Farmland non-grain monitoring method based on remote sensing data

Abstract

The invention discloses a farmland non-grain monitoring method based on remote sensing data. The monitoring method comprises the following steps: acquiring geographic data of a target monitoring area; acquiring a crop growth characteristic time sequence corresponding to the target monitoring area based on the geographic data of the target monitoring area; acquiring a plurality of remote sensing images of the target monitoring area based on the crop growth characteristic time sequence; determining a plurality of remote sensing image sets based on the plurality of remote sensing images; each remote sensing image set comprises a plurality of remote sensing images of the target monitoring area; and acquiring cultivated land use information of the target monitoring area according to the remote sensing image set one by one. The technical problems of low timeliness and low accuracy in the farmland non-grain monitoring process in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of land monitoring, and particularly relates to a cultivated land non-food monitoring method based on remote sensing data. BACKGROUND

[0002] Cultivated land is a strategic resource for food security, but with the acceleration of urbanization and the adjustment of agricultural structure, the problem of cultivated land non-food (changing to economic crops or digging ponds for aquaculture) is becoming increasingly prominent.

[0003] Traditional manual patrol methods are high in cost and low in efficiency; satellite remote sensing technology can achieve rapid, dynamic and accurate cultivated land monitoring.

[0004] At present, the existing technology usually adopts manual visual methods or general change detection algorithms to interpret and analyze the remote sensing images of cultivated land to determine whether the cultivated land appears non-food or non-agricultural. However, the process of cultivated land non-food or non-agricultural is dynamic. For the process of cultivated land non-food, the current monitoring method is low in timeliness and difficult to meet the demand for high-frequency and rapid dynamic change monitoring of cultivated land. SUMMARY

[0005] The present application aims to provide a cultivated land non-food monitoring method based on remote sensing data, which solves the technical problem of low timeliness and accuracy in the monitoring process of cultivated land non-food in the prior art.

[0006] The present application provides a cultivated land non-food monitoring method based on remote sensing data, which comprises:

[0007] obtaining geographical data of a target monitoring area;

[0008] based on the geographical data of the target monitoring area, obtaining a crop growth characteristic time sequence corresponding to the target monitoring area;

[0009] based on the crop growth characteristic time sequence, collecting a plurality of remote sensing images of the target monitoring area;

[0010] based on the plurality of remote sensing images, confirming a plurality of remote sensing image sets; each remote sensing image set comprises a plurality of remote sensing images of the target monitoring area;

[0011] obtaining cultivated land use information of the target monitoring area according to the remote sensing image sets one by one.

[0012] Further, the geographical data comprises geographical coordinate data, water source condition data, altitude data and illumination data; based on the geographical data of the target monitoring area, the crop growth characteristic time sequence corresponding to the target monitoring area is obtained, comprising:

[0013] Based on the geographic coordinate data, obtain historical planting data and agricultural purchase data in the target monitoring area and construct a crop variety database;

[0014] Based on the water source condition data, elevation data and light data of the target monitoring area, obtain the crop variety constraint condition;

[0015] Based on the crop variety constraint condition, screen in the crop variety database to obtain a plurality of plantable crop varieties corresponding to the target monitoring area, and a plurality of characteristic growth stages corresponding to each plantable crop variety; the plantable crop varieties include food crops and non-food crops;

[0016] Based on the plurality of characteristic growth stages corresponding to the plantable crop varieties, confirm the characteristic acquisition time sequence corresponding to the plantable crop varieties; the characteristic acquisition time sequence includes a plurality of acquisition time points; the plurality of acquisition time points correspond one-to-one to the plurality of characteristic growth stages;

[0017] Based on the characteristic acquisition time sequence of the plurality of plantable crop varieties, obtain a crop growth characteristic time sequence; each acquisition time point in the crop growth characteristic time sequence is provided with a crop label of the corresponding plantable crop variety.

