Crop planting structure identification system based on remote sensing data of unmanned aerial vehicle

Abstract

The invention discloses a crop planting structure identification system based on unmanned aerial vehicle remote sensing data, and belongs to the technical field of agricultural planting, and the system comprises an unmanned aerial vehicle remote sensing detection module which obtains remote sensing data related to crops in a target area; the data processing and transmitting module is used for preprocessing the acquired remote sensing data and transmitting the preprocessed remote sensing data to the central processing center; an unmanned aerial vehicle detection unit, a recognition unit and an analysis unit are arranged in the central processing center, and the unmanned aerial vehicle detection unit is used for monitoring and correcting the flight path of the unmanned aerial vehicle; the identification unit is used for identifying a crop planting structure in the target area; and the analysis unit is used for carrying out growth abnormity early warning analysis on the identified crops. According to the invention, by training the sample data of crops in different periods and comparing the obtained remote sensing data through the identification unit, the crop type in each planting area can be determined more accurately through comparison and analysis in multiple periods.

Description

Technical Field

[0001] The present invention belongs to the technical field of agricultural planting, and in particular relates to a crop planting structure recognition system based on unmanned aerial vehicle remote sensing data. Background Art

[0002] The crop planting structure refers to the proportion of crop types planted in a region or country. By identifying the crop planting structure, we can understand and manage the distribution and growth status of different crops in the farmland, which is of great significance for optimizing the agricultural production layout and improving resource utilization efficiency.

[0003] Traditional crop planting structure identification is mostly carried out through manual surveys and sampling statistics. Manual surveys involve professionals visiting the field to record crop distribution. Although the data is accurate, it is time-consuming, labor-intensive and costly. Sampling statistics select some plots for detailed investigation and then infer the overall layout, which is prone to large errors and poor timeliness. In addition, sampling statistics cannot provide a more comprehensive understanding of the growth of crops in the planting area, which makes it difficult to understand the growth status of crops in a timely manner, and thus cannot effectively respond to possible crop disasters, which can easily cause large-scale crop losses and reduced yields. Summary of the invention

[0004] The purpose of the present invention is to provide a crop planting structure recognition system based on UAV remote sensing data to solve the problems faced in the above-mentioned background technology.

[0005] The purpose of the present invention can be achieved through the following technical solutions:

[0006] A crop planting structure recognition system based on unmanned aerial vehicle remote sensing data, the system comprising:

[0007] A UAV remote sensing detection module, wherein the UAV remote sensing detection module is used to perform non-contact long-distance detection of a target area to obtain remote sensing data related to crops in the target area;

[0008] A data processing and transmission module, which is used to pre-process the acquired remote sensing data and then transmit it to a central processing center;

[0009] A central processing center, wherein the central processing center is provided with a drone detection unit, an identification unit, and an analysis unit;

[0010] The drone detection unit is connected to the drone and is used to monitor the flight path of the drone and correct the path when the flight path of the drone deviates;

[0011] The identification unit is used to identify the crop planting structure in the target area according to the acquired remote sensing data;

[0012] The analysis unit is used to further analyze the acquired remote sensing data and perform growth abnormality early warning analysis on the identified crops.

[0013] Furthermore, the data processing and transmission module works as follows:

[0014] Correct the acquired remote sensing data to eliminate atmospheric scattering and absorption effects, and correct the distorted lens and deformed projection:

[0015] The image stitching technology is used to reorganize and merge the acquired images into a complete image, and the stitched image is color balanced;

[0016] The processed remote sensing data is transmitted to the central processing center via wireless transmission technology.

[0017] Furthermore, the working method of the drone detection unit is:

[0018] Set the flight path of the drone, and determine the allowable error area of ​​the drone flight according to the set flight path, drive the drone to fly according to the set flight path, obtain the position of the drone in real time, and determine whether the drone is within the allowable error area. If the drone is not within the allowable error area, obtain the distance H between the position of the drone and the set flight path, and obtain the wind speed v and wind direction at this time;

[0019] When the wind direction is downwind, the formula Control the drone to fly in the direction of the set flight path Y S time;

[0020] When the wind direction is against the wind, the formula Control the drone to fly in the direction of the set flight path Y N time;

[0021] Among them, v 0 Adjust the offset speed for the set drone, α is the conversion coefficient.

