Tropical agricultural ecology remote sensing monitoring equipment

Abstract

The invention relates to the technical field of remote sensing monitoring equipment, and discloses tropical agricultural ecology remote sensing monitoring equipment which comprises an unmanned aerial vehicle platform, a remote sensing unmanned aerial vehicle is arranged on the unmanned aerial vehicle platform, and a monitoring sensor assembly is loaded on the remote sensing unmanned aerial vehicle. The monitoring sensor assembly comprises a high-resolution camera device, a hyperspectral sensor, a thermal infrared sensor and a laser radar, a data receiving module, a data processing module and a display module are arranged in the unmanned aerial vehicle platform, and the monitoring sensor assembly is mainly used for obtaining soil images and data information. The data processing module can judge soil ecological information according to images and data, and the monitoring sensor assembly and the data processing module are matched to obtain the erosion degree and fertility condition of soil in the planting area, the moisture content, the soil type, the moisture evaporation condition and the slope and direction information of the soil. The growth state of plants in a planting area and the ecological environment of soil are monitored, and producers are better assisted in formulating targeted agricultural management measures.

Description

Technical Field

[0001] The invention relates to the technical field of remote sensing monitoring equipment, in particular to tropical agricultural ecological remote sensing monitoring equipment. Background Art

[0002] Tropical agriculture refers to agriculture developed in tropical areas with high temperatures and high rainfall, long summers and no winters throughout the year, and abundant water and heat resources. It features crops such as rice and sugarcane, as well as a variety of tropical cash crops, cash trees, and tropical fruits. The world's tropical crops are mainly distributed in South Asia, Southeast Asia, Africa (except parts of North Africa and South Africa), Latin America (except most of Argentina and Chile), and Oceania (except central and southern Australia and New Zealand). In China, tropical crops are mainly distributed in Hainan and parts of Guangdong, Guangxi, Yunnan, Fujian, Hunan, Sichuan, Guizhou, Tibet, and Taiwan. Among them, Hainan Island and Xishuangbanna are the most suitable areas for the growth of tropical crops.

[0003] Currently, monitoring of tropical crops is usually carried out using drones equipped with high-definition cameras, which are mainly used to monitor plant growth conditions and pests and diseases. This tends to overlook the impact of soil ecology on tropical crops, thereby affecting the effectiveness of agricultural planting monitoring and failing to assist producers in agricultural management. Summary of the invention

[0004] In order to overcome the above-mentioned defects of the prior art, the present invention provides a tropical agricultural ecological remote sensing monitoring device to solve the problems existing in the above-mentioned background technology.

[0005] The present invention provides the following technical solutions: a tropical agricultural ecological remote sensing monitoring device, comprising an unmanned aerial vehicle platform, on which a remote sensing unmanned aerial vehicle is arranged, and on which a monitoring sensor component is loaded, and the monitoring sensor component comprises a high-resolution camera, a hyperspectral sensor, a thermal infrared sensor and a laser radar (LiDAR). A data receiving module, a data processing module and a display module are arranged inside the unmanned aerial vehicle platform, and the monitoring sensor component is mainly used to obtain soil images and data information, and the data processing module can judge soil ecological information according to the images and data. The monitoring sensor component and the data processing module cooperate to realize the monitoring of the growth status of plants in the planting area and the ecological environment of the soil, and can better assist producers in formulating targeted agricultural management measures.

[0006] Furthermore, the high-resolution camera equipment is used to capture high-resolution images of the soil surface, and the data processing module will pre-process the collected images, including denoising, image enhancement (such as rotation, flipping, scaling, etc.), radiation correction, atmospheric correction and geometric correction, and then identify the degree of exposure and vegetation coverage of the soil surface through image analysis, and indirectly judge the degree of soil erosion and fertility.

[0007] Furthermore, image analysis includes feature extraction, type recognition and result calculation. Feature extraction includes color feature extraction, texture feature extraction, shape feature extraction and depth feature extraction. Color feature extraction is to extract the difference in color between bare soil and vegetation through color histogram. Texture feature extraction is to extract the surface texture features of bare soil and vegetation by using the gray-level co-occurrence matrix method. Shape feature extraction is to obtain the shape features of bare soil and vegetation through edge detection and contour extraction. Deep feature extraction uses a pre-trained convolutional neural network (CNN) as a feature extractor to extract high-level features in the image. The extracted features are trained using random forests to establish a classification model. The pre-processed image is input into the trained classification model. The model will classify the pixels or areas in the image according to the extracted features, distinguish between bare soil and vegetation, and quantify the classification results. Calculate the bare soil area and vegetation coverage, so as to judge the degree of soil erosion and fertility.