[0018] Further, based on the crop growth characteristic time sequence, a plurality of remote sensing images of the target monitoring area are acquired, including:

[0019] Based on each acquisition time point in the characteristic acquisition time sequence, a plurality of acquisition time windows are obtained; each acquisition time window includes at least one acquisition time point;

[0020] Based on a plurality of acquisition time points in the same acquisition time window, a new image acquisition task is created; each image acquisition task includes an image acquisition range, an image quality requirement and a plurality of crop labels;

[0021] According to the image acquisition task, remote sensing images of the target monitoring area are obtained and labeled with crop labels; each remote sensing image is labeled with at least one crop label.

[0022] Further, based on the plurality of remote sensing images, a plurality of remote sensing image sets are confirmed, including:

[0023] Based on the plurality of plantable crop varieties of the target area, a plurality of image empty sets are constructed and each image empty set is provided with set parameters; the set parameters include crop labels and set capacities;

[0024] When an image acquisition task is completed, the remote sensing images are stored in at least one corresponding image empty set according to the crop labels of the remote sensing images;

[0025] When the number of remote sensing images in an image empty set reaches the corresponding set capacity, the remote sensing image set is obtained.

[0026] Further, based on the several plantable crop varieties of the target region, several image null sets are constructed and each image null set is set with a set parameter, including:

[0027] Based on the several plantable crop varieties of the target region, the number of plantable crop varieties is obtained, and the crop label and the number of characteristic growth stages of each plantable crop variety are obtained;

[0028] Based on the number of plantable crop varieties, several image null sets are constructed, and each image null set is labeled one by one based on the several crop labels;

[0029] Based on the number of characteristic growth stages corresponding to the crop label of the image null set and the lower limit of the preset capacity, the set capacity of the image null set is set.

[0030] Further, the cultivated land use information of the target monitoring region is obtained one by one according to the remote sensing image set, including:

[0031] Obtain the plot division information of the target monitoring region;

[0032] Whenever a remote sensing image set is generated, each remote sensing image in the remote sensing image set is divided based on the plot division information, and several plot remote sensing images are obtained;

[0033] Based on the several plot remote sensing images of the same plot, a remote sensing image subset is constructed and labeled with a crop label; the crop label of each remote sensing image subset is consistent with the crop label of the remote sensing image set;

[0034] The crop analysis is performed on each plot corresponding remote sensing image subset, and whether the actual planted crop of each plot is consistent with the crop label is judged one by one;

[0035] Obtain a plurality of plots whose actual planted crops are consistent with the crop label, and determine whether the plurality of plots are grain plots according to the crop label and output a monitoring result.

[0036] Further, the crop analysis is performed on each plot corresponding remote sensing image subset, and whether the actual planted crop of each plot is consistent with the crop label is judged one by one, including:

[0037] Based on the crop label, a crop analysis model is obtained;

[0038] Image processing is performed on each plot remote sensing image in the plot corresponding remote sensing image subset to obtain image feature information;

[0039] The image feature information is input into the crop analysis model to obtain an analysis result; the analysis result includes that the actual planted crop is consistent with the crop label, or the actual planted crop is not consistent with the crop label.

[0040] Further, based on the crop marker, a crop analysis model is obtained, comprising:

[0041] A feature relationship database corresponding to the crop marker is obtained; the feature relationship database comprises image feature information, analysis results, and a mapping relationship between the image feature information and the analysis results;

[0042] The neural network model is trained through the feature relationship database to obtain the crop analysis model.

[0043] Further, the cultivated land use information of the target monitoring area is obtained according to the remote sensing image set one by one, and further comprising:

[0044] A number of plots where the actual planted crops do not match the crop marker are obtained, and when the next remote sensing image set is generated, a corresponding remote sensing image subset is obtained for crop analysis.

[0045] Further, the cultivated land use information of the target monitoring area is obtained according to the remote sensing image set one by one, and further comprising:

[0046] The crop maturity time is recorded for the plot where the actual planted crops match the crop marker;

[0047] Based on the remote sensing image set generated after the crop maturity time, a remote sensing image subset of the plot is obtained and crop analysis is performed.