[0022] Furthermore, the identification unit works as follows:

[0023] Obtain a large number of crop samples from different periods from big data, and train them to obtain training models of crops at different periods;

[0024] Acquire image data collected in the current period, identify image information in the image, and divide the target area into multiple planting areas according to the similarity of the images;

[0025] Input the images of each planting area into the training model of the corresponding period to identify the type of crops in each planting area;

[0026] When the recognition result is that the crop type in the planting area is only a single type, the crop type in the planting area is determined to be the recognized type; otherwise, continue to acquire image data at different periods for recognition until the recognition result is that the crop type in the planting area is only a single type, then the crop type in the planting area is determined to be the recognized type.

[0027] Furthermore, the working method of the analysis unit is:

[0028] Obtain image data from each planting area, use image recognition technology to extract the color features, texture features, shape features, and coverage features of the planting area, and input the color features, texture features, shape features, and coverage features into the pre-constructed color feature evaluation value set, texture feature evaluation value set, shape feature evaluation value set, and coverage feature evaluation value set to obtain the color evaluation value G of the planting area. R , texture evaluation value G W , shape evaluation value G X And the coverage G H ;

[0029] So through the formula The deviation degree G of the planting area is obtained;

[0030] When G>G th When , it is judged that there is growth abnormality in the planting area;

[0031] in, is the standard color evaluation value of the planting area in the same period, is the standard texture evaluation value of the planting area in the same period, is the standard shape evaluation value of the planting area in the same period, is the standard coverage rate of the planting area in the same period, ΔG R is the color evaluation comparison value set, ΔG W is the texture evaluation comparison value set, ΔG X is the shape evaluation comparison value set, ΔG H Set the coverage comparison value for G th is the preset deviation threshold.

[0032] Furthermore, the analysis unit working method also includes:

[0033] Get the deviation degree G of all planting areas in the target area i ,

[0034] By formula The overall status value K of the target area is obtained.

[0035] When K>K th , it is determined that there is an abnormality in the target area;

[0036] Where n is the number of planting areas, and i∈[1,n], G>G in all planting areas th The number of planting areas, Pr is the concentration coefficient value, K th is the preset overall status threshold.

[0037] Furthermore, the concentration coefficient value Pr is obtained by:

[0038] The target area is divided into m large areas according to the area. Each large area contains the same number of planting areas. The number of planting areas with abnormal growth in each large area is obtained.

[0039] By formula The concentration coefficient value Pr is obtained;

[0040] Among them, b j is the number of planting areas with abnormal growth in the jth region, and j∈[1,m], max b It is the number of planting areas with the most growth abnormalities.

[0041] Beneficial effects of the present invention:

[0042] The present invention adopts UAV remote sensing technology to identify crop planting structures, eliminating manual investigation and sampling statistics, which can save a lot of human resource costs. At the same time, the UAV detection unit monitors the UAV flight path in real time to prevent the UAV flight from deviating from the predetermined trajectory and make timely corrections. This can ensure that the acquired data is clearer and more accurate, reduce subsequent data processing, and effectively reduce subsequent identification errors.

[0043] The present invention trains sample data of crops at different periods, and then compares the acquired remote sensing data through a recognition unit. Through comparison and analysis of multiple periods, the type of crops in each planting area can be accurately determined.

[0044] The analysis unit provided in the present invention can not only judge the growth conditions of crops in each planting area, thereby issuing timely warnings to reduce crop losses, but also monitor the crops in the entire target area, and can know an abnormal situation in the entire target area and the concentrated distribution of abnormalities, so as to more conveniently understand which approximate location in the target area has a larger abnormality, so as to prescribe the right remedy and promptly rectify the target area to ensure the crop yield in the entire target area.

[0045] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0047] Figure 1 It is a system block diagram of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. 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 creative work are within the scope of protection of the present invention.

