A fast and effective method to determine the area of ​​crops

Abstract

The application discloses a method for rapidly and effectively judging crop area, and collects data including collecting images with certain resolution and filling in plot area S1, obtaining area S a from a classification decision tree and visually interpreting area S b with corresponding spectral characteristics, obtaining area set {S2} in combination with S1, S a and S b , layering plot area set {S2}, sampling each layer after layering, calculating sample capacity, randomly sampling sample plots for RTK actual measurement, performing precision evaluation, respectively calculating error rates of each layer, judging whether the result is within the allowable error rate, reprocessing the area set {S2}' if the result does not conform to the requirement, and verifying until the result is qualified, and finally estimating total sowing area of crops; the method combines remote sensing and sampling investigation, and improves the accuracy of plot area and plot position.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of remote sensing data measurement, in particular to a method for quickly and effectively judging crop area. BACKGROUND

[0002] Currently, there are two methods for crop sowing area statistics: one is to select survey sites in multiple stages according to the random principle within the scope of a province, city or autonomous region, and then to sample and measure the crop sowing area, and then to calculate the sowing area by stages; the other is to classify and extract crops by using remote sensing images to obtain the crop sowing area.

[0003] However, the traditional sampling and measurement method is inevitably disturbed by human factors in the data acquisition process, and problems such as misreporting and omission are difficult to avoid. Since the total population is calculated by sampling survey, the deviation in the survey process will directly affect the final calculation result. The accuracy of the sampling survey area is between 60% and 90%.

[0004] In addition, satellite remote sensing cannot accurately obtain full-coverage remote sensing images of the survey area during the crop growth period due to the influence of weather and operation cycle. In addition, the accuracy is low due to the limitation of resolution. The high cost of aerial remote sensing limits it to a small range, and it is difficult to obtain periodic aerial photographs. The uncertainty of the remote sensing image extraction technology also makes it impossible to independently complete the area measurement work with perfect statistical significance. The data accuracy of remote sensing images depends on the completeness of the image data, and the accuracy is between 50% and 85%. There is no scientific and effective method to determine the number of samples for sampling.

[0005] Finally, the traditional sampling survey relies on the data reported by the basic units, and the accuracy and authenticity cannot be guaranteed. The optical resolution of the mainstream commercial satellite image is generally 0.5 meters, and the positioning accuracy is 0.5-1 meters, which cannot identify small crop areas. Areas below 0.5 mu cannot be identified, and the accurate measurement of crop sowing area is slightly insufficient. Using RTK with centimeter-level positioning accuracy for field measurement also has the problems of too much work and being limited by the terrain of the measurement area, which cannot guarantee that all work can be completed within a specified time. SUMMARY

[0006] The present application aims to provide a method for quickly and effectively judging crop area to solve the problems in the background art.

[0007] To solve the above technical problems, the present application provides the following technical solution: a method for quickly and effectively judging crop area, comprising the following steps:

[0008] Step S100: collecting data to be processed, the data to be processed including panoramic images of the crop area and distinguishing features of the crops; processing the panoramic images of the crop area to obtain crop plot images;

[0009] Step S200: Sampling and stratifying the crop plot images in step S100, calculating the sample size of each sampling layer, and randomly selecting sample plots from the sample size for RTK measurement to obtain RTK measurement values;

[0010] Step S300: Based on the RTK measurement values ​​obtained in step S200, the error rate, overall error and square root error of each sampling layer are calculated respectively;

[0011] Step S400: Determine whether the area of ​​crops is qualified based on the error rate and square root difference of step S300. If qualified, estimate the total sowing area of ​​each sampled layer. If unqualified, return to step S100 to process the data until it is qualified, and then estimate the total sowing area of ​​each sampled layer.

