Non-contact soil humidity detection system and method based on polarization imaging

Abstract

The invention belongs to the technical field of intelligent perception and agricultural informatization, and particularly relates to a non-contact soil humidity detection system and method based on polarization imaging. The problems that an existing traditional soil humidity detection method is poor in real-time performance and strong in dependence on illumination, and a contact type detection system and method cannot represent a real physical state in a characterization mode and cannot conduct mobile measurement are solved. The system specifically comprises an illumination control unit, a polarization imaging unit connected with the illumination control unit through a data bus, an attitude measurement unit connected with the polarization imaging unit through a data bus, and a data processing and analysis unit connected with the polarization imaging unit and the attitude measurement unit through data buses. And the display and output module is connected with the data processing and analyzing unit through a data bus.

Description

Technical Field

[0001] This invention belongs to the field of intelligent sensing and agricultural information technology, and in particular relates to a non-contact soil moisture detection system and method based on polarization imaging. Background Technology

[0002] Soil moisture is a crucial parameter in agricultural management, hydrological regulation, ecological monitoring, and geological disaster early warning. Its spatial distribution and dynamic changes directly affect crop growth, water resource allocation, foundation stability, and ecosystem water cycle. Obtaining high-precision, real-time soil surface moisture information is a vital prerequisite for achieving precision irrigation, intelligent agricultural machinery operation planning, and road foundation safety monitoring. Currently, the main soil moisture detection methods used in the industry include contact sensor methods, remote sensing / spectral inversion methods, and thermal infrared imaging methods.

[0003] Contact sensor methods are currently the most mature technology, with representative devices including time-domain reflectometers, frequency-domain reflectometers, and resistive moisture sensors. These devices indirectly calculate moisture content by inserting probes into the soil and measuring its dielectric constant or resistivity. While this method offers high accuracy in laboratories or at fixed monitoring points, it suffers from challenges in large-area or mobile monitoring scenarios, including deployment difficulties, probe susceptibility to corrosion or damage, high maintenance costs, and limited spatial coverage, making it difficult to meet the demands of real-time and dynamic monitoring.

[0004] Remote sensing or spectral inversion methods utilize multispectral or hyperspectral sensors mounted on platforms such as satellites and drones to retrieve water content by measuring surface reflectance and shortwave infrared absorption characteristics. These methods can reflect soil moisture distribution at a regional scale, but their inversion results are easily affected by factors such as illumination angle, atmospheric scattering, surface roughness, and vegetation cover, leading to unstable measurement accuracy. Existing research shows that reflectance differences within the same plot can exceed 20% under different incident angles, causing significant moisture estimation bias. Furthermore, spectral methods primarily reflect the moisture state of the very shallow topsoil and are insufficiently responsive to moisture changes below this level, failing to reflect the true moisture gradient.

[0005] Thermal infrared imaging is based on the thermophysical relationship between surface temperature and moisture. It monitors the surface temperature field and inverts the humidity distribution using an infrared camera or thermal imager. Although this method does not require contact with the soil and has some usability at night, it is easily affected by external factors such as solar radiation, wind speed, shadows, soil thermal inertia, and vegetation cover, resulting in large fluctuations in the inversion results and making it difficult to achieve high-precision real-time measurements. Summary of the Invention

[0006] The problem that this invention aims to solve is that existing traditional soil moisture detection methods suffer from poor real-time performance, strong dependence on light, inability of contact detection systems and methods to characterize the true physical state, and inability to perform measurements while mobile.

[0007] To address the aforementioned problems, this invention provides a non-contact soil moisture detection system and method based on polarization imaging, characterized by: an illumination control unit 101, a polarization imaging unit 201 connected to the illumination control unit via a data bus, an attitude measurement unit 301 connected to the polarization imaging unit via a data bus, a data processing and analysis unit 401 connected to the polarization imaging unit and the attitude measurement unit via data buses respectively, and a display and output module 501 connected to the data processing and analysis unit via a data bus; The illumination control unit records the incident light intensity, incident angle and incident polarization direction, and provides stable illumination under low light conditions; The polarization imaging unit uses a polarization camera to simultaneously acquire four polarization angle reflection images, which are used to calculate the degree of polarization and polarization angle distribution.

[0008] The attitude measurement unit is used to record the camera attitude angle and distance from the ground surface in real time to correct errors caused by changes in viewing angle. At the same time, it records the geographic coordinates of the soil moisture measurement to support subsequent large-scale soil moisture distribution fusion.

[0009] The data processing and analysis unit is used for image preprocessing, polarization feature extraction, optical reflection modeling, humidity inversion, and real-time output of humidity estimates.

[0010] The display and output module is used to display the humidity distribution map in real time and upload the results to the monitoring platform via a wireless link.

[0011] Furthermore, the polarization imaging unit includes an adjustable gimbal and a polarization camera with multi-channel polarization synchronous acquisition function.

[0012] Furthermore, the four polarization angles of the polarization camera are set to 0°, 45°, 90°, and 135°, respectively.

[0013] Furthermore, the polarization camera is mounted on an adjustable gimbal, which is used to precisely control the incident angle of the polarization camera.

[0014] Furthermore, the illumination control unit includes a natural light monitoring module and a controllable polarization light source module.

[0015] Furthermore, the natural light monitoring module is used to record the incident light intensity, incident angle, and polarization direction in real time.