[0049] In one embodiment, a crop planting structure recognition system based on drone remote sensing data is disclosed. Figure 1 As shown, the identification system includes:

[0050] UAV remote sensing detection module: The UAV remote sensing detection module is used to conduct non-contact long-distance detection of the target area to obtain remote sensing data related to crops in the target area;

[0051] A data processing and transmission module is used to pre-process the acquired remote sensing data and then transmit it to the central processing center;

[0052] The central processing center includes a drone detection unit, an identification unit, and an analysis unit;

[0053] The drone detection unit is connected to the drone to monitor the flight path of the drone and correct the path when the drone's flight path deviates;

[0054] The identification unit is used to identify the crop planting structure in the target area based on the acquired remote sensing data;

[0055] The analysis unit is used to further analyze the acquired remote sensing data and conduct early warning analysis of growth abnormalities on the identified crops.

[0056] Through the above technical scheme, the present application adopts UAV remote sensing technology to identify the crop planting structure, eliminating manual investigation and sampling statistics, which can save a lot of human resource costs. At the same time, the UAV detection unit monitors the UAV flight path in real time to prevent the UAV flight from deviating from the predetermined trajectory and make timely corrections, which can ensure that the acquired data is clearer and more accurate, reduce subsequent data processing, and effectively reduce subsequent identification errors; in addition, by training sample data of crops at different periods, and then comparing the acquired remote sensing data through the identification unit, through comparison and analysis of multiple periods, the type of crops in each planting area can be accurately determined; the analysis unit set in the present application can not only judge the growth of crops in each planting area, so as to give timely warning to reduce crop losses, but also monitor the crops in the entire target area to determine whether the target area is abnormal, so as to prescribe the right medicine and rectify the target area in time to ensure the crop yield of the entire target area.

[0057] In one embodiment of the present invention, the working method of the data processing and transmission module is as follows:

[0058] Correct the acquired remote sensing data to eliminate atmospheric scattering and absorption effects, and correct the distorted lens and deformed projection:

[0059] The image stitching technology is used to reorganize and merge the acquired images into a complete image, and the stitched image is color balanced;

[0060] The processed remote sensing data is transmitted to the central processing center via wireless transmission technology.

[0061] After the drone acquires remote sensing data of the target area, atmospheric correction is used to eliminate atmospheric scattering and absorption effects to make the reflectivity in the image closer to the actual value. Geometric correction is used to correct lens distortion and projection deformation to ensure that the image pixel position accurately corresponds to the ground coordinates. Image stitching technology is then used to reorganize and merge the acquired images and stitch them into a complete image. Finally, color balancing technology is used to adjust the color differences between adjacent images to ensure overall visual consistency. After preprocessing, it is more convenient for subsequent analysis, and then wireless transmission technology is used to transmit the values ​​to the central processing center for subsequent analysis.

[0062] The working method of the drone detection unit is as follows: set the drone flight path, and determine the drone flight allowable error area according to the set flight path, drive the drone to fly according to the set flight path, obtain the drone's position in real time, and determine whether the drone is within the allowable error area. When the drone is not within the allowable error area, obtain the distance H between the drone's position and the set flight path, and obtain the wind speed v and wind direction at this time;

[0063] When the wind direction is downwind, the formula Control the drone to fly in the direction of the set flight path Y S time;

[0064] When the wind direction is against the wind, the formula Control the drone to fly in the direction of the set flight path Y N time;

[0065] Among them, v 0 Adjust the offset speed for the set drone, α is the conversion coefficient.