[0012] Furthermore, step S100 specifically includes:

[0013] Step S110: Collect crop plot images of a certain resolution, send the crop plot images to the grassroots units, and the grassroots units outline the crop plot range on the crop plot images, and record the area of ​​the crop plot range as S1;

[0014] Step S120: The distinguishing features of crops include the key period for crop identification. Multiple images of the key period for crop identification are collected as an image set. The image set is based on the spectral features of the crops, and a classification decision tree is constructed for the crops in the image set. The area of ​​the crops in the image set is extracted as S. a ;

[0015] Step S130: Visually interpret the crop image information in step S120 to obtain an area S b ;

[0016] Step S140: Based on the area S1 of the crop plot in step S110 and the area S obtained by the classification decision tree in step S120 a The area S obtained by visual interpretation in step S130 b , draw the boundaries of all crop plots on the preset image base map, and obtain the area set of all crops {S2};

[0017] Visual interpretation requires little equipment and is simple and convenient. It can obtain the required information from remote sensing images at any time, reducing the waste of a large amount of manpower and material resources in the early stage of data preparation. Visual interpretation can effectively identify the boundaries of crop plots based on human experience and knowledge through the color and shape characteristics of the image and interpretation marks. It is an irreplaceable part of remote sensing applications.

[0018] Moreover, the accuracy of area extraction from remote sensing images is improved by combining the area of ​​manually drawn plots, the area extracted by building a classification decision tree, and the area obtained by visual interpretation.

[0019] Furthermore, in step S200, sampling stratification is performed to stratify the plots according to the area set {S2}. The specific stratification process is as follows:

[0020] The plots are divided into multiple sampling layers according to their size, including small plot layer, medium plot layer and large plot layer;

[0021] Among them, when the area of ​​a plot is less than the first area threshold, the plot is a small plot, and the small plot is stored in the small plot layer {G1}, and the total area of ​​the plots in the small plot layer S is estimated. 总1 When the area of ​​a plot is greater than the first area threshold and less than the second area threshold, the plot is a medium-sized plot. The medium-sized plot is stored in the medium-sized plot layer {G2} and the total area of ​​the plots in the medium-sized plot layer S is estimated. 总2 When the area of ​​a plot is greater than the second area threshold, the plot is considered a large plot. The large plot is stored in the large plot layer {G3} and the total area of ​​the plots in the large plot layer S is estimated. 总3 , and the first area threshold is smaller than the second area threshold.

[0022] Stratifying plots is used to ensure data integrity and reliability, without neglecting smaller plots. RTK measurements improve accuracy, and stratified sampling can improve sample representativeness. Taking into account the differences between small, medium, and large plots, the accuracy of total area estimates can be improved overall. In addition, sampling grouping is flexible and convenient to implement and easy to organize.

[0023] Furthermore, the sample size calculation process in step S200 is as follows:

[0024] Step S210: Determine whether there is crop plot area data from previous years, if not, go to step S220, if yes, go to step S250;

[0025] Step S220: randomly selecting 30 or more plots from each sampling layer as initial plot samples and performing RTK measurement to obtain the sample measured area, where n≥30;

[0026] Step S230: Calculate the initial sample standard deviation of each sampling layer based on the measured sample area in step S210. If the initial sample standard deviation is qualified, the area of ​​each sampling layer is qualified, and the sample standard deviation is used as the estimated value of the population standard deviation of each sampling layer. The calculation formula is as follows:

[0027]

[0028] Among them, GS1, 2, 3 i are the areas of the i-th sample plot in the small plot layer, medium plot layer, and large plot layer, respectively. i refers to the sample plot in each plot layer. GS1i, is the area of ​​the i-th sample plot in the small plot layer GS1, GS2i, is the area of ​​the i-th sample plot in the medium plot layer GS2, and GS3i, is the area of ​​the i-th sample plot in the large plot layer GS3; GR1,2,3 i are the RTK measured areas of the i-th sample plot in the small plot layer, medium plot layer, and large plot layer, respectively;

[0029] Step S240: Calculate the sample size of each sampling layer using the formula:

[0030]

[0031] Where k is the quantile value of the standard normal distribution for a certain confidence level, e is the error limit, and N is the population size;

[0032] Step S250: Use the standard deviation of the area of ​​each layer of crop plots in previous years as the estimated standard deviation when calculating the sampling sample size this year; use the estimated standard deviation as the estimated value of the overall standard deviation of each sampling layer in step S230, and substitute it into step S230 for calculation.