[0008] Furthermore, the hyperspectral sensor can obtain the reflectance information of the soil in different spectral bands, and correct the collected spectral data through the data processing module, including radiation correction, atmospheric correction and geometric correction, and then extract the features related to the soil moisture content, organic matter content and soil type from the preprocessed spectral data, and then screen the extracted features according to the relevance and importance, and select the features with predictive ability for the target variables (such as soil moisture content, organic matter content and soil type) as the feature group, use the support vector machine, establish a prediction model according to the selected features, and infer the soil moisture content, organic matter content and soil type by analyzing the relevant features through the prediction model.

[0009] Furthermore, the hyperspectral sensor can also detect the "red edge position" of vegetation. By detecting the "red edge position" formed by the spectral characteristics of vegetation, the data processing module analyzes the changes in the "red edge position" of vegetation, thereby confirming the current growth status of vegetation and land pollution status.

[0010] The description of the red edge position includes the position of the red edge and the slope of the red edge. With the changes in chlorophyll content, biomass and phenology, the red edge position will move along the wavelength axis in the coordinates of the spectrum curve. When the chlorophyll content of the vegetation is high and the growth is vigorous, the "red edge" position will shift toward the infrared direction, which is called "red shift"; when the vegetation suffers from environmental stress, pests and diseases, or the chlorophyll content is reduced due to phenological changes, the "red edge" position will shift toward the blue light direction, which is called "blue shift". Vegetation coverage is related to the leaf area index. The higher the vegetation coverage, the greater the leaf area index, the greater the slope of the red edge, the better the corresponding vegetation growth state, and the red edge position will appear "red shift"; conversely, the red edge position will be correspondingly "blue shifted".

[0011] By analyzing the correlation between chlorophyll content or vegetation growth status parameters and soil heavy metal content, an inversion model for soil heavy metal content is established to reasonably predict and deduce the soil heavy metal pollution situation.

[0012] Furthermore, the thermal infrared sensor is used to measure the temperature of the soil surface. During the day, soil with high humidity may appear cold anomaly on the thermal infrared image due to its large heat capacity and slow temperature increase; while soil with low humidity will heat up quickly and appear hot anomaly. The soil surface temperature data monitored by the thermal infrared sensor can be used to identify the thermal anomaly area of ​​the soil, and the cooling effect intensity of water evaporation can be calculated through the data processing module, thereby indirectly evaluating the water evaporation situation of the soil.

[0013] Furthermore, the laser radar (LiDAR) can provide high-precision terrain and vegetation height information. By using the laser radar equipment to scan the target area, a large amount of distance measurement data is obtained, and then the collected laser radar data is preprocessed. Then, based on the processed laser radar data, an interpolation algorithm or a grid generation algorithm is used to generate a high-precision topographic map. Then, on the generated topographic map, a slope calculation tool or algorithm is used to extract slope information. Finally, a soil erosion model - the water area erosion prediction and assessment model WEPP is used to assess the soil erosion risk. Erosion factors include rainfall intensity, soil type, vegetation coverage and terrain characteristics. Rainfall intensity can be obtained from the local meteorological bureau, and soil type, vegetation coverage and terrain characteristics can be measured by the monitoring sensor components carried by itself. The erosion factors are input into the water area erosion prediction and assessment model WEPP to obtain the soil erosion risk level, and then the soil status of the planting area is evaluated.

[0014] Furthermore, the display module includes a display screen and a human-computer interaction device, which can display the data information collected by the monitoring sensor component on the drone and the data information analyzed and processed by the data processing module.

[0015] Beneficial effects:

[0016] 1. The tropical agricultural ecological remote sensing monitoring equipment is equipped with a monitoring sensor component, which is equipped with high-resolution camera equipment, hyperspectral sensors, thermal infrared sensors and laser radars. It can monitor the growth status of plants in the planting area and the ecological environment of the soil in many aspects, thereby increasing the effect of agricultural planting monitoring.

[0017] 2. The tropical agricultural ecological remote sensing monitoring equipment is equipped with a data processing module, which can quickly analyze and process the image, spectrum, temperature and terrain data information transmitted by the monitoring sensor components, so as to obtain the degree of soil erosion and fertility, moisture content and soil type, water evaporation, and soil slope and slope information in the planting area, deduce the growth status of plants in the current planting area and the ecological environment of the soil, and better assist producers in formulating targeted agricultural management measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 This is an overall module structure diagram of a tropical agricultural ecological remote sensing monitoring device proposed by the present invention. DETAILED DESCRIPTION

[0019] 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.