[0066] The above scheme provides a specific working method of the drone detection unit. First, the drone flight path is set, and according to the set flight path, the allowable error area of ​​the drone flight is determined, the drone is driven to fly according to the set flight path, the position of the drone is obtained in real time, and it is determined whether the drone is in the allowable error area. When the drone is not in the allowable error area, it means that the drone has a position deviation at this time, so the acquired image will deviate from the expected one. In order to ensure the integrity of the acquired image, the drone needs to be adjusted in time, specifically: the distance H between the position of the drone and the set flight path is obtained, and the wind speed v and wind direction at this time are obtained. Since the flight of the drone is easily affected by the wind, a large adjustment is made. Therefore, in order to better adjust the drone distance, the drone position is adjusted by adjusting the drone flight time. When the wind direction is a tailwind, the formula is used at this time. Control the drone to fly in the direction of the set flight path Y S time, we know that when the wind is favorable, the drone is affected by the wind thrust and the driving speed is faster, so the drone flight time needs to be shortened compared to the original time. It can be seen that the greater the wind speed, the more time is reduced; similarly, when the wind direction is against the wind, the drone is affected by wind resistance and the driving speed is slower, so the drone flight time needs to be increased compared to the original time. It can be seen that the greater the wind resistance, the more time needs to be increased. At this time, through the formula Control the drone to fly in the direction of the set flight path Y N In this way, when the UAV has a deviation error, the position of the UAV can be adjusted in time according to the current wind speed and wind direction to ensure the integrity of the image obtained by the UAV and facilitate subsequent analysis.

[0067] The working method of the recognition unit is: obtain a large number of crop samples at different periods from big data, and train them to obtain training models of crops at different periods;

[0068] Acquire image data collected in the current period, identify image information in the image, and divide the target area into multiple planting areas according to the similarity of the images;

[0069] Input the images of each planting area into the training model of the corresponding period to identify the type of crops in each planting area;

[0070] When the recognition result is that the crop type in the planting area is only a single type, the crop type in the planting area is determined to be the recognized type; otherwise, continue to acquire image data at different periods for recognition until the recognition result is that the crop type in the planting area is only a single type, then the crop type in the planting area is determined to be the recognized type.

[0071] The above technical solution provides a specific working method of the recognition unit. First, a large number of crop samples of different periods are obtained from big data, and training is performed to obtain training models of crops at different periods. Then, image data collected in the current period is obtained, and image information in the image is identified. According to the similarity of the image, the target area is divided into multiple planting areas; the images of each planting area are input into the training model of the corresponding period, so as to identify the types of crops in each planting area. Since the same planting area may have multiple types of crops in the final result when identifying, when the identification result is that the crop type in the planting area is only a single type, it means that the type of crop planted is the identified crop type. Otherwise, continue to obtain image data of crops at different periods, analyze and identify, until the identification result is that the crop type in the planting area is only a single type, then determine that the crop type in the planting area is the identified type. The use of drone remote sensing technology to obtain image data of multiple periods in the planting area for comparison and analysis can accurately determine the types of crops in each planting area, which can greatly reduce recognition errors.

[0072] The working method of the analysis unit is as follows: image data of each planting area is obtained, and the color features, texture features, shape features and coverage features of the planting area are extracted by using image recognition technology, and the color features, texture features, shape features and coverage features are input into the color feature evaluation value set, texture feature evaluation value set, shape feature evaluation value set and coverage feature evaluation value set constructed in advance to obtain the color evaluation value G of the planting area. R , texture evaluation value G W , shape evaluation value G X And the coverage G H ;

[0073] So through the formula The deviation degree G of the planting area is obtained;

[0074] When G>G th When , it is judged that there is growth abnormality in the planting area;

[0075] in, is the standard color evaluation value of the planting area in the same period, is the standard texture evaluation value of the planting area in the same period, is the standard shape evaluation value of the planting area in the same period, is the standard coverage rate of the planting area in the same period, ΔG R is the color evaluation comparison value set, ΔG W is the texture evaluation comparison value set, ΔG X is the shape evaluation comparison value set, ΔG H Set the coverage comparison value for G th is the preset deviation threshold.