[0048] Compared with the prior art, the beneficial effects of the present application are:

[0049] In the present application, the crop growth feature time sequence is obtained through the geographic data of the target monitoring area, and the remote sensing image is collected according to the crop growth feature time sequence; and the remote sensing image set is generated according to a number of remote sensing images so as to analyze according to a plurality of remote sensing images in the remote sensing image set, thereby further obtaining the cultivated land use information of the target monitoring area. Avoid the situation of inaccurate and untimely analysis caused by single remote sensing image. The technical problem of low timeliness and accuracy in the monitoring process of the existing technology of non-grain cultivation of cultivated land is solved. BRIEF DESCRIPTION OF DRAWINGS

[0050] Figure 1 The method steps of the present application are a cultivated land non-grain monitoring method based on remote sensing data. DETAILED DESCRIPTION

[0051] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0052] As shown in Figure 1 A farmland non-grain monitoring method based on remote sensing data, the monitoring method comprises:

[0053] S1: acquiring geographic data of a target monitoring area;

[0054] In the embodiment, the target monitoring area includes at least one land plot; for hilly areas or densely populated areas, there are multiple farmland land plots in the same target monitoring area, and the multiple farmland land plots belong to different persons, resulting in different crops planted in each land plot and greater difficulty in farmland non-grain management.

[0055] S2: based on the geographic data of the target monitoring area, acquiring a crop growth characteristic time sequence corresponding to the target monitoring area;

[0056] In the embodiment, the geographic data of the target monitoring area affects the types of crops planted in the target monitoring area; according to the growth stages of different types of crops, a crop growth characteristic time sequence is formed, so that corresponding remote sensing images are collected at each growth stage of each crop, facilitating subsequent analysis.

[0057] S3: based on the crop growth characteristic time sequence, collecting a plurality of remote sensing images of the target monitoring area;

[0058] S4: based on the plurality of remote sensing images, confirming a plurality of remote sensing image sets; each remote sensing image set includes a plurality of remote sensing images of the target monitoring area;

[0059] In the embodiment, each remote sensing image set is generated at different time, and the remote sensing image set is generated in time to facilitate analysis.

[0060] S5: acquiring farmland use information of the target monitoring area according to each remote sensing image set.

[0061] In the embodiment, the farmland use information includes that the farmland is non-grain or the farmland is not non-grain.

[0062] The specific implementation process of the embodiment includes:

[0063] In the embodiment, the crop growth characteristic time sequence is acquired through the geographic data of the target monitoring area, and the remote sensing images are collected according to the crop growth characteristic time sequence; and the remote sensing image set is generated according to a plurality of remote sensing images, so as to analyze the plurality of remote sensing images in the remote sensing image set, thereby further acquiring the cultivated land use information of the target monitoring area. The inaccurate and untimely analysis caused by a single remote sensing image is avoided. The technical problem of low timeliness and accuracy in the monitoring process of the non-grain cultivation of the cultivated land in the prior art is solved.

[0064] In the embodiment, the geographic data includes geographic coordinate data, water source condition data, altitude data and illumination data; based on the geographic data of the target monitoring area, the corresponding crop growth characteristic time sequence of the target monitoring area is acquired, including:

[0065] S21: based on the geographic coordinate data, the historical planting data and the agricultural material purchase data in the target monitoring area are acquired and the crop variety database is constructed;

[0066] In the embodiment, the geographic coordinate data includes the central coordinate of the target monitoring area and a plurality of coordinates of the boundary of the target monitoring area.

[0067] In the embodiment, the agricultural material purchase data entering the area is acquired through the interface connected with a plurality of agricultural material purchase platforms; in some embodiments, the geographic coordinate data and the selling data of the local agricultural material store are acquired as the agricultural material purchase data; in order to avoid the influence of the self-retained planting of the cultivator on the analysis result, the crop variety database is further generated in combination with the historical planting data.