[0033] When processing data, the required data is divided into two situations: one with crop plot area data from previous years and the other without crop plot area data from previous years. This can effectively reduce the heavy workload of calculating all crop areas, and can also provide a solid basis for subsequent related work, further improving the accuracy of the data.

[0034] Furthermore, in step S200, sample plots are randomly selected from the sample volume for RTK measurement to obtain RTK measurement values. The specific process is as follows:

[0035] The RTK measurement values ​​of each sampling layer are recorded as {R1}, {R2}, and {R3}. The total area of ​​RTK measurement of each sampling layer is recorded as R1, R2, and R3 respectively; the area of ​​each plot sample in {G1} is recorded as GR1 1...n , the area of ​​each plot sample in {G2} is recorded as GR21...n , the area of ​​each plot sample in {G3} is recorded as GR3 1...n .

[0036] The area values ​​of the internal processing corresponding to {G1}, {G2} and {G3} are recorded as {I1}, {I2}, {I3} respectively, and the overall area is recorded as I1, I2, I3 respectively; and the area of ​​each plot sample in the internal processing corresponding to {G1} is recorded as GS1 1...n , the area of ​​each plot sample in the internal processing corresponding to {G2} is recorded as GS2 1...n , the area of ​​each plot sample in the internal processing corresponding to {G3} is recorded as GS3 1...n .

[0037] Furthermore, the error rate, overall error, and square root error of each layer of data in step S300 are calculated as follows:

[0038] Step S310: Calculate the error rate of {G1}, {G2}, {G3}. The error rate of a certain sampling layer = (the estimated area value S of the sampling layer) 总i -The RTK measured area R of the sampling layer i ) / RTK measured area of ​​the sampling layer R i ; where i = {1, 2, 3}, the table represents the small plot layer, medium plot layer and large plot layer;

[0039] Step S320: Calculate the overall error, square root difference and coefficient of variation of each sampling layer data respectively, using the formula:

[0040] G 1,2,3 Overall error E 1,2,3 =(I 1,2,3 -R 1,2,3 ) / R 1,2,3

[0041] G 1,2,3 Square root difference

[0042] G 1,2,3 The coefficient of variation

[0043] Among them I 1,2,3 is the total area of ​​each layer processed by the internal work, R 1,2,3 is the total area of ​​the RTK measured areas of each sampling layer, n 1,2,3 is the sample size of each sampling stratum.

[0044] Calculating the overall error is helpful to serve as a criterion for judging whether the area of ​​the sampled plot is qualified. The coefficient of variation based on this is further compared. When the data measurement scales differ too much or the data dimensions are different, the use of the coefficient of variation for comparison does not require the reference data average value, which can eliminate the influence of different units or averages on the degree of variation of each sampling layer.

[0045] Furthermore, the specific process of determining the crop area in step S400 is as follows:

[0046] Step S410: E 1,2,3 The absolute value of the error rate is compared with the standard value of the error rate. The absolute value of the coefficient of variation error rate was compared with the standard value;

[0047] Step S420: If E exists 1,2,3 The absolute values ​​of the error rates are all less than the standard values ​​and If the absolute values ​​of the above are all smaller than the standard value of the error rate of the coefficient of variation, then the area of ​​the sampling layer is determined to be available; otherwise, the area of ​​the sampling layer is unavailable and the internal work needs to be reprocessed. Suppose the new area set obtained by the internal work is {S2}', and the new area set {S2}' is subjected to the process from step S200 to step S400 until the overall error of each sampling layer in the new area set {S2}' is smaller than the standard value of the error rate and the coefficient of variation is smaller than the standard value of the error rate of the coefficient of variation.

[0048] The standard error rate is used as the basis for judgment, and unqualified ones will be reprocessed. This process increases the accuracy of the area set, making the data preparation error before estimating the total sown area smaller, and the estimated actual overall area will be more accurate.

[0049] Furthermore, the specific algorithm for estimating the total sown area of ​​crops in step S400 is as follows:

[0050]

[0051] Among them, S 总i is the total area of ​​the i-th layer of each sampling layer, i={1,2,3}, S 总 It is the total sown area of ​​crops.