[0076] The above scheme provides an analysis unit to determine whether there is abnormal growth in a single planting area. First, image data in each planting area is obtained, and the color features, texture features, shape features and coverage features of the planting area are extracted using image recognition technology. The color features, texture features, shape features and coverage features are input into the color feature evaluation value set, texture feature evaluation value set, shape feature evaluation value set and coverage feature evaluation value set constructed in advance to obtain the color evaluation value G of the planting area. R , texture evaluation value G W , shape evaluation value G X And the coverage G H , and then through the formula The deviation degree G of the planting area is obtained; the obtained color evaluation value, texture evaluation value, shape evaluation value and coverage are compared with the standard color evaluation value, texture evaluation value, shape evaluation value and coverage rate of the same period. It can be seen that when the difference is larger, the possibility of abnormal crop growth is greater. Therefore, after the deviation degree G is obtained, it is compared with the preset deviation degree threshold G th For comparison, when G>G th In this way, the difference between various characteristic parameters of crops in the planting area and historical standard characteristic parameters can be analyzed, which can make it easier for managers to understand the growth of crops and manage the planting areas with abnormal growth in time to reduce crop losses.

[0077] It should be noted that the standard color evaluation value in the same period Standard texture rating Standard shape evaluation value Standard coverage The color evaluation comparison value ΔG is based on the historical data of normal crop growth. R, Texture evaluation comparison value ΔG W , shape evaluation comparison value ΔG X , coverage comparison value ΔG H , the preset deviation threshold G th It can be formulated based on empirical data, which will not be described in detail here.

[0078] The working method of the analysis unit also includes: obtaining the deviation degree G of all planting areas in the target area i ,

[0079] By formula The overall status value K of the target area is obtained.

[0080] When K>K th , it is determined that there is an abnormality in the target area;

[0081] Where n is the number of planting areas, and i∈[1,n], G>G in all planting areas th The number of planting areas, Pr is the concentration coefficient value, and the method for obtaining the concentration coefficient value Pr is as follows: divide the target area into m large areas according to the area, each large area contains the same number of planting areas, and obtain the number b of planting areas with abnormal growth in each large area,

[0082] By formula The concentration coefficient value Pr is obtained;

[0083] Among them, b j is the number of planting areas with abnormal growth in the jth region, and j∈[1,m], max b It is the number of planting areas with the most growth abnormalities.

[0084] The above technical solution provides a method for the analysis unit to perform an overall analysis on the target area. First, the target area is divided into m large areas according to the area. Each large area contains the same number of planting areas. The number of planting areas with abnormal growth in each large area is obtained. The formula The concentration coefficient value Pr is obtained, and the entire target area is divided into multiple large areas. The number of planting areas with abnormal growth in each large area is obtained. The number can not only know the abnormal situation and concentrated distribution of the entire target area, but also make it more convenient to understand which approximate location in the target area has a larger abnormality, which is convenient for subsequent symptomatic treatment. Then the deviation degree G of all planting areas in the target area is obtained. i , through the formula The overall condition value K of the target area is obtained, and the formula It is expressed as the proportion of planting areas with abnormal growth in the target area. The formula is It indicates the average deviation of the entire target area. It can be seen that the larger the overall condition value is, the greater the possibility of abnormality in the target area is. Therefore, it is compared with the preset overall condition threshold K th Compare, when K>K th In this way, the deviation degree of the planting area in the entire target area can be used to determine whether the entire target area is abnormal. This not only allows us to know an abnormal situation in the entire target area and the distribution of abnormalities, but also makes it easier to understand which approximate location in the target area has a large abnormality, which is convenient for subsequent remedies and timely rectification of the target area to ensure the subsequent crop yield of the entire target area.

[0085] The above contents are merely examples and explanations of the concept of the present invention. The technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the claims, they should all fall within the protection scope of the present invention.

Claims

1. A crop planting structure recognition system based on UAV remote sensing data, characterized in that: The system comprises: A UAV remote sensing detection module, wherein the UAV remote sensing detection module is used to perform non-contact long-distance detection of a target area to obtain remote sensing data related to crops in the target area; A data processing and transmission module, which is used to pre-process the acquired remote sensing data and then transmit it to a central processing center; A central processing center, wherein the central processing center is provided with a drone detection unit, an identification unit, and an analysis unit; The drone detection unit is connected to the drone and is used to monitor the flight path of the drone and correct the path when the flight path of the drone deviates; The identification unit is used to identify the crop planting structure in the target area according to the acquired remote sensing data; The analysis unit is used to further analyze the acquired remote sensing data and perform growth abnormality early warning analysis on the identified crops.