[0068] It should be noted that in the embodiment, the crop planting parameters of each crop variety in the crop variety database are set through expert experience and big data when the crop variety database is generated. The crop planting parameters include a plurality of requirements of the crop variety for the planting environment. In the embodiment, the plurality of requirements include water source requirement, altitude requirement and illumination requirement.

[0069] S22: based on the water source condition data, the altitude data and the illumination data of the target monitoring area, the crop variety constraint condition is acquired;

[0070] In the embodiment, the water source constraint condition is acquired according to the water source condition data; the altitude constraint condition is acquired according to the altitude data; the illumination constraint condition is acquired according to the illumination data; and the crop variety constraint condition is acquired based on the water source constraint condition, the altitude constraint condition and the illumination constraint condition. In the embodiment, the water source constraint condition includes water taking difficulty and natural annual rainfall distribution data.

[0071] S23: screening in the crop variety database based on the crop variety constraint condition, obtaining a plurality of plantable crop varieties corresponding to the target monitoring area, and a plurality of characteristic growth stages corresponding to each plantable crop variety; the plantable crop varieties include food crops and non-food crops;

[0072] In this embodiment, a plurality of plantable crop varieties of the target monitoring area are obtained by screening in the crop variety database based on the crop variety constraint condition. In this embodiment, each plantable crop variety has a plurality of characteristic growth stages due to its growth characteristics, and the same plantable crop variety has obvious differences in different characteristic growth stages, which is easy to distinguish from other plantable crop varieties. For example, the growth stages of rice include seedling stage, tillering stage, jointing stage, booting stage, heading stage, flowering stage, filling stage and mature stage; the growth stages of wheat include emergence stage, tillering stage, overwintering stage, green stage, rising stage, jointing stage, booting stage, heading stage, flowering stage, filling stage and mature stage.

[0073] S24: confirming the characteristic acquisition time sequence corresponding to the plantable crop variety based on the plurality of characteristic growth stages corresponding to the plantable crop variety; the characteristic acquisition time sequence includes a plurality of acquisition time points; the plurality of acquisition time points correspond to the plurality of characteristic growth stages one by one;

[0074] In this embodiment, the remote sensing image of the target monitoring area is acquired in each characteristic growth stage of the plantable crop variety; the acquisition time of each characteristic growth stage constitutes the characteristic acquisition time sequence; in this embodiment, the characteristic acquisition time sequence is represented as (t1, t2, …, t n ) v.n v.n represents the nth plantable crop variety, which also serves as a crop marker in this embodiment; t n represents the nth acquisition time point.

[0075] S25: obtaining the crop growth characteristic time sequence based on the characteristic acquisition time sequence of the plurality of plantable crop varieties; each acquisition time point in the crop growth characteristic time sequence is provided with a crop marker of the corresponding plantable crop variety.

[0076] In this embodiment, the acquisition time points in the characteristic acquisition time sequence of the plurality of plantable crop varieties are rearranged in the order of time, to constitute the crop growth characteristic sequence of the target monitoring area. During the rearrangement process, the crop marker of the corresponding plantable crop variety is marked for each acquisition time point. For example, the crop growth characteristic sequence is represented as [t (1，V.1) , t (1，V.2) , t (1，V.3) , t (2，V.2) ……] in an embodiment.

[0077] In this embodiment, based on the time series of crop growth characteristics, a plurality of remote sensing images of the target monitoring area are collected, including:

[0078] S31: Based on each collection time in the feature collection time series, a plurality of collection time windows are obtained; each collection time window includes at least one collection time;

[0079] In this embodiment, according to the distribution of each collection time in the feature collection time series, adjacent and interval time less than the preset time interval are divided into the same collection time window. In this embodiment, the number of collection times in the same collection time window does not exceed the preset number, and the preset number includes 3, 4, 5 and 6. The preset time interval includes 24 hours, 36 hours and 48 hours.