[0052] Compared with the existing technology, the present invention has the following beneficial effects: it combines the advantages of remote sensing and sampling surveys, uses spatial sampling technology to measure crop planting areas based on remote sensing data resources, and greatly improves the accuracy of plot area and plot location; in the data preparation stage, it verifies and supplements data from multiple sources, avoids the problems of over-reporting, misreporting, and omission of reports by grassroots units, and the accuracy and authenticity of areas, compensates for the problem of poor image quality caused by weather, and greatly improves the fault tolerance rate; adopts a stratified sampling method, fully considering the problem of different errors in the estimation of plots of different sizes;

[0053] It also addresses the problems of remote sensing images being unable to distinguish small-area features and the positioning accuracy of the images themselves; the issue of determining the sample size for crop sampling surveys; the difficulty of measuring terrain in the survey area due to the use of RTK measurements, and the inability of the time schedule to meet the project's time requirements; and the improvement of the accuracy of crop planting area statistics from 60% to 90% in traditional sampling surveys and 50% to 85% in remote sensing image identification to over 98%. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0055] Figure 1 It is a flow chart of a method for quickly and effectively determining the area of ​​crops according to the present invention; DETAILED DESCRIPTION

[0056] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0057] See also Figure 1 The present invention provides a technical solution: a method for quickly and effectively determining the area of ​​crops, comprising the following steps:

[0058] Step S100: collecting data to be processed, the data to be processed including panoramic images of the crop area and distinguishing features of the crops; processing the panoramic images of the crop area to obtain crop plot images;

[0059] Step S110: Collect crop plot images of a certain resolution and send them to grassroots units, including village committees and community workers. The grassroots units outline the crop plot range on the crop plot images, and record the area of ​​the crop plot range as S1;

[0060] Step S120: The distinguishing features of crops include the key identification period of crops. Multiple images of the key identification period of crops are collected and set as an image set. The image set is judged according to the phenological calendar of crops. The image set is based on the spectral characteristics of crops, and a classification decision tree is constructed for the crops in the image set. The area of ​​crops in the image set is extracted as S a ; The classification decision tree is explained based on the algorithm structure and algorithm theory, and then these decision tree algorithms are used to conduct remote sensing crop cover classification experiments;

[0061] Step S130: Visually interpret the crop image information in step S120 to obtain an area S b ,Visual interpretation is based on the different color features of different crops presented in true color images, false color images and NDVI;

[0062] Step S140: Based on the area S1 of the crop plot in step S110 and the area S obtained by the classification decision tree in step S120 a The area S obtained by visual interpretation in step S130 b , draw the boundaries of all crop plots on a preset image base map, and obtain the area set of all crops {S2}. The preset image base map is an image base map obtained by the China Land Observation Satellite Data Center or purchased commercially;

[0063] A phenological calendar, also known as a natural calendar or agricultural calendar, compiles years of observational data on a region's natural phenology, crop phenology, pest outbreaks, and agricultural activities, and arranges them into tables by date of occurrence to help track farming seasons. Using a phenological calendar as a basis for identifying key crop periods provides a reliable basis and also makes crop image data more accurate. NDVI, the Normalized Difference Vegetation Index, is a remote sensing image that measures the difference between the reflectance values ​​in the near-infrared band and the red band divided by the sum of the two. The NDVI is a key parameter reflecting crop growth and nutritional information.

[0064] Visual interpretation requires little equipment and is simple and convenient. It can obtain the required information from remote sensing images at any time, reducing the waste of a large amount of manpower and material resources in the early stage of data preparation. Visual interpretation can effectively identify the boundaries of crop plots based on human experience and knowledge through the color and shape characteristics of the image and interpretation marks. It is an irreplaceable part of remote sensing applications.

[0065] Moreover, the accuracy of area extraction from remote sensing images is improved by combining the area of ​​manually drawn plots, the area extracted by building a classification decision tree, and the area obtained by visual interpretation.