2. The crop planting structure recognition system based on UAV remote sensing data according to claim 1 is characterized in that: The working method of the data processing and transmission module is: Correct the acquired remote sensing data to eliminate atmospheric scattering and absorption effects, and correct the distorted lens and deformed projection: The image stitching technology is used to reorganize and merge the acquired images into a complete image, and the stitched image is color balanced; The processed remote sensing data is transmitted to the central processing center via wireless transmission technology.

3. The crop planting structure recognition system based on UAV remote sensing data according to claim 1 is characterized in that: The working method of the drone detection unit is: Set the flight path of the drone, and determine the allowable error area of ​​the drone flight according to the set flight path, drive the drone to fly according to the set flight path, obtain the position of the drone in real time, and determine whether the drone is within the allowable error area. If the drone is not within the allowable error area, obtain the distance H between the position of the drone and the set flight path, and obtain the wind speed v and wind direction at this time; When the wind direction is downwind, the formula Control the drone to fly in the direction of the set flight path Y S time; When the wind direction is against the wind, the formula Control the drone to fly in the direction of the set flight path Y N time; Among them, v0 is the set drone adjustment offset speed, and α is the conversion coefficient.

4. The crop planting structure recognition system based on UAV remote sensing data according to claim 1 is characterized in that: The working method of the identification unit is: Obtain a large number of crop samples from different periods from big data, and train them to obtain training models of crops at different periods; Acquire image data collected in the current period, identify image information in the image, and divide the target area into multiple planting areas according to the similarity of the images; Input the images of each planting area into the training model of the corresponding period to identify the type of crops in each planting area; When the recognition result is that the crop type in the planting area is only a single type, the crop type in the planting area is determined to be the recognized type; otherwise, continue to acquire image data at different periods for recognition until the recognition result is that the crop type in the planting area is only a single type, then the crop type in the planting area is determined to be the recognized type.

5. The crop planting structure recognition system based on UAV remote sensing data according to claim 1 is characterized in that: The working method of the analysis unit is: Obtain image data from each planting area, use image recognition technology to extract the color features, texture features, shape features, and coverage features of the planting area, and input the color features, texture features, shape features, and coverage features into the pre-constructed color feature evaluation value set, texture feature evaluation value set, shape feature evaluation value set, and coverage feature evaluation value set to obtain the color evaluation value G of the planting area. R , texture evaluation value G W , shape evaluation value G X And the coverage G H ; So through the formula The deviation degree G of the planting area is obtained; When G>G th When , it is judged that there is growth abnormality in the planting area; in, is the standard color evaluation value of the planting area in the same period, is the standard texture evaluation value of the planting area in the same period, is the standard shape evaluation value of the planting area in the same period, is the standard coverage rate of the planting area in the same period, ΔG R is the color evaluation comparison value set, ΔG W is the texture evaluation comparison value set, ΔG X is the shape evaluation comparison value set, ΔG H Set the coverage comparison value for G th is the preset deviation threshold.

6. The crop planting structure recognition system based on UAV remote sensing data according to claim 5 is characterized in that: The analysis unit working method also includes: Get the deviation degree G of all planting areas in the target area i , By formula The overall status value K of the target area is obtained. When K>K th , it is determined that there is an abnormality in the target area; Where n is the number of planting areas, and i∈[1,n], G>G in all planting areas th The number of planting areas, Pr is the concentration coefficient value, K th is the preset overall status threshold.

7. The crop planting structure recognition system based on UAV remote sensing data according to claim 6 is characterized in that: The concentration coefficient value Pr is obtained by: The target area is divided into m large areas according to the area. Each large area contains the same number of planting areas. The number of planting areas with abnormal growth in each large area is obtained. By formula The concentration coefficient value Pr is obtained; Among them, b j is the number of planting areas with abnormal growth in the jth region, and j∈[1,m], max b It is the number of planting areas with the most abnormal growth.

Images