[0020] Embodiment 1

[0021] See also Figure 1 A tropical agricultural ecological remote sensing monitoring device includes an unmanned aerial vehicle platform, a remote sensing unmanned aerial vehicle is arranged on the unmanned aerial vehicle platform, and a monitoring sensor component is loaded on the remote sensing unmanned aerial vehicle. The monitoring sensor component includes a high-resolution camera device, a hyperspectral sensor, a thermal infrared sensor and a laser radar (LiDAR). A data receiving module, a data processing module and a display module are arranged inside the unmanned aerial vehicle platform. The monitoring sensor component is mainly used to obtain soil images and data information. The data processing module can judge the soil ecological information according to the image and data. The monitoring sensor component and the data processing module cooperate to realize the growth status of plants in the planting area and the ecological environment of the soil, which can better assist producers to formulate targeted agricultural management measures.

[0022] High-resolution cameras are used to capture high-resolution images of the soil surface. The data processing module will pre-process the collected images, including denoising, image enhancement (such as rotation, flipping, scaling, etc.), radiation correction, atmospheric correction and geometric correction. Image analysis is then used to identify the degree of soil surface exposure and vegetation coverage, and indirectly determine the degree of soil erosion and fertility.

[0023] Image analysis includes feature extraction, type recognition and result calculation. Feature extraction includes color feature extraction, texture feature extraction, shape feature extraction and depth feature extraction. Color feature extraction is to extract the color difference characteristics of bare soil and vegetation through color histogram. Texture feature extraction is to extract the surface texture characteristics of bare soil and vegetation by using the gray-level co-occurrence matrix method. Shape feature extraction is to obtain the shape characteristics of bare soil and vegetation through edge detection and contour extraction. Deep feature extraction uses a pre-trained convolutional neural network (CNN) as a feature extractor to extract high-level features in the image. The extracted features are trained using random forests to establish a classification model. The pre-processed image is input into the trained classification model. The model will classify the pixels or areas in the image according to the extracted features, distinguish between bare soil and vegetation, and quantify the classification results. Calculate the bare soil area and vegetation coverage, so as to judge the degree of soil erosion and fertility.

[0024] The hyperspectral sensor can obtain the reflectance information of the soil in different spectral bands, and correct the collected spectral data through the data processing module, including radiation correction, atmospheric correction and geometric correction, and then extract the features related to soil moisture content, organic matter content and soil type from the preprocessed spectral data. Then, the extracted features are screened according to the relevance and importance, and the features with predictive ability for the target variables (such as soil moisture content, organic matter content and soil type) are selected as the feature group. A support vector machine is used to establish a prediction model based on the selected features, and the soil moisture content, organic matter content and soil type are inferred by analyzing the relevant features through the prediction model.

[0025] Hyperspectral sensors can also detect the "red edge position" of vegetation. By detecting the "red edge position" formed by the spectral characteristics of vegetation, the data processing module analyzes the changes in the "red edge position" of vegetation, thereby confirming the current growth status of vegetation and land pollution status.

[0026] The description of the red edge position includes the position of the red edge and the slope of the red edge. With the changes in chlorophyll content, biomass and phenology, the red edge position will move along the wavelength axis in the coordinates of the spectrum curve. When the chlorophyll content of the vegetation is high and the growth is vigorous, the "red edge" position will shift toward the infrared direction, which is called "red shift"; when the vegetation suffers from environmental stress, pests and diseases, or the chlorophyll content is reduced due to phenological changes, the "red edge" position will shift toward the blue light direction, which is called "blue shift". Vegetation coverage is related to the leaf area index. The higher the vegetation coverage, the greater the leaf area index, the greater the slope of the red edge, the better the corresponding vegetation growth state, and the red edge position will appear "red shift"; conversely, the red edge position will be correspondingly "blue shifted".

[0027] By analyzing the correlation between chlorophyll content or vegetation growth status parameters and soil heavy metal content, an inversion model for soil heavy metal content is established to reasonably predict and deduce the soil heavy metal pollution situation.

[0028] Embodiment 2

[0029] See also Figure 1 Thermal infrared sensors are used to measure the temperature of the soil surface. During the day, soil with high humidity may appear cold anomaly on thermal infrared images due to its large heat capacity and slow temperature increase. However, soil with low humidity will warm up quickly and appear hot anomaly. The soil surface temperature data monitored by the thermal infrared sensor can be used to identify the thermal anomaly area of ​​the soil. The cooling effect intensity of water evaporation can be calculated through the data processing module, thereby indirectly evaluating the water evaporation of the soil.