[0066] Step S200: Sampling and stratifying the crop plot images in step S100, calculating the sample size of each sampling layer, and randomly selecting sample plots from the sample size for RTK measurement to obtain RTK measurement values;

[0067] In step S200, sampling stratification is performed to stratify the plots according to the area set {S2}. The specific stratification process is as follows:

[0068] The plots are divided into multiple sampling layers according to their size, including small plot layer, medium plot layer and large plot layer;

[0069] Among them, when the area of ​​a plot is less than the first area threshold, the plot is a small plot, and the small plot is stored in the small plot layer {G1}, and the total area of ​​the plots in the small plot layer S is calculated. 总1 When the area of ​​a plot is greater than the first area threshold and less than the second area threshold, the plot is a medium-sized plot. The medium-sized plot is stored in the medium-sized plot layer {G2} and the total area of ​​the plots in the medium-sized plot layer S is calculated. 总2 When the area of ​​a plot is greater than the second area threshold, the plot is a large plot, and the large plot is stored in the large plot layer {G3}, and the total area of ​​the plots in the large plot layer S is calculated. 总3 , and the first area threshold is smaller than the second area threshold.

[0070] Stratifying plots is used to ensure data integrity and reliability, without neglecting smaller plots. RTK measurements improve accuracy, and stratified sampling can improve sample representativeness. Taking into account the differences between small, medium, and large plots, the accuracy of total area estimates can be improved overall. In addition, stratified sampling is flexible and convenient to implement and easy to organize.

[0071] The sample size calculation process in step S200 is as follows:

[0072] Step S210: Determine whether there is crop plot area data from previous years, if not, go to step S220, if yes, go to step S250;

[0073] Step S220: randomly selecting 30 or more plots from each sampling layer as initial plot samples and performing RTK measurement to obtain the sample measured area, where n≥30;

[0074] Step S230: Calculate the initial sample standard deviation of each sampling layer according to the sample area in step S210. If the initial sample standard deviation is qualified, the area of each sampling layer is qualified, and the sample standard deviation is taken as the estimation value of the population standard deviation of each sampling layer. The calculation formula is as follows:

[0075]

[0076] GS1,2,3 i are the areas of the i-th sample plot in the small plot layer, the medium plot layer and the large plot layer respectively, i refers to the sample plot in each plot layer, GS1i is the area of the i-th sample plot in the small plot layer GS1, GS2i is the area of the i-th sample plot in the medium plot layer GS2, and GS3i is the area of the i-th sample plot in the large plot layer GS3; GR1,2,3 i are the RTK measured areas of the i-th sample plot in the small plot layer, the medium plot layer and the large plot layer respectively;

[0077] The estimation value of the population standard deviation of the small plot layer is:

[0078] The estimation value of the population standard deviation of the medium plot layer is:

[0079] The estimation value of the population standard deviation of the large plot layer is:

[0080] Step S240: Calculate the sample size of each sampling layer using the formula:

[0081]

[0082] where k is the quantile value of the standard normal distribution for a certain confidence level, e is the error limit, and N is the population size.

[0083] The sample size calculation formula for the small sampling layer is:

[0084] The sample size calculation formula for the medium sampling layer is:

[0085] The sample size calculation formula for the large sampling layer is:

[0086] Step S250: Take the area standard deviation of each layer of the previous year's crop as the standard deviation estimation value for this year's sampling sample size calculation; take the standard deviation estimation value as the estimation value of the population standard deviation of each sampling layer in step S230, and substitute it into step S230 for calculation.

[0087] The error margin e is a multiple of the standard error, and the multiplier factor depends on the desired level of confidence in the estimate in the survey, i.e. where k is the quantile value of the standard normal distribution corresponding to a certain confidence level, which can be obtained from the standard normal distribution table.

[0088] In the process of data, the required data is divided into two cases of having and not having the area data of the previous year's crop land, which can effectively reduce the heavy workload of calculating all crop areas, provide a basis for subsequent related work, and further improve the accuracy of data.

[0089] In step S200, a sample plot is randomly selected from the sample size for RTK measurement, and an RTK measurement value is obtained, and the specific process is as follows:

[0090] The RTK measurement values of each sampling layer are recorded as {R1}, {R2}, and {R3}, and the RTK measurement total areas of each layer are recorded as R1, R2, and R3, respectively. The area of each plot sample in {G1} is recorded as GR1 1...n , the area of each plot sample in {G2} is recorded as GR2 1...n , and the area of each plot sample in {G3} is recorded as GR3 1...n .