[0080] S32: Based on a plurality of collection times in the same collection time window, a new image collection task is created; each image collection task includes an image collection range, an image quality requirement, and a plurality of crop markers;

[0081] In this embodiment, according to the plurality of collection times in the same collection time window, a plurality of crop markers are obtained. According to a plurality of coordinates of the region boundary of the target monitoring area, the image collection range is confirmed. The image quality requirement includes the resolution requirement and the cloud amount requirement.

[0082] In this embodiment, according to the image collection task, an imaging request for remote sensing images is submitted to the satellite operator.

[0083] S33: According to the image collection task, the remote sensing images of the target monitoring area are obtained and marked with crop markers; each remote sensing image is marked with at least one crop marker.

[0084] In this embodiment, the remote sensing images collected at any time within the collection time window and meeting the image quality requirements are all used as the remote sensing images corresponding to each collection time within the collection time window. After obtaining the remote sensing images, all crop markers corresponding to the collection times are marked on the remote sensing images. For example, a remote sensing image is represented as P[(1, V.1), (1, V.2), (1, V.3)].

[0085] In this embodiment, based on a plurality of remote sensing images, a plurality of remote sensing image sets are confirmed, including:

[0086] S41: Based on a plurality of plantable crop varieties of the target area, a plurality of image empty sets are constructed and each image empty set is set with a set parameter; the set parameter includes a crop marker and a set capacity;

[0087] In this embodiment, each image empty set corresponds to a crop marker. In this embodiment, it is denoted as v. n;

[0088] S42: when an image acquisition task is executed, the remote sensing image is stored in the corresponding image set according to the crop label of the remote sensing image;

[0089] In this embodiment, when the crop label of the remote sensing image has multiple crop labels, the remote sensing image is stored in multiple image sets respectively; for example, a remote sensing image P[(1, V.1), (1, V.2), (1, V.3)], which has crop labels V.1, V.2 and V.3. It is stored in image set v.1 , {} v.2 , {} v.3 , and represented as {P(1, V.1)} v.1 , {P(1, V.2)} v.2 , {P(1, V.3)} v.3

[0090] S43: when the number of remote sensing images in an image set reaches the corresponding set capacity, the remote sensing image set is obtained.

[0091] In this embodiment, when the number of remote sensing images in each image set reaches the set capacity, the remote sensing image set is immediately generated to facilitate the next analysis in time. For the image set whose number of remote sensing images does not reach the set capacity, the execution result of the subsequent image acquisition task is continued to be waited.

[0092] In this embodiment, based on the number of plantable crop varieties in the target area, a plurality of image sets are constructed and set parameters for each image set, including:

[0093] S411: based on the number of plantable crop varieties in the target area, the number of plantable crop varieties, and the number of stages of the characteristic growth stage of each plantable crop variety are obtained;

[0094] S412: based on the number of plantable crop varieties, a plurality of image sets are constructed, and each of the plurality of image sets is labeled based on a plurality of crop labels;

[0095] In this embodiment, the number of image sets is consistent with the number of plantable crop varieties, and the plurality of image sets correspond to the plurality of plantable crop varieties one by one.

[0096] S413: based on the number of stages of the characteristic growth stage corresponding to the crop label of the image set and the lower limit of the preset capacity, the set capacity of the image set is set.

[0097] In this embodiment, the upper limit of the set capacity of each image empty set is not more than the number of stages of the characteristic growth stage of the corresponding plantable crop variety. The lower limit of the preset capacity in this embodiment includes 2 remote sensing images. In this embodiment, the corresponding set capacity is set according to the plantable crop variety based on expert experience.

[0098] In this embodiment, the cultivated land use information of the target monitoring area is obtained according to the remote sensing image set one by one, including:

[0099] S51: Obtain the plot division information of the target monitoring area.