[0030] Laser radar (LiDAR) can provide high-precision terrain and vegetation height information. By using the LiDAR device to scan the target area and obtain a large amount of distance measurement data, the collected LiDAR data is then preprocessed. Then, based on the processed LiDAR data, an interpolation algorithm or a grid generation algorithm is used to generate a high-precision terrain map. Then, on the generated terrain map, a slope calculation tool or algorithm is used to extract the slope information. Finally, a soil erosion model, the Water Erosion Prediction and Assessment Model WEPP, is used to assess the soil erosion risk. Erosion factors include rainfall intensity, soil type, vegetation coverage and terrain characteristics. Rainfall intensity can be obtained from the local meteorological bureau, and soil type, vegetation coverage and terrain characteristics can be measured by the monitoring sensor components carried by the plant. The erosion factors are input into the Water Erosion Prediction and Assessment Model WEPP to obtain the soil erosion risk level, and then the soil status of the planting area is assessed.

[0031] The display module includes a display screen and a human-computer interaction device, which can display the data information collected by the monitoring sensor components on the drone and the data information analyzed and processed by the data processing module.

[0032] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A tropical agricultural ecological remote sensing monitoring device, including an unmanned aerial vehicle platform, characterized in that: A remote sensing drone is arranged on the drone platform, and a monitoring sensor assembly is loaded on the remote sensing drone. The monitoring sensor assembly includes a high-resolution camera, a hyperspectral sensor, a thermal infrared sensor and a laser radar (LiDAR). A data receiving module, a data processing module and a display module are arranged inside the drone platform. The monitoring sensor assembly is mainly used to obtain soil images and data information. The data processing module can judge soil ecological information according to the images and data. The monitoring sensor assembly and the data processing module cooperate to realize the monitoring of the growth status of plants in the planting area and the ecological environment of the soil, which can better assist producers to formulate targeted agricultural management measures.

2. A tropical agricultural ecological remote sensing monitoring device according to claim 1, characterized in that: The high-resolution camera equipment is used to capture high-resolution images of the soil surface. The data processing module will pre-process the collected images and then perform image analysis to identify the degree of exposure and vegetation coverage of the soil surface, and indirectly determine the degree of soil erosion and fertility.

3. A tropical agricultural ecological remote sensing monitoring device according to claim 2, characterized in that: Image analysis includes feature extraction, type recognition and result calculation. Feature extraction includes color feature extraction, texture feature extraction, shape feature extraction and depth feature extraction. Color feature extraction is to extract the color difference characteristics of bare soil and vegetation through color histogram. Texture feature extraction is to extract the surface texture characteristics of bare soil and vegetation by using the gray-level co-occurrence matrix method. Shape feature extraction is to obtain the shape characteristics of bare soil and vegetation through edge detection and contour extraction. Deep feature extraction is to extract high-level features in the image by using a pre-trained convolutional neural network (CNN) as a feature extractor, and the extracted features are trained using random forest to establish a classification model.

4. A tropical agricultural ecological remote sensing monitoring device according to claim 1, characterized in that: The hyperspectral sensor can obtain the reflectance information of the soil in different spectral bands, and correct the collected spectral data through the data processing module, and then extract the features related to the soil moisture content, organic matter content and soil type from the preprocessed spectral data, and then screen the extracted features according to the relevance and importance, and select the features with predictive ability for the target variables (such as soil moisture content, organic matter content and soil type) as the feature group, use the support vector machine, establish a prediction model according to the selected features, and infer the soil moisture content, organic matter content and soil type by analyzing the relevant features through the prediction model.

5. A tropical agricultural ecological remote sensing monitoring device according to claim 4, characterized in that: The hyperspectral sensor can also detect the "red edge position" of vegetation. By detecting the "red edge position" formed by the spectral characteristics of vegetation, the data processing module analyzes the changes in the "red edge position" of vegetation, thereby confirming the current growth status of vegetation and land pollution status.

6. A tropical agricultural ecological remote sensing monitoring device according to claim 1, characterized in that: The thermal infrared sensor is used to measure the temperature of the soil surface. The soil surface temperature data monitored by the thermal infrared sensor is used to calculate the cooling effect intensity of water evaporation through the data processing module, thereby indirectly evaluating the water evaporation of the soil.

7. The tropical agricultural ecological remote sensing monitoring device according to claim 1, characterized in that: The laser radar (LiDAR) can provide high-precision terrain and vegetation height information. By using the laser radar device to scan the target area and obtain a large amount of distance measurement data, the collected laser radar data is preprocessed, and then an interpolation algorithm or a grid generation algorithm is used to generate a high-precision terrain map based on the processed laser radar data. Then, on the generated terrain map, a slope calculation tool or algorithm is used to extract the slope information. Finally, a soil erosion model, the Water Erosion Prediction and Assessment Model WEPP, is used to assess the soil erosion risk. The erosion factors include rainfall intensity, soil type, vegetation coverage and terrain characteristics.

8. The tropical agricultural ecological remote sensing monitoring device according to claim 1, characterized in that: The display module includes a display screen and a human-computer interaction device, and can display the data information collected by the monitoring sensor component on the drone and the data information analyzed and processed by the data processing module.

Images