[0091] The area values corresponding to {G1}, {G2}, and {G3} in the office processing are recorded as {I1}, {I2}, and {I3}, respectively, and the total areas are recorded as I1, I2, and I3, respectively. The area of each plot sample in {G1} corresponding to the office processing is recorded as GS1 1...n , the area of each plot sample in {G2} corresponding to the office processing is recorded as GS2 1...n , and the area of each plot sample in {G3} corresponding to the office processing is recorded as GS3 1...n .

[0092] Step S300: Based on the RTK measurement value obtained in step S200, the error rate, total error, and square root difference of each layer are calculated respectively.

[0093] RTK real-time dynamic measurement technology is a real-time differential GPS technology based on carrier phase observation. It is a breakthrough in the development of measurement technology. It consists of three parts: base station receiver, data link, and mobile station receiver. A receiver is placed on the base station as a reference station to continuously observe the satellite, and its observation data and station information are sent to the mobile station in real time through radio transmission equipment. While receiving GPS satellite signals, the mobile station GPS receiver receives the data transmitted by the base station through wireless receiving equipment, and then calculates the three-dimensional coordinates and accuracy of the mobile station in real time based on the principle of relative positioning. It is divided into radio mode and network communication mode.

[0094] Step S310: Calculate the error rate of {G1}, {G2}, {G3}, the error rate of a certain sampling layer = (the estimated area value S of the sampling layer 总i -The RTK measured area R of the sampling layer i ) / RTK measured area of ​​the sampling layer R i ; where i = {1, 2, 3}, the table represents small plots, medium plots and large plots;

[0095] For example, the error rate of the small plot layer is calculated as follows: (S 总1 -R1) / R1; Error rate of medium-sized plot layer = (S 总2 -R2) / R2; Error rate of large plot layer = (S 总3 -R3) / R3;

[0096] Step S320: Calculate the overall error, square root difference and coefficient of variation of each layer of sampled data respectively, using the formula:

[0097] G 1,2,3 Overall error E 1,2,3 =(I 1,2,3 -R 1,2,3 ) / R 1,2,3

[0098] For example, the overall error of the small plot layer {G1} is E1 = (I1-R1) / R1; the overall error of the medium plot layer {G2} is E2 = (I2-R2) / R2; the overall error of the large plot layer {G3} is E3 = (I3-R3) / R3;

[0099] G 1,2,3 Square root difference

[0100] For example, the square root difference of the small plot layer {G1} Square root difference of medium-sized plot layer {G2} Square root difference of large plot layer {G3}

[0101] G 1,2,3 The coefficient of variation

[0102] Coefficient of variation of small plot layer {G1} Coefficient of variation of medium-sized plot layer {G2} Coefficient of variation of large plot layer {G3}

[0103] Among them I 1,2,3 is the total area of ​​each layer processed by the internal work, R 1,2,3 is the total area of ​​the RTK measured areas of each sampling layer, n 1,2,3 is the sample size of each sampling stratum.

[0104] Calculating the overall error is helpful to serve as a criterion for judging whether the area of ​​the sampled plot is qualified. The coefficient of variation based on this is further compared. When the data measurement scales differ too much or the data dimensions are different, the use of the coefficient of variation for comparison does not require the reference data average value, which can eliminate the influence of different units or averages on the degree of variation of each sampling layer.

[0105] Step S400: Determine whether the area of ​​crops is qualified based on the error rate and square root difference of step S300. If qualified, estimate the total sowing area of ​​each sampled layer. If unqualified, return to step S100 to process the data until it is qualified, and then estimate the total sowing area of ​​each sampled layer.

[0106] Step S410: E 1,2,3 The absolute value of the error rate is compared with the standard value of the error rate. The absolute value of the coefficient of variation error rate was compared with the standard value;

[0107] Step S420: If E exists 1,2,3 The absolute values ​​of the error rates are all less than the standard values ​​and If the absolute values ​​of the values ​​of {S2}' and {S2}' are all less than the standard value of the error rate of the coefficient of variation, then the area of ​​the sampling layer is determined to be usable; otherwise, the area of ​​the sampling layer is unusable and needs to be reprocessed. Suppose the new area set obtained by the internal processing is {S2}', and the new area set {S2}' is subjected to the process from step S200 to step S400 again until the overall error of each sampling layer in the new area set {S2}' is less than the standard value of the error rate and the coefficient of variation is less than the standard value of the error rate of the coefficient of variation;

[0108] The standard error rate is used as the basis for judgment, and unqualified ones will be reprocessed. This process increases the accuracy of the area set, making the data preparation error before estimating the total sown area smaller, and the estimated actual overall area will be more accurate.

[0109] The estimation of the total sowing area of the crops in step S400 is as follows:

[0110]

[0111] wherein, S 总i is the total area of the i-th layer of the sample layers, i = {1, 2, 3}, S 总 is the total sowing area of the crops.

[0112] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one entity or action from another, without necessarily requiring or implying any such actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0113] Finally, it should be noted that the above only represents the preferred embodiments of the present application, and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, modifications or equivalent replacements to the technical solutions described in the foregoing embodiments can still be made by those skilled in the art. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for quickly and effectively determining the area of ​​crops, characterized in that: The following steps are involved: Step S100: collecting data to be processed, wherein the data to be processed includes a panoramic image of a crop area and distinguishing features of the crops; Processing the panoramic image of the crop area to obtain a crop plot image; Step S100 specifically includes: Step S110: collecting crop plot images of a certain resolution, and sending the crop plot images to the grassroots units. The grassroots units outline the crop plot range on the crop plot images, and record the area of ​​the crop plot range as S1; Step S120: The distinguishing features of the crops include the key identification period of the crops. Multiple images of the key identification period of the crops are collected as an image set. A classification decision tree is constructed for the crops in the image set, and the area of ​​the crops in the image set is extracted as S. a ; Step S130: Visually interpret the image information of the crops in step S120 to obtain an area S b ; Step S140: Based on the area S1 of the crop plot in step S110 and the area S obtained by the classification decision tree in step S120 a The area S obtained by visual interpretation in step S130 b , draw the boundaries of all crop plots on the preset image base map, and obtain the area set of all crops {S2}; Step S200: Sampling and stratifying the crop plot images in step S100, calculating the sample size of each sampling layer, and randomly selecting sample plots from the sample size for RTK measurement to obtain RTK measurement values; In step S200, sampling stratification is performed, and the specific stratification process is as follows: The plots are divided into multiple sampling layers according to the size of the plots, and the sampling layers include a small plot layer, a medium plot layer and a large plot layer; Among them, when the area of ​​a plot is less than the first area threshold, the plot is a small plot, and the small plot is stored in the small plot layer {G1}, and the total area of ​​the plots in the small plot layer S is estimated. 总1 When the area of ​​a plot is greater than the first area threshold and less than the second area threshold, the plot is a medium-sized plot. The medium-sized plot is stored in the medium-sized plot layer {G2} and the total area of ​​the plots in the medium-sized plot layer S is estimated. 总2 When the area of ​​a plot is greater than the second area threshold, the plot is considered a large plot. The large plot is stored in the large plot layer {G3} and the total area of ​​the plots in the large plot layer S is estimated. 总3 , and the first area threshold is smaller than the second area threshold; The sample size calculation process of step S200 is as follows: Step S210: Determine whether there is crop plot area data from previous years, if not, go to step S220, if yes, go to step S250; Step S220: randomly selecting 30 or more plots from each sampling layer as initial plot samples and performing RTK measurement to obtain the sample measured area, where n≥30; Step S230: Calculate the initial sample standard deviation of each sampling layer based on the measured sample area in step S210. If the initial sample standard deviation is qualified, the area of ​​each sampling layer is qualified, and the sample standard deviation is used as the estimated value of the population standard deviation of each sampling layer. The calculation formula is as follows: Among them, GS1, 2, 3 i are the areas of the i-th sample plot in the small plot layer, medium plot layer and large plot layer, GR1, 2, 3 i are the RTK measured areas of the i-th sample plot in the small plot layer, medium plot layer, and large plot layer, respectively; Step S240: Calculate the sample size of each sampling layer using the formula: Where k is the quantile value of the standard normal distribution for a certain confidence level, e is the error margin, and N is the population size; Step S250: Using the standard deviation of the area of ​​each crop layer in previous years as the estimated standard deviation for calculating the sample size for this year; using the estimated standard deviation as the estimated value of the overall standard deviation of each sampling layer in step S230, and substituting it into step S230 for calculation; In step S200, sample plots are randomly selected from the sample volume for RTK measurement to obtain RTK measurement values. The specific process is as follows: The RTK measurement values ​​of each sampling layer are recorded as {R1}, {R2}, and {R3}. The total area of ​​the RTK measurement of each sampling layer is recorded as R1, R2, and R3 respectively; the area of ​​each plot sample in {G1} is recorded as GR1 1...n , the area of ​​each plot sample in {G2} is recorded as GR2 1...n , the area of ​​each plot sample in {G3} is recorded as GR3 1...n ; The area values ​​of the internal processing corresponding to {G1}, {G2} and {G3} are recorded as {I1}, {I2}, {I3} respectively, and the overall area is recorded as I1, I2, I3 respectively; and the area of ​​each plot sample in the internal processing corresponding to {G1} is recorded as GS1 1...n , the area of ​​each plot sample in the internal processing corresponding to {G2} is recorded as GS2 1...n , the area of ​​each plot sample in the internal processing corresponding to {G3} is recorded as GS3 1...n ; Step S300: Based on the RTK measurement values ​​obtained in step S200, the error rate, overall error and square root error of each sampling layer are calculated respectively; Step S400: Determine whether the area of ​​crops is qualified based on the error rate and square root difference of step S300. If qualified, estimate the total sowing area of ​​each sampled layer. If unqualified, return to step S100 to process the data until it is qualified, and then estimate the total sowing area of ​​each sampled layer.

2. The method for quickly and effectively determining the crop area according to claim 1, characterized in that: The specific calculation process of the error rate, overall error and square root difference of each layer of data in step S300 is as follows: Step S310: Calculate the error rate of {G1}, {G2}, {G3}. The error rate of a certain sampling layer = (the estimated area value S of the sampling layer) 总i -The RTK measured area R of the sampling layer i ) / RTK measured area of ​​the sampling layer R i ; where i = {1, 2, 3}, the table represents the small plot layer, medium plot layer and large plot layer; Step S320: Calculate the overall error, square root difference and coefficient of variation of the data of each sampling layer respectively, using the formula: G 1,2,3 Overall error E 1,2,3 =(I 1,2,3 -R 1,2,3 ) / R 1,2,3 G 1,2,3 Square root difference G 1,2,3 Coefficient of variation CV 1,2,3 =σG 1,2,3 / R 1,2,3 *n 1,2,3 Among them I 1,2,3 is the total area of ​​each layer processed by the internal work, R 1,2,3 is the total area of ​​the RTK measured areas of each sampling layer, n 1,2,3 is the sample size of each sampling stratum.

3. The method for quickly and effectively determining the crop area according to claim 2, characterized in that: The specific process of determining the crop area in step S400 is as follows: Step S410: E 1,2,3 The absolute value of the error rate is compared with the standard value of the error rate. The absolute value of the coefficient of variation error rate was compared with the standard value; Step S420: If there is E 1,2,3 The absolute values ​​of the error rates are all less than the standard value and If the absolute values ​​of the above are all smaller than the standard value of the error rate of the coefficient of variation, then the area of ​​the sampling layer is determined to be available; otherwise, the area of ​​the sampling layer is unavailable and needs to be reprocessed. Suppose the new area set obtained by the internal processing is {S2}', and the new area set {S2}' is subjected to the process from step S200 to step S400 until the overall error of each sampling layer in the new area set {S2}' is smaller than the standard value of the error rate and the coefficient of variation is smaller than the standard value of the error rate of the coefficient of variation.

4. The method for quickly and effectively determining the crop area according to claim 3, characterized in that: The specific algorithm for estimating the total sown area of ​​crops in step S400 is as follows: Among them, S 总i is the total area of ​​the i-th layer of the sampling layers, i = {1, 2, 3}, S 总 It is the total sown area of ​​crops.

Images