[0100] In this embodiment, the target monitoring area includes a plurality of plots; the plurality of plots can plant the same crop, or can plant different types of crops; for densely populated areas such as mountainous areas, the ownership of the plurality of plots in the target monitoring area involves multiple users. Each user will often choose the variety of crops according to his own ability and needs. In this embodiment, the plot division information includes plot boundary information.

[0101] S52: Each time a remote sensing image set is generated, the remote sensing images in the remote sensing image set are divided based on the plot division information to obtain a plurality of plot remote sensing images.

[0102] In this embodiment, the boundary information in the remote sensing image is recognized according to the plot boundary information, and the plot remote sensing image of each plot is intercepted;

[0103] S53: Based on a plurality of plot remote sensing images of the same plot, a remote sensing image subset is constructed and a crop label is marked; the crop label of each remote sensing image subset is consistent with the crop label of the remote sensing image set;

[0104] In this embodiment, a plurality of plot remote sensing images of the same plot constitute a remote sensing image subset. The crop label and the set capacity of the remote sensing image subset are consistent with the crop label and the set capacity of the corresponding remote sensing image set, respectively.

[0105] S54: Perform crop analysis on the remote sensing image subset corresponding to each plot, and judge whether the actual planted crop of each plot is consistent with the crop label one by one;

[0106] In this embodiment, crop analysis is performed based on a plurality of plot remote sensing images arranged in time sequence in the remote sensing image subset of the same plot, to determine whether the actual planted crop of the plot is consistent with the crop variety corresponding to the crop label.

[0107] S55: Obtain a plurality of plots in which the actual planted crop is consistent with the crop label, and determine whether the plurality of plots are grain land according to the crop label and output a monitoring result.

[0108] In the embodiment, when the actual planted crops and the crop marks of a plot are consistent, if the crop variety corresponding to the current crop mark is a grain crop, it is determined that the plot has not been non-grain; if the crop variety corresponding to the current crop mark is a non-grain crop, it is determined that the plot has been non-grain.

[0109] In the embodiment, the cultivated land use information of the target monitoring area is obtained according to the remote sensing image set one by one, and the method further includes:

[0110] S56: Obtain a plurality of plots in which the actual planted crops are inconsistent with the crop marks, and obtain the corresponding remote sensing image subset for crop analysis when the next remote sensing image set is generated.

[0111] For a plurality of plots in which the actual planted crops are consistent with the crop marks, no crop analysis is performed on the plots when the next remote sensing image set is generated; in some embodiments, no remote sensing image subset of the corresponding plots is generated when the next remote sensing image set is generated. The workload of crop analysis is reduced.

[0112] In the embodiment, the cultivated land use information of the target monitoring area is obtained according to the remote sensing image set one by one, and the method further includes:

[0113] S57: Record the crop maturity time of the plot in which the actual planted crops are consistent with the crop marks.

[0114] In the embodiment, each plot of the target monitoring area is planted with at least one season of crops; for a plot planted with multiple seasons of crops, the next season of crops will be planted after the current season of crops matures; therefore, in the embodiment, the crop maturity time of each plot with a confirmed crop variety is recorded; so that the remote sensing image of the next season of crops can be obtained in time.

[0115] S58: Based on the remote sensing image set generated after the crop maturity time, obtain the remote sensing image subset of the plot and perform crop analysis.

[0116] In the embodiment, the remote sensing image set generated after the crop maturity time includes all remote sensing images in the remote sensing image set, and the collection time of each remote sensing image is not earlier than the crop maturity time.

[0117] In the embodiment, the remote sensing image subset corresponding to each plot is subjected to crop analysis, and whether the actual planted crops of each plot are consistent with the crop marks is judged one by one, including:

[0118] S541: Obtain a crop analysis model based on the crop marks.

[0119] In the embodiment, a corresponding crop analysis model is pre-trained for each crop variety; this makes the crop analysis result more accurate; and it is convenient to process a large number of remote sensing image subsets generated at the same time.

[0120] S542: image processing is performed on each plot remote sensing image in the subset of remote sensing images corresponding to a plot to obtain image feature information;

[0121] In this embodiment, the image processing includes a pre-processing step and a feature extraction step; the pre-processing step includes radiation calibration, atmospheric correction and geometric correction; the feature extraction step includes extracting enhanced vegetation index, normalized difference vegetation index, leaf area vegetation index, red edge chlorophyll vegetation index and texture features in the plot remote sensing image as image feature information; and the image feature information in the same plot remote sensing image is marked with a corresponding acquisition time mark. In this embodiment, the image feature information of a plot remote sensing image is represented as (EVI, NDVI, LAI, RECI, WL) (tm，v.n) ; wherein EVI represents enhanced vegetation index; NDVI represents normalized difference vegetation index; LAI represents leaf area vegetation index; RECI represents red edge chlorophyll vegetation index; WL represents texture features; tm represents acquisition time; and v.n represents crop mark.

[0122] S543: input the image feature information into a crop analysis model to obtain an analysis result; the analysis result includes that the actual planted crop is consistent with the crop mark or that the actual planted crop is inconsistent with the crop mark.

[0123] In this embodiment, the input of the crop analysis model includes image feature information; and the output of the crop analysis model includes "consistent" or "inconsistent".

[0124] In this embodiment, the crop analysis model is obtained based on the crop mark, and includes:

[0125] S5411: obtain a feature relationship database corresponding to the crop mark; the feature relationship database includes image feature information, analysis result and mapping relationship between the image feature information and the analysis result;

[0126] In this embodiment, the feature relationship database is obtained by constructing historical data and expert experience; the feature relationship database includes enhanced vegetation index, normalized difference vegetation index, leaf area vegetation index, red edge chlorophyll vegetation index and texture features at a plurality of acquisition times, analysis structure and mapping relationship between the enhanced vegetation index, normalized difference vegetation index, leaf area vegetation index, red edge chlorophyll vegetation index and texture features at the plurality of acquisition times and the analysis result.

[0127] S5412: train a neural network model through the feature relationship database to obtain the crop analysis model.

[0128] It is to be noted that, in the present text, relational terms 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 actual such 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 also include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0129] The above embodiments are used to illustrate the technical solutions of the present application but not to limit the present application. Although the present application is explained in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application.

Claims

1. A method for monitoring the conversion of arable land to non-grain crops based on remote sensing data, characterized in that: The monitoring method includes: Acquire geographic data of the target monitoring area; Based on the geographic data of the target monitoring area, obtain the time series of crop growth characteristics corresponding to the target monitoring area; Based on the time series of crop growth characteristics, several remote sensing images of the target monitoring area were collected; Based on several remote sensing images, several remote sensing image sets are identified; each remote sensing image set includes multiple remote sensing images of the target detection area; Based on the remote sensing image set, farmland use information for the target monitoring area is obtained one by one.

2. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 1, characterized in that: Geographic data includes geographic coordinate data, water source condition data, altitude data, and illumination data; Based on the geographical data of the target monitoring area, obtain the time series of crop growth characteristics corresponding to the target monitoring area, including: Based on geographic coordinate data, historical planting data and agricultural input purchase data in the target monitoring area are obtained and a crop variety database is constructed. Based on water source conditions data, altitude data, and light data of the target monitoring area, crop variety constraints are obtained. Based on crop variety constraints, the crop variety database is used to screen and obtain several cultivable crop varieties corresponding to the target monitoring area, as well as several characteristic growth stages corresponding to each cultivable crop variety; cultivable crop varieties include food crops and non-food crops. Based on several characteristic growth stages corresponding to cultivable crop varieties, the characteristic collection time series corresponding to cultivable crop varieties is identified; the characteristic collection time series includes several collection times; several collection times correspond one-to-one with several characteristic growth stages. Based on the feature collection time series of several cultivable crop varieties, a crop growth feature time series is obtained; each collection time in the crop growth feature time series is marked with a corresponding crop tag for the cultivable crop variety.

3. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 2, characterized in that: Based on the time series of crop growth characteristics, several remote sensing images of the target monitoring area were collected, including: Based on each acquisition moment in the feature acquisition time series, several acquisition time windows are obtained; each acquisition time window includes at least one acquisition moment. Create new image acquisition tasks based on several acquisition moments within the same acquisition time window; each image acquisition task includes the image acquisition range, image quality requirements, and several crop markers. According to the image acquisition task, acquire remote sensing images of the target monitoring area and mark crop tags; each remote sensing image is marked with at least one crop tag.

4. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 3, characterized in that: Based on several remote sensing images, several sets of remote sensing images were identified, including: Based on several cultivable crop varieties in the target area, several empty sets of images are constructed and set set parameters for each empty set of images; the set parameters include crop labels and set capacity; When an image acquisition task is completed, the crop label of the remote sensing image is stored in at least one corresponding empty image set; When the number of remote sensing images in an image empty set reaches the corresponding set capacity, the remote sensing image set is obtained.

5. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 4, characterized in that: Based on several cultivable crop varieties in the target region, several empty image sets are constructed, and set parameters are set for each empty image set, including: Based on several cultivable crop varieties in the target area, obtain the number of cultivable crop varieties, as well as the crop markers and the number of characteristic growth stages for each cultivable crop variety. Based on the number of cultivable crop varieties, several empty image sets are constructed, and each empty image set is labeled one by one based on several crop labels. Based on the number of crop growth stages corresponding to the empty set of the image, and the preset lower limit of the capacity, the set capacity of the empty set of the image is set.

6. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 4, characterized in that: Based on the remote sensing image set, farmland use information for the target monitoring area is obtained one by one, including: Obtain land parcel delineation information for the target monitoring area; Whenever a set of remote sensing images is generated, the remote sensing images in the set are divided based on the land parcel division information to obtain several land parcel remote sensing images; Based on remote sensing images of several plots of land in the same area, a subset of remote sensing images is constructed and crop labels are assigned; the crop labels of each subset of remote sensing images are consistent with the crop labels of the set of remote sensing images. Crop analysis is performed on a subset of remote sensing images corresponding to each plot, and it is determined one by one whether the actual crops planted in each plot match the crop labels. The system acquires multiple plots of land that match the actual planted crops and crop tags, determines whether these plots are designated for grain production based on the crop tags, and outputs the monitoring results.

7. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 6, characterized in that: Crop analysis is performed on a subset of remote sensing images corresponding to each plot, determining whether the actual crops planted in each plot match the crop labels, including: Based on crop tags, obtain crop analysis models; Image processing is performed on the remote sensing images of each plot in the subset of remote sensing images corresponding to a plot to obtain image feature information; Image feature information is input into the crop analysis model to obtain analysis results; the analysis results include whether the actual planted crop matches the crop label or whether the actual planted crop does not match the crop label.

8. The method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 7, characterized in that: Based on crop markers, obtain crop analysis models, including: Obtain the feature relationship database corresponding to crop tags; the feature relationship database includes image feature information, analysis results, and the mapping relationship between image feature information and analysis results; A crop analysis model is obtained by training a neural network model using a feature relation database.

9. A method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 6, characterized in that: Based on the remote sensing image set, the farmland use information of the target monitoring area is obtained one by one, including: Several plots of land where the actual crops planted do not match the crop labels are identified. When the next set of remote sensing images is generated, the corresponding subsets of remote sensing images are acquired one by one for crop analysis.

10. A method for monitoring the conversion of arable land to non-grain crops based on remote sensing data as described in claim 9, characterized in that: Based on the remote sensing image set, the farmland use information of the target monitoring area is obtained one by one, including: Record the crop maturity time for plots where the actual planted crops match the crop markings; Based on a set of remote sensing images generated after crop maturity, a subset of remote sensing images of the plot is obtained and crop analysis is performed.

Images