[0016] Furthermore, the controllable polarization light source module is used to automatically adjust the output power and polarization direction to maintain the stability of the reflected signal.

[0017] Furthermore, the polarization camera, adjustable gimbal, and data processing unit are set in the same reference coordinate system to maintain consistency between optical and attitude measurements.

[0018] Furthermore, the present invention also provides a non-contact soil moisture detection method based on polarization imaging, characterized by comprising the following steps: Step S31: Construction and acquisition of polarization imaging unit, synchronous imaging is achieved by using optical beam splitter or polarizer array, and time series is added after image acquisition; Step S32: Configure and compensate the illumination control unit to achieve consistent illumination conditions in different environments; Step S33: Integration and calibration of the attitude measurement unit, unifying the clock for attitude parameters and distance parameters, calculating geometric errors under different viewpoints, and ensuring the consistency of the inversion results; Step S34: The data processing and analysis unit is constructed to realize real-time calculation of image acquisition, feature extraction, reflection modeling and humidity inversion; Step S35, Display and Output: The calculated humidity distribution results are output to the display terminal via the communication interface, generating a color heat map and overlaying geographic coordinate information; Furthermore, the construction of the data processing and analysis unit in step S34 specifically includes the following steps: Step S341: Image preprocessing to ensure consistent pixel positions across the four channels; Step S342, polarization characteristic calculation, calculate the degree of polarization. and polarization angle ; Step S343: Optical reflection modeling, fusion of incident angles And the direction of illumination, to correct the polarization reflectivity; Step S344, Humidity inversion and result output, based on real-time calculations. Inversely calculate pixel-level water content Generate a humidity distribution matrix and statistical image.

[0019] The present invention provides a non-contact soil moisture detection system and method based on polarization imaging that can work under natural light conditions. It utilizes the polarization characteristics of reflected light combined with an optical reflection model to achieve stable and interference-resistant moisture inversion, and can be embedded in drones, sampling robots, monitoring robots, agricultural machinery or mobile platforms. Attached Figure Description

[0020] Figure 1 Block diagram of a non-contact soil moisture detection system based on polarization imaging; Figure 2Flowchart of a non-contact soil moisture detection method based on polarization imaging; Figure 3 Flowchart of the method based on data processing and analysis units.

[0021] The reference numerals in the attached figures are as follows: 101, illumination control unit; 201, polarization imaging unit; 301, attitude measurement unit; 401, data processing and analysis unit; 501, display and output module. Detailed Implementation

[0022] To better understand the purpose, structure, and function of this invention, the following detailed description of a non-contact soil moisture detection system and method based on polarization imaging, in conjunction with the accompanying drawings, is provided.

[0023] like Figure 1 This invention illustrates a non-contact soil moisture detection system based on polarization imaging, comprising an illumination control unit 101, a polarization imaging unit 201 connected to the illumination control unit via a data bus, an attitude measurement unit 301 connected to the polarization imaging unit via a data bus, a data processing and analysis unit 401 connected to the polarization imaging unit and the attitude measurement unit via data buses respectively, and a display and output module 501 connected to the data processing and analysis unit via a data bus.

[0024] Furthermore, the polarization imaging unit 201 includes a linear polarization or beam-splitting polarization camera 203 with multi-channel polarization synchronous acquisition function, which is mounted on the adjustable gimbal 202.

[0025] Specifically, the gimbal 202 is driven by a servo motor, which can precisely control the camera's angle of incidence within the range of 0° to 60°. .

[0026] Specifically, the polarization camera is configured with four polarization directions of 0°, 45°, 90°, and 135° to achieve synchronous imaging through optical beam splitting or polarizer arrays.

[0027] Specifically, the polarization camera 203, attitude measurement unit 301, and data processing and analysis unit 401 are placed in the same reference coordinate system to reduce time drift and maintain consistency between optical and attitude measurements.

[0028] During the acquisition process, the polarization camera 203 synchronously acquires reflection images in four polarization directions according to a set frame rate. And attach a timestamp for subsequent calculations.

[0029] Furthermore, the illumination control unit 101 includes: a natural light monitoring module 102 and a controllable polarization light source module 103.

[0030] Specifically, the natural light monitoring module 102 is used to record the incident light intensity in real time. Angle of incidence and polarization direction .

[0031] Specifically, the controllable polarization light source module 103 is used to automatically start when the light is insufficient or changes drastically, so as to provide the target area with stable intensity and adjustable polarization direction lighting.

[0032] The natural light monitoring module 102 transmits the illumination parameters to the data processing and analysis unit 401 via a bus for modeling and compensation.

[0033] For example, when the system of the present invention detects that the ambient illuminance is lower than the threshold, the controllable polarization light source module 103 automatically adjusts the output power and polarization direction to maintain the stability of the reflected signal.

[0034] Furthermore, the attitude measurement unit 301, which is set in the same reference coordinate system as the polarization camera 203, includes: an inertial measurement unit 302, a laser rangefinder 303, and a positioning module 304.

[0035] Specifically, the inertial measurement unit 302 outputs the pitch, roll, and yaw angle information of the polarization camera 203 in real time.

[0036] Specifically, the laser rangefinder 303 measures the distance d from the lens of the polarization camera 203 to the ground surface.

[0037] Specifically, the positioning module records geographic coordinates in real time.

[0038] The attitude data output in real time by the inertial measurement unit 302 and the distance parameters from the lens of the polarization camera 203 to the ground surface measured by the laser rangefinder 303