and high-precision inversion of corn biomass. Summary of the Invention

[0007] The purpose of the present invention is to solve the problem of low accuracy of existing corn biomass inversion methods and propose a corn biomass inversion method based on drone microwave radiation data.

[0008] The technical solution adopted by the present invention to solve the above technical problems is: a corn biomass inversion method based on drone microwave radiation data, the method specifically comprising the following steps:

[0009] Step S1, obtaining the drone microwave radiation data WBData of the study area, and obtaining the width WBWidth and height WBHeight of the drone microwave radiation data according to the drone microwave radiation data WBData;

[0010] And calculate the individual reference structure WBRef of the drone microwave radiation data image;

[0011] Step S2, establishing a hierarchical microwave penetration transformation attribute description module MKDX for the top male inflorescence of a corn plant, wherein the input of the module MKDX is the row MKDXRow and column MKDXCol to be calculated by MKDX, and the output of the module MKDX is the attribute array MKDXOutput;

[0012] Step S3, establishing a microwave penetration transformation attribute description module MKSS for the upper layer of corn plant without ear stem, the input of the module MKSS is the row MKSSRow and column MKSSCol that MKSS needs to calculate, and the output of MKSS is the attribute array MKSSOutput;

[0013] Step S4, establishing a microwave penetration transformation attribute description module MKYG for the middle layer of corn plants with ears and stems, the input of the module MKYG is the row MKYGRow and column MKYGCol that MKYG needs to calculate, and the output of MKYG is the attribute array MKYGOutput;

[0014] Step S5, establishing a corn plant multi-layer penetration transformation attribute description module MKDCC, wherein the input of MKDCC is the row MKDCCRow and column MKDCCCol that MKDCC needs to calculate, and the output of the module MKDCC is obtained by calculation using the modules MKDX, MKSS, and MKYG. The output of MKDCC is the attribute array MKDCCOutput;

[0015] Step S6: input the corn biomass dataset YMList, use the module MKDCC and the dataset YMList to construct the corn biomass decision data table YMDecision, and use YMDecision to construct the multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel;

[0016] Step S7: Utilize the module MKDCC and the model DCCForestModel to perform corn biomass inversion prediction and obtain the corn biomass inversion prediction result FYResult.

[0017] The beneficial effects of the present invention are:

[0018] The method of the present invention establishes a multi-layer penetration transformation attribute description module based on drone microwave radiation data. This module can describe the characteristics of corn at various levels on a vertical scale. Then, a random forest model is used to learn these characteristics and regress corn biomass. The method of the present invention can obtain accurate pixel-level corn biomass estimation results from drone microwave imagery. The method of the present invention can fully utilize the high precision of drone microwave radiation data to meet the requirements of precise monitoring at the farmland scale. By layering the original data, it effectively prevents the confusion of data from different locations of the corn plant in decision-making. It can effectively discover the correlation between drone microwave radiation data and corn biomass, thereby effectively improving the accuracy of corn biomass estimation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flow chart of a corn biomass inversion method based on drone microwave radiation data. DETAILED DESCRIPTION

[0020] Specific implementation method 1: Combination Figure 1 This embodiment describes a corn biomass inversion method based on drone microwave radiation data, which specifically includes the following steps:

[0021] Step S1, obtaining the UAV microwave radiation data WBData of the study area (i.e., the area to be inverted), and obtaining the width WBWidth and height WBHeight of the UAV microwave radiation data according to the UAV microwave radiation data WBData;

[0022] And calculate the individual reference structure WBRef of the drone microwave radiation data image;

[0023] Step S2, establishing a hierarchical microwave penetration transformation attribute description module MKDX for the top male inflorescence of a corn plant, wherein the input of the module MKDX is the row MKDXRow and column MKDXCol to be calculated by MKDX, and the output of the module MKDX is the attribute array MKDXOutput;

[0024] Step S3, establishing a microwave penetration transformation attribute description module MKSS for the upper layer of corn plant without ear stem, the input of the module MKSS is the row MKSSRow and column MKSSCol that MKSS needs to calculate, and the output of MKSS is the attribute array MKSSOutput;

[0025] Step S4, establishing a microwave penetration transformation attribute description module MKYG for the middle layer of corn plants with ears and stems, the input of the module MKYG is the row MKYGRow and column MKYGCol that MKYG needs to calculate, and the output of MKYG is the attribute array MKYGOutput;

[0026] Step S5, establishing a multi-layer penetration transformation attribute description module MKDCC for corn plants (in the present invention, the multi-layer penetration transformation attributes include the microwave penetration transformation attributes of the corn plant's top male inflorescence layer, the microwave penetration transformation attributes of the corn plant's upper layer without ear stem layer, and the microwave penetration transformation attributes of the corn plant's middle layer with ear stem layer). The input of MKDCC is the rows MKDCCRow and columns MKDCCCol to be calculated by MKDCC. The output of the module MKDCC is calculated using the modules MKDX, MKSS, and MKYG. The output of MKDCC is the attribute array MKDCCOutput.

[0027] Step S6: input the corn biomass dataset YMList, use the module MKDCC and the dataset YMList to construct the corn biomass decision data table YMDecision, and use YMDecision to construct the multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel;

[0028] Step S7: Utilize the module MKDCC and the model DCCForestModel to perform corn biomass inversion prediction and obtain the corn biomass inversion prediction result FYResult.

[0029] Specific embodiment 2: This embodiment differs from specific embodiment 1 in that the specific process of step S1 is as follows:

[0030] Step S101: Obtain the drone microwave radiation data WBData of the study area. WBData is an image array containing 8 elements, where:

[0031] The first element is a microwave brightness temperature image obtained using a vertically polarized 18 GHz microwave radiometer at an observation angle of 55 degrees.

[0032] The observation inclination is the angle between the observation angle and the horizontal plane;

[0033] The second element is a microwave brightness temperature image obtained using a horizontally polarized 18 GHz microwave radiometer at an observation angle of 55 degrees.

[0034] The third element is a microwave brightness temperature image obtained using a vertically polarized 18 GHz microwave radiometer at an observation angle of 35 degrees.

[0035] The fourth element is a microwave brightness temperature image obtained using a horizontally polarized 18 GHz microwave radiometer at an observation angle of 35 degrees.

[0036] The fifth element is a microwave brightness temperature image obtained using a vertically polarized 36 GHz microwave radiometer at an observation angle of 55 degrees.

[0037] The sixth element is a microwave brightness temperature image obtained using a horizontally polarized 36 GHz microwave radiometer at an observation angle of 55 degrees.

[0038] The seventh element is a microwave brightness temperature image obtained using a vertically polarized 36 GHz microwave radiometer at an observation angle of 35 degrees.

[0039] The eighth element is a microwave brightness temperature image obtained using a horizontal polarization at a frequency of 36 GHz and an observation tilt of 35 degrees using a drone-mounted microwave radiometer.

[0040] Step S102: The width of the drone microwave radiation data WBWidth=the width of the first element in WBData;

[0041] Step S103: The height of the drone microwave radiation data WBHeight=the height of the first element in WBData;

[0042] Step S104: Create an individual reference structure WBRef of the drone microwave radiation data image, where WBRef is a floating-point array containing 16 elements.

[0043] Initialize all elements in the array to 0;

[0044] Step S105, initialize the element counter InitCounter in the array = 1;

[0045] Step S106, initialization phase: the first temporary variable InitTemp1 = the average value of all pixels of the image of the InitCounterth element in WBData;

[0046] Step S107, initialization phase: the second temporary variable InitTemp2 = the standard deviation of all pixels of the image of the InitCounterth element in WBData;

[0047] Step S108, initialization phase: the third temporary variable InitTemp3 = InitTemp1 - 1.8 × InitTemp2;

[0048] Step S109, initialization phase: fourth temporary variable InitTemp4 = InitTemp1 + 1.8 × InitTemp2;

[0049] Step S110: If InitCounter is greater than 4, update the values ​​of InitTemp3 and InitTemp4 to: InitTemp3 = InitTemp3 - 0.4 × InitTemp2 and InitTemp4 = InitTemp4 + 0.4 × InitTemp2;

[0050] The right side of the formula is the InitTemp3 value and InitTemp4 value before the update, and the left side of the formula is the InitTemp3 value and InitTemp4 value after the update;

[0051] If InitCounter is less than or equal to 4, there is no need to process InitTemp3 and InitTemp4;

[0052] Step S111, set the value of the InitCounter×2-1th element of WBRef to InitTemp3, and set the value of the InitCounter×2th element of WBRef to InitTemp4;

[0053] Step S112: Set InitCounter=InitCounter+1;

[0054] Step S113: If InitCounter is less than or equal to 8, go to step S106; otherwise, go to step S114.

[0055] Step S114, end.

[0056] Other steps and parameters are the same as those in the first embodiment.

[0057] Specific embodiment 3: This embodiment differs from specific embodiment 1 or 2 in that the specific process of step S2 is as follows:

[0058] Step S201, establishing a hierarchical microwave penetration transformation attribute description module MKDX of the male inflorescence at the top of the corn plant, wherein the input of the module MKDX is the row MKDXRow and column MKDXCol that MKDX needs to calculate;

[0059] Step S202, create the first temporary variable MKDXTemp1 of MKDX = take out the value of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the third element of WBData;

[0060] Step S203, create the second temporary variable MKDXTemp2 of MKDX = take out the value of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the fourth element of WBData;

[0061] Step S204, create the third temporary variable MKDXTemp3 of MKDX = take out the value of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the 7th element of WBData;

[0062] Step S205, create the fourth temporary variable MKDXTemp4 of MKDX = take out the value of the pixel located at the MKDXRowth row and MKDXColth column in the image of the 8th element of WBData;

[0063] Step S206, create the fifth temporary variable MKDXTemp5 of MKDX = take out the average value of the eight neighboring pixels of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the third element of WBData;

[0064] Step S207, create the sixth temporary variable MKDXTemp6 of MKDX = take out the average value of the eight neighboring pixels of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the fourth element of WBData;

[0065] Step S208: Create an MKDX output attribute array MKDXOutput = a floating-point array of 4 elements, and initialize the value of each element of the array to 0;

[0066] Step S209: Create the seventh temporary variable MKDXTemp7 of MKDX as:

[0067] MKDXTemp7=(MKDXTemp2-MKDXTemp1) / (MKDXTemp3-MKDXTemp4) / abs(WBRef[5]-WBRef[6])

[0068] Where abs(·) is the calculated absolute value, “ / ” represents the division operation, WBRef[5] represents the value of the 5th element in the array WBRef, and WBRef[6] represents the value of the 6th element in the array WBRef;

[0069] Step S210: Create the eighth temporary variable MKDXTemp8 of MKDX as:

[0070] MKDXTemp8=MKDXTemp7×MKDXTemp5 / abs(WBRef[5]-WBRef[6])

[0071] Step S211, create the ninth temporary variable MKDXTemp9 of MKDX as:

[0072] MKDXTemp9=MKDXTemp7×MKDXTemp6 / abs(WBRef[7]-WBRef[8])

[0073] Among them, WBRef[7] represents the value of the 7th element in the array WBRef, and WBRef[8] represents the value of the 8th element in the array WBRef;

[0074] Step S212: Create the 10th temporary variable MKDXTemp10 of MKDX as:

[0075] MKDXTemp10=(MKDXTemp3+MKDXTemp4) / 2

[0076] Step S213, set the first element MKDXOutput[1] in the output attribute array MKDXOutput to MKDXTemp7; set the second element MKDXOutput[2] in the output attribute array MKDXOutput to MKDXTemp8; set the third element MKDXOutput[3] in the output attribute array MKDXOutput to MKDXTemp9; set the fourth element MKDXOutput[4] in the output attribute array MKDXOutput to MKDXTemp10;

[0077] Step S214: Return the output attribute array MKDXOutput as the result of MKDX.

[0078] Other steps and parameters are the same as those in the first or second embodiment.

[0079] Specific embodiment 4: This embodiment differs from any one of specific embodiments 1 to 3 in that the specific process of step S3 is as follows:

[0080] Step S301, establishing a microwave penetration transformation attribute description module MKSS for the upper layer of corn plant without ear and stem, the input of the module MKSS is the row MKSSRow and column MKSSCol that MKSS needs to calculate;

[0081] Step S302, create the first temporary variable MKSSTemp1 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the first element image of WBData;

[0082] Step S303, create the second temporary variable MKSSTemp2 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the second element image of WBData;

[0083] Step S304, create the third temporary variable MKSSTemp3 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the fifth element image of WBData;

[0084] Step S305, create the fourth temporary variable MKSSTemp4 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the sixth element image of WBData;

[0085] Step S306: Create the fifth temporary variable MKSSTemp5 of MKSS:

[0086] MKSSTemp5=(MKSSTemp1-WBRef[1]) / (WBRef[1]-WBRef[2])

[0087] Where WBRef[1] represents the value of the first element in the array WBRef, and WBRef[2] represents the value of the second element in the array WBRef;

[0088] Step S307: Create the sixth temporary variable MKSSTemp6 of MKSS as:

[0089] MKSSTemp6=(MKSSTemp2-WBRef[3]) / (WBRef[3]-WBRef[4])

[0090] Among them, WBRef[3] represents the value of the third element in the array WBRef, and WBRef[4] represents the value of the fourth element in the array WBRef;

[0091] Step S308: Create the seventh temporary variable MKSSTemp7 of MKSS as:

[0092] MKSSTemp7=(MKSSTemp3-WBRef

[13] ) / (WBRef

[10] -WBRef[9])

[0093] Among them, WBRef[9] represents the value of the 9th element in the array WBRef, WBRef

[10] represents the value of the 10th element in the array WBRef, and WBRef

[13] represents the value of the 13th element in the array WBRef;

[0094] Step S309: Create the eighth temporary variable MKSSTemp8 of MKSS as:

[0095] MKSSTemp8=(MKSSTemp4-WBRef

[15] ) / (WBRef

[12] -WBRef

[11] )

[0096] Among them, WBRef

[11] represents the value of the 11th element in the array WBRef, WBRef

[12] represents the value of the 12th element in the array WBRef, and WBRef

[15] represents the value of the 15th element in the array WBRef;

[0097] Step S310: Create an MKSS output attribute array MKSSOutput = a floating-point array of 2 elements;

[0098] The first element MKSSOutput[1] in the attribute array MKSSOutput is:

[0099] MKSSOutput[1]=tanh( (MKSSTemp5-MKSSTemp6) / (MKSSTemp5+MKSSTemp6)×2 )

[0100] Among them, tanh is the calculation of hyperbolic tangent function;

[0101] The second element MKSSOutput[2] in the attribute array MKSSOutput is:

[0102] MKSSOutput[2]=tanh( (MKSSTemp7-MKSSTemp8) / (MKSSTemp7+MKSSTemp8)×2 )

[0103] Where tanh is the hyperbolic tangent function;

[0104] Step S311: Return the attribute array MKSSOutput as the result of MKSS.

[0105] The other steps and parameters are the same as those in the first to third embodiments.

[0106] Specific embodiment 5: This embodiment differs from any one of specific embodiments 1 to 4 in that the specific process of step S4 is as follows:

[0107] Step S401: establishing a microwave penetration transformation attribute description module MKYG for the middle layer of corn plants with ears and stems, wherein the input of MKYG is the rows MKYGRow and columns MKYGCol that MKYG needs to calculate;

[0108] Step S402: Create the first temporary variable MKYGTemp1 of MKYG = the average value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the first element of WBData;

[0109] Step S403: Create a second temporary variable MKYGTemp2 of MKYG, which is the mean value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the third element of WBData;

[0110] Step S404: Create the third temporary variable MKYGTemp3 of MKYG, which is the mean value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the fifth element of WBData;

[0111] Step S405: Create the fourth temporary variable MKYGTemp4 of MKYG = the average value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the seventh element of WBData;

[0112] Step S406: Create the fifth temporary variable MKYGTemp5 of MKYG:

[0113] MKYGTemp5=(MKYGTemp1-WBRef

[14] ) / (WBRef[1]-WBRef[9])

[0114] Among them, WBRef[1] represents the value of the first element in the array WBRef, WBRef[9] represents the value of the ninth element in the array WBRef, and WBRef

[14] represents the value of the fourteenth element in the array WBRef;

[0115] Step S407: Create the sixth temporary variable MKYGTemp6 of MKYG:

[0116] MKYGTemp6=(MKYGTemp2-WBRef

[16] ) / (WBRef[5]-WBRef

[13] )

[0117] Among them, WBRef[5] represents the value of the 5th element in the array WBRef, WBRef

[13] represents the value of the 13th element in the array WBRef, and WBRef

[16] represents the value of the 16th element in the array WBRef;

[0118] Step S408: Create the seventh temporary variable MKSSTemp7 of MKYG as:

[0119] MKSSTemp7=tanh(MKYGTemp5-MKYGTemp6)

[0120] Step S409: Create the eighth temporary variable MKYGTemp8 of MKYG:

[0121] MKYGTemp8=(MKYGTemp1-MKYGTemp2) / (MKYGTemp3-MKYGTemp4)

[0122] Step S410, create an MKYG output attribute array MKYGOutput = a floating point array of 4 elements;

[0123] Set the first element MKYGOutput[1] in the attribute array MKYGOutput to MKYGTemp5, the second element MKYGOutput[2] in the attribute array MKYGOutput to MKYGTemp6, the third element MKYGOutput[3] in the attribute array MKYGOutput to MKYGTemp7, and the fourth element MKYGOutput[4] in the attribute array MKYGOutput to MKYGTemp8;

[0124] Step S411: Return MKYGOutput as the result of MKYG.

[0125] The other steps and parameters are the same as those in the first to fourth embodiments.

[0126] Specific embodiment 6: This embodiment differs from any one of specific embodiments 1 to 5 in that the specific process of step S5 is as follows:

[0127] Step S501: Establish a corn plant multi-layer penetration transformation attribute description module MKDCC, where the input of MKDCC is the row MKDCCRow and column MKDCCCol that MKDCC needs to calculate;

[0128] Step S502: Establish a first-level temporary variable MKDCCTemp1 in the corn plant multi-layer penetration transformation attribute description module, access MKDX for calculation, input MKDX row MKDXRow=MKDCCRow and input MKDX column MKDXCol=MKDCCCol, and return the attribute array MKDXOutput output by MKDX;

[0129] Step S503: Establish a second-level temporary variable MKDCCTemp2 in the corn plant multi-layer penetration transformation attribute description module, access the MKSS for calculation, input the MKSS row MKSSRow=MKDCCRow and the MKSS column MKSSCol=MKDCCCol, and return the attribute array MKSSOutput output by the MKSS;

[0130] Step S504: Establish a third-level temporary variable MKDCCTemp3 in the corn plant multi-layer penetration transformation attribute description module, access MKYG for calculation, input MKYG row MKYGRow=MKDCCRow and MKYG column MKYGCol=MKDCCCol, and return the attribute array MKYGOutput output by MKYG;

[0131] Step S505, create an MKDCC output attribute array MKDCCOutput = create a floating point array containing 10 elements;

[0132] Set the first element MKDCCOutput[1] in the attribute array MKDCCOutput to MKDCCTemp1[1]; the second element MKDCCOutput[2] in the attribute array MKDCCOutput to MKDCCTemp1[2]; the third element MKDCCOutput[3] in the attribute array MKDCCOutput to MKDCCTemp1[3]; the fourth element MKDCCOutput[4] in the attribute array MKDCCOutput to MKDCCTemp1[4];

[0133] Among them, MKDCCTemp1[1] represents the first element in the temporary variable of the first level, MKDCCTemp1[2] represents the second element in the temporary variable of the first level, MKDCCTemp1[3] represents the third element in the temporary variable of the first level, and MKDCCTemp1[4] represents the fourth element in the temporary variable of the first level;

[0134] Set the fifth element MKDCCOutput[5] in the attribute array MKDCCOutput to MKDCCTemp2[1]; the sixth element MKDCCOutput[6] in the attribute array MKDCCOutput to MKDCCTemp2[2];

[0135] Among them, MKDCCTemp2[1] represents the first element in the second-level temporary variable, and MKDCCTemp2[2] represents the second element in the second-level temporary variable;

[0136] Set the 7th element MKDCCOutput[7] in the attribute array MKDCCOutput to MKDCCTemp3[1]; the 8th element MKDCCOutput[8] in the attribute array MKDCCOutput to MKDCCTemp3[2]; the 9th element MKDCCOutput[9] in the attribute array MKDCCOutput to MKDCCTemp3[3]; the 10th element MKDCCOutput

[10] in the attribute array MKDCCOutput to MKDCCTemp3[4];

[0137] Among them, MKDCCTemp3[1] represents the first element in the temporary variable of the third level, MKDCCTemp3[2] represents the second element in the temporary variable of the third level, MKDCCTemp3[3] represents the third element in the temporary variable of the third level, and MKDCCTemp3[4] represents the fourth element in the temporary variable of the third level;

[0138] Step S506: Return the attribute array MKDCCOutput as the result of MKDCC.

[0139] The other steps and parameters are the same as those in the first to fifth embodiments.

[0140] Specific embodiment 7: This embodiment differs from any one of specific embodiments 1 to 6 in that the specific process of step S6 is as follows:

[0141] Step S601: Input the corn biomass dataset YMList. YMList is a list containing the YMROW field, the YMCOL field, and the SWL field. The YMROW field, the YMCOL field, and the SWL field are:

[0142] YMROW field: the row where the data collection point is located in the drone microwave radiation data image;

[0143] YMCOL field: the column where the data collection point is located in the drone microwave radiation data image;

[0144] SWL field: biomass of corn;

[0145] Step S602: Create a corn biomass decision data table YMDecision containing the DCCFeature field and the DSWL field. YMDecision contains two fields:

[0146] DCCFeature field: multi-layer penetration transformation attributes;

[0147] DSWL field: biomass of corn used for inversion prediction;

[0148] The specific process of step S602 is:

[0149] Step 1: Initialize the multi-layer penetration transform attribute analysis counter DCCCounter = 1;

[0150] Step 2: Create a new row of data YMDecisionRow for YMDecision;

[0151] Step 3: Multi-layer penetration transform attribute description temporary variable DCCTemp1 = access MKDCC for calculation, input MKDCCRow of MKDCC row = the value of YMROW field of DCCCounter-th element in YMList, input MKDCCCol of MKDCC column = the value of YMCOL field of DCCCounter-th element in YMList, and return MKDCCOutput;

[0152] Step 4. Set the value of the DCCFeature field of YMDecisionRow to DCCTemp1;

[0153] Step 5. Set the value of the DSWL field of YMDecisionRow to the value of the SWL field of the DCCCounterth element in YMList;

[0154] Step 6: Set DCCCounter=DCCCounter+1;

[0155] Step 7: If DCCCounter is less than or equal to the total number of elements in YMList, go to step 2; otherwise, stop the iteration and obtain the corn biomass decision data table YMDecision.

[0156] Step S603: Establish a multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel, where the input of the model DCCForestModel is the DCCFeature field of YMDecision, and the label of the model DCCForestModel prediction output is the DSWL field of YMDecision;

[0157] Step S604: Use all the data of YMDecision to train the DCCForestModel model;

[0158] The trained model DCCForestModel is able to predict biomass;

[0159] Step S605, end.

[0160] The other steps and parameters are the same as those in the first to sixth embodiments.

[0161] Specific embodiment eight: This embodiment differs from any one of specific embodiments one to seven in that the multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel is a data regression decision forest.

[0162] The other steps and parameters are the same as those in the first to seventh embodiments.

[0163] Specific embodiment 9: This embodiment differs from any one of specific embodiments 1 to 8 in that the specific process of step S7 is as follows:

[0164] Step S701, establishing corn biomass inversion prediction result FYResult = establishing a two-dimensional matrix with the number of rows being WBHeight and the number of columns being WBWidth, and initializing the value of each element in the two-dimensional matrix to be 0;

[0165] Step S702, inversion row counter FYROW=1;

[0166] Step S703, inversion column counter FYCOL=1;

[0167] Step S704: The first temporary variable FYTemp1 used for inversion is connected to MKDCC for calculation, the row MKDCCRow=FYROW of MKDCC is input, the column MKDCCCol=FYCOL of MKDCC is input, and the output MKDCCOutput of MKDCC is returned;

[0168] Step S705, the second temporary variable FYTemp2 for inversion = FYTemp1 is connected to the model DCCForestModel, and the inversion prediction result of the FYROWth row and the FYCOLth column in FYResult is obtained through the DCCForestModel model;

[0169] Step S706: Set the value of the FYROWth row and the FYCOLth column of FYResult to FYTemp2;

[0170] Step S707: Set FYCOL=FYCOL+1;

[0171] Step S708: If FYCOL is less than or equal to WBWidth, go to step S703; otherwise, go to step S708;

[0172] Step S709: set FYROW=FYROW+1;

[0173] Step S710: If FYROW is less than or equal to WBHeight, go to step S702; otherwise, go to step S710.

[0174] Step S711, outputting FYResult as the corn biomass inversion result;

[0175] Step S712, end.

[0176] The other steps and parameters are the same as those in Specific Embodiments 1 to 8.

[0177] The above examples are merely illustrative of the calculation model and process of the present invention and are not intended to limit the embodiments of the present invention. Persons skilled in the art will readily appreciate that other variations or modifications based on the above description are possible. This list of embodiments is not exhaustive; however, any obvious variations or modifications derived from the technical solution of the present invention remain within the scope of protection of the present invention.

Claims

1. A corn biomass inversion method based on drone microwave radiation data, characterized in that: The method specifically comprises the following steps: Step S1, obtaining the drone microwave radiation data WBData of the study area, and obtaining the width WBWidth and height WBHeight of the drone microwave radiation data according to the drone microwave radiation data WBData; And calculate the individual reference structure WBRef of the drone microwave radiation data image; Step S2, establishing a hierarchical microwave penetration transformation attribute description module MKDX for the top male inflorescence of a corn plant, wherein the input of the module MKDX is the row MKDXRow and column MKDXCol to be calculated by MKDX, and the output of the module MKDX is the attribute array MKDXOutput; The specific process of step S2 is: Step S201, establishing a hierarchical microwave penetration transformation attribute description module MKDX of the male inflorescence at the top of the corn plant, wherein the input of the module MKDX is the row MKDXRow and column MKDXCol that MKDX needs to calculate; Step S202, create the first temporary variable MKDXTemp1 of MKDX = take out the value of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the third element of WBData; Step S203, create the second temporary variable MKDXTemp2 of MKDX = take out the value of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the fourth element of WBData; Step S204, create the third temporary variable MKDXTemp3 of MKDX = take out the value of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the 7th element of WBData; Step S205, create the fourth temporary variable MKDXTemp4 of MKDX = take out the value of the pixel located at the MKDXRowth row and MKDXColth column in the image of the 8th element of WBData; Step S206, create the fifth temporary variable MKDXTemp5 of MKDX = take out the average value of the eight neighboring pixels of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the third element of WBData; Step S207, create the sixth temporary variable MKDXTemp6 of MKDX = take out the average value of the eight neighboring pixels of the pixel located at the MKDXRowth row and the MKDXColth column in the image of the fourth element of WBData; Step S208: Create an MKDX output attribute array MKDXOutput = a floating-point array of 4 elements, and initialize the value of each element of the array to 0; Step S209: Create the seventh temporary variable MKDXTemp7 of MKDX as: MKDXTemp7=(MKDXTemp2-MKDXTemp1) / (MKDXTemp3-MKDXTemp4) / abs(WBRef[5]-WBRef[6]) Where abs(·) is the calculated absolute value, WBRef[5] represents the value of the 5th element in the array WBRef, and WBRef[6] represents the value of the 6th element in the array WBRef; Step S210: Create the eighth temporary variable MKDXTemp8 of MKDX as: MKDXTemp8=MKDXTemp7×MKDXTemp5 / abs(WBRef[5]-WBRef[6]) Step S211, create the ninth temporary variable MKDXTemp9 of MKDX as: MKDXTemp9=MKDXTemp7×MKDXTemp6 / abs(WBRef[7]-WBRef[8]) Among them, WBRef[7] represents the value of the 7th element in the array WBRef, and WBRef[8] represents the value of the 8th element in the array WBRef; Step S212: Create the 10th temporary variable MKDXTemp10 of MKDX as: MKDXTemp10=(MKDXTemp3+MKDXTemp4) / 2 Step S213, set the first element MKDXOutput[1] in the output attribute array MKDXOutput to MKDXTemp7; set the second element MKDXOutput[2] in the output attribute array MKDXOutput to MKDXTemp8; set the third element MKDXOutput[3] in the output attribute array MKDXOutput to MKDXTemp9; set the fourth element MKDXOutput[4] in the output attribute array MKDXOutput to MKDXTemp10; Step S214: Return the output attribute array MKDXOutput as the result of MKDX; Step S3, establishing a microwave penetration transformation attribute description module MKSS for the upper layer of corn plant without ear stem, the input of the module MKSS is the row MKSSRow and column MKSSCol that MKSS needs to calculate, and the output of MKSS is the attribute array MKSSOutput; Step S4, establishing a microwave penetration transformation attribute description module MKYG for the middle layer of corn plants with ears and stems, the input of the module MKYG is the row MKYGRow and column MKYGCol that MKYG needs to calculate, and the output of MKYG is the attribute array MKYGOutput; Step S5, establishing a corn plant multi-layer penetration transformation attribute description module MKDCC, wherein the input of MKDCC is the row MKDCCRow and column MKDCCCol that MKDCC needs to calculate, and the output of the module MKDCC is obtained by calculation using the modules MKDX, MKSS, and MKYG. The output of MKDCC is the attribute array MKDCCOutput; Step S6: input the corn biomass dataset YMList, use the module MKDCC and the dataset YMList to construct the corn biomass decision data table YMDecision, and use YMDecision to construct the multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel; Step S7: Utilize the module MKDCC and the model DCCForestModel to perform corn biomass inversion prediction and obtain the corn biomass inversion prediction result FYResult.

2. The corn biomass inversion method based on drone microwave radiation data according to claim 1 is characterized in that: The specific process of step S1 is: Step S101: Obtain the drone microwave radiation data WBData of the study area. WBData is an image array containing 8 elements, where: The first element is a microwave brightness temperature image obtained using a vertically polarized 18 GHz microwave radiometer at an observation angle of 55 degrees. The second element is a microwave brightness temperature image obtained using a horizontally polarized 18 GHz microwave radiometer at an observation angle of 55 degrees. The third element is a microwave brightness temperature image obtained using a vertically polarized 18 GHz microwave radiometer at an observation angle of 35 degrees. The fourth element is a microwave brightness temperature image obtained using a horizontally polarized 18 GHz microwave radiometer at an observation angle of 35 degrees. The fifth element is a microwave brightness temperature image obtained using a vertically polarized 36 GHz microwave radiometer at an observation angle of 55 degrees. The sixth element is a microwave brightness temperature image obtained using a horizontally polarized 36 GHz microwave radiometer at an observation angle of 55 degrees. The seventh element is a microwave brightness temperature image obtained using a vertically polarized 36 GHz microwave radiometer at an observation angle of 35 degrees. The eighth element is a microwave brightness temperature image obtained using a horizontal polarization at a frequency of 36 GHz and an observation tilt of 35 degrees using a drone-mounted microwave radiometer. Step S102: The width of the drone microwave radiation data WBWidth=the width of the first element in WBData; Step S103: The height of the drone microwave radiation data WBHeight=the height of the first element in WBData; Step S104: Create an individual reference structure WBRef of the drone microwave radiation data image, where WBRef is a floating-point array containing 16 elements. Initialize all elements in the array to 0; Step S105, initialize the element counter InitCounter in the array = 1; Step S106, initialization phase: the first temporary variable InitTemp1 = the average value of all pixels of the image of the InitCounterth element in WBData; Step S107, initialization phase: the second temporary variable InitTemp2 = the standard deviation of all pixels of the image of the InitCounterth element in WBData; Step S108, initialization phase: the third temporary variable InitTemp3 = InitTemp1 - 1.8 × InitTemp2; Step S109, initialization phase: fourth temporary variable InitTemp4 = InitTemp1 + 1.8 × InitTemp2; Step S110: If InitCounter is greater than 4, update the values ​​of InitTemp3 and InitTemp4 to: InitTemp3 = InitTemp3 - 0.4 × InitTemp2 and InitTemp4 = InitTemp4 + 0.4 × InitTemp2; If InitCounter is less than or equal to 4, there is no need to process InitTemp3 and InitTemp4; Step S111, set the value of the InitCounter×2-1th element of WBRef to InitTemp3, and set the value of the InitCounter×2th element of WBRef to InitTemp4; Step S112: Set InitCounter=InitCounter+1; Step S113: If InitCounter is less than or equal to 8, go to step S106; otherwise, go to step S114. Step S114, end.

3. The corn biomass inversion method based on drone microwave radiation data according to claim 2 is characterized in that: The specific process of step S3 is: Step S301, establishing a microwave penetration transformation attribute description module MKSS for the upper layer of corn plant without ear and stem, the input of the module MKSS is the row MKSSRow and column MKSSCol that MKSS needs to calculate; Step S302, create the first temporary variable MKSSTemp1 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the first element image of WBData; Step S303, create the second temporary variable MKSSTemp2 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the second element image of WBData; Step S304, create the third temporary variable MKSSTemp3 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the fifth element image of WBData; Step S305, create the fourth temporary variable MKSSTemp4 of MKSS = retrieve the value of the pixel located at the MKSSRowth row and the MKSSColth column in the sixth element image of WBData; Step S306: Create the fifth temporary variable MKSSTemp5 of MKSS as: MKSSTemp5=(MKSSTemp1-WBRef[1]) / (WBRef[1]-WBRef[2]) Where WBRef[1] represents the value of the first element in the array WBRef, and WBRef[2] represents the value of the second element in the array WBRef; Step S307: Create the sixth temporary variable MKSSTemp6 of MKSS as: MKSSTemp6=(MKSSTemp2-WBRef[3]) / (WBRef[3]-WBRef[4]) Among them, WBRef[3] represents the value of the third element in the array WBRef, and WBRef[4] represents the value of the fourth element in the array WBRef; Step S308: Create the seventh temporary variable MKSSTemp7 of MKSS as: MKSSTemp7=(MKSSTemp3-WBRef[13]) / (WBRef[10]-WBRef[9]) Among them, WBRef[9] represents the value of the 9th element in the array WBRef, WBRef[10] represents the value of the 10th element in the array WBRef, and WBRef[13] represents the value of the 13th element in the array WBRef; Step S309: Create the eighth temporary variable MKSSTemp8 of MKSS as: MKSSTemp8=(MKSSTemp4-WBRef[15]) / (WBRef[12]-WBRef[11]) Among them, WBRef[11] represents the value of the 11th element in the array WBRef, WBRef[12] represents the value of the 12th element in the array WBRef, and WBRef[15] represents the value of the 15th element in the array WBRef; Step S310: Create an MKSS output attribute array MKSSOutput = a floating-point array of 2 elements; The first element MKSSOutput[1] in the attribute array MKSSOutput is: MKSSOutput[1]=tanh( (MKSSTemp5-MKSSTemp6) / (MKSSTemp5+MKSSTemp6)×2 ) Among them, tanh is the calculation of hyperbolic tangent function; The second element MKSSOutput[2] in the attribute array MKSSOutput is: MKSSOutput[2]=tanh( (MKSSTemp7-MKSSTemp8) / (MKSSTemp7+MKSSTemp8)×2 ) Where tanh is the hyperbolic tangent function; Step S311: Return the attribute array MKSSOutput as the result of MKSS.

4. The corn biomass inversion method based on drone microwave radiation data according to claim 3 is characterized in that: The specific process of step S4 is as follows: Step S401: establishing a microwave penetration transformation attribute description module MKYG for the middle layer of corn plants with ears and stems, wherein the input of MKYG is the rows MKYGRow and columns MKYGCol that MKYG needs to calculate; Step S402: Create the first temporary variable MKYGTemp1 of MKYG = the average value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the first element of WBData; Step S403: Create a second temporary variable MKYGTemp2 of MKYG, which is the mean value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the third element of WBData; Step S404: Create the third temporary variable MKYGTemp3 of MKYG, which is the mean value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the fifth element of WBData; Step S405: Create the fourth temporary variable MKYGTemp4 of MKYG = the average value of the eight neighboring pixels of the pixel at the MKYGRowth row and the MKYGColth column in the image of the seventh element of WBData; Step S406: Create the fifth temporary variable MKYGTemp5 of MKYG: MKYGTemp5=(MKYGTemp1-WBRef[14]) / (WBRef[1]-WBRef[9]) Among them, WBRef[1] represents the value of the first element in the array WBRef, WBRef[9] represents the value of the ninth element in the array WBRef, and WBRef[14] represents the value of the fourteenth element in the array WBRef; Step S407: Create the sixth temporary variable MKYGTemp6 of MKYG: MKYGTemp6=(MKYGTemp2-WBRef[16]) / (WBRef[5]-WBRef[13]) Among them, WBRef[5] represents the value of the 5th element in the array WBRef, WBRef[13] represents the value of the 13th element in the array WBRef, and WBRef[16] represents the value of the 16th element in the array WBRef; Step S408: Create the seventh temporary variable MKSSTemp7 of MKYG as: MKSSTemp7=tanh(MKYGTemp5-MKYGTemp6) Step S409: Create the eighth temporary variable MKYGTemp8 of MKYG: MKYGTemp8=(MKYGTemp1-MKYGTemp2) / (MKYGTemp3-MKYGTemp4) Step S410, create an MKYG output attribute array MKYGOutput = a floating point array of 4 elements; Set the first element MKYGOutput[1] in the attribute array MKYGOutput to MKYGTemp5, the second element MKYGOutput[2] in the attribute array MKYGOutput to MKYGTemp6, the third element MKYGOutput[3] in the attribute array MKYGOutput to MKYGTemp7, and the fourth element MKYGOutput[4] in the attribute array MKYGOutput to MKYGTemp8; Step S411: Return MKYGOutput as the result of MKYG.

5. The corn biomass inversion method based on drone microwave radiation data according to claim 4 is characterized in that: The specific process of step S5 is as follows: Step S501: Establish a corn plant multi-layer penetration transformation attribute description module MKDCC, where the input of MKDCC is the row MKDCCRow and column MKDCCCol that MKDCC needs to calculate; Step S502: Establish a first-level temporary variable MKDCCTemp1 in the corn plant multi-layer penetration transformation attribute description module, access MKDX for calculation, input MKDX row MKDXRow=MKDCCRow and input MKDX column MKDXCol=MKDCCCol, and return the attribute array MKDXOutput output by MKDX; Step S503: Establish a second-level temporary variable MKDCCTemp2 in the corn plant multi-layer penetration transformation attribute description module, access the MKSS for calculation, input the MKSS row MKSSRow=MKDCCRow and the MKSS column MKSSCol=MKDCCCol, and return the attribute array MKSSOutput output by the MKSS; Step S504: Establish a third-level temporary variable MKDCCTemp3 in the corn plant multi-layer penetration transformation attribute description module, access MKYG for calculation, input MKYG row MKYGRow=MKDCCRow and MKYG column MKYGCol=MKDCCCol, and return the attribute array MKYGOutput output by MKYG; Step S505, create an MKDCC output attribute array MKDCCOutput = create a floating point array containing 10 elements; Set the first element MKDCCOutput[1] in the attribute array MKDCCOutput to MKDCCTemp1[1]; the second element MKDCCOutput[2] in the attribute array MKDCCOutput to MKDCCTemp1[2]; the third element MKDCCOutput[3] in the attribute array MKDCCOutput to MKDCCTemp1[3]; the fourth element MKDCCOutput[4] in the attribute array MKDCCOutput to MKDCCTemp1[4]; Among them, MKDCCTemp1[1] represents the first element in the temporary variable of the first level, MKDCCTemp1[2] represents the second element in the temporary variable of the first level, MKDCCTemp1[3] represents the third element in the temporary variable of the first level, and MKDCCTemp1[4] represents the fourth element in the temporary variable of the first level; Set the fifth element MKDCCOutput[5] in the attribute array MKDCCOutput to MKDCCTemp2[1]; the sixth element MKDCCOutput[6] in the attribute array MKDCCOutput to MKDCCTemp2[2]; Among them, MKDCCTemp2[1] represents the first element in the second-level temporary variable, and MKDCCTemp2[2] represents the second element in the second-level temporary variable; Set the 7th element MKDCCOutput[7] in the attribute array MKDCCOutput to MKDCCTemp3[1]; the 8th element MKDCCOutput[8] in the attribute array MKDCCOutput to MKDCCTemp3[2]; the 9th element MKDCCOutput[9] in the attribute array MKDCCOutput to MKDCCTemp3[3]; the 10th element MKDCCOutput[10] in the attribute array MKDCCOutput to MKDCCTemp3[4]; Among them, MKDCCTemp3[1] represents the first element in the temporary variable of the third level, MKDCCTemp3[2] represents the second element in the temporary variable of the third level, MKDCCTemp3[3] represents the third element in the temporary variable of the third level, and MKDCCTemp3[4] represents the fourth element in the temporary variable of the third level; Step S506: Return the attribute array MKDCCOutput as the result of MKDCC.

6. The corn biomass inversion method based on drone microwave radiation data according to claim 5 is characterized in that: The specific process of step S6 is as follows: Step S601: Input the corn biomass dataset YMList. YMList is a list containing the YMROW field, the YMCOL field, and the SWL field. The YMROW field, the YMCOL field, and the SWL field are: YMROW field: the row where the data collection point is located in the drone microwave radiation data image; YMCOL field: the column where the data collection point is located in the drone microwave radiation data image; SWL field: biomass of corn; Step S602: Create a corn biomass decision data table YMDecision containing the DCCFeature field and the DSWL field. YMDecision contains two fields: DCCFeature field: multi-layer penetration transformation attributes; DSWL field: biomass of corn used for inversion prediction; The specific process of step S602 is: Step 1: Initialize the multi-layer penetration transform attribute analysis counter DCCCounter = 1; Step 2: Create a new row of data YMDecisionRow for YMDecision; Step 3: Multi-layer penetration transform attribute description temporary variable DCCTemp1 = access MKDCC for calculation, input MKDCCRow of MKDCC row = the value of YMROW field of DCCCounter-th element in YMList, input MKDCCCol of MKDCC column = the value of YMCOL field of DCCCounter-th element in YMList, and return MKDCCOutput; Step 4. Set the value of the DCCFeature field of YMDecisionRow to DCCTemp1; Step 5. Set the value of the DSWL field of YMDecisionRow to the value of the SWL field of the DCCCounterth element in YMList; Step 6: Set DCCCounter=DCCCounter+1; Step 7: If DCCCounter is less than or equal to the total number of elements in YMList, go to step 2; otherwise, stop the iteration and obtain the corn biomass decision data table YMDecision; Step S603: Establish a multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel, where the input of the model DCCForestModel is the DCCFeature field of YMDecision, and the label of the model DCCForestModel prediction output is the DSWL field of YMDecision; Step S604: Use all the data of YMDecision to train the DCCForestModel model; Step S605, end.

7. The corn biomass inversion method based on drone microwave radiation data according to claim 6 is characterized in that: The multi-layer penetration transformation attribute inversion prediction decision forest model DCCForestModel is a data regression decision forest.

8. The corn biomass inversion method based on drone microwave radiation data according to claim 7 is characterized in that: The specific process of step S7 is as follows: Step S701, establishing corn biomass inversion prediction result FYResult = establishing a two-dimensional matrix with the number of rows being WBHeight and the number of columns being WBWidth, and initializing the value of each element in the two-dimensional matrix to be 0; Step S702, inversion row counter FYROW=1; Step S703, inversion column counter FYCOL=1; Step S704: The first temporary variable FYTemp1 used for inversion is connected to MKDCC for calculation, the row MKDCCRow=FYROW of MKDCC is input, the column MKDCCCol=FYCOL of MKDCC is input, and the output MKDCCOutput of MKDCC is returned; Step S705, the second temporary variable FYTemp2 for inversion = FYTemp1 is connected to the model DCCForestModel, and the inversion prediction result of the FYROWth row and the FYCOLth column in FYResult is obtained through the DCCForestModel model; Step S706: Set the value of the FYROWth row and the FYCOLth column of FYResult to FYTemp2; Step S707: Set FYCOL=FYCOL+1; Step S708: If FYCOL is less than or equal to WBWidth, go to step S703; otherwise, go to step S708; Step S709: set FYROW=FYROW+1; Step S710: If FYROW is less than or equal to WBHeight, go to step S702; otherwise, go to step S710. Step S711, outputting FYResult as the corn biomass inversion result; Step S712, end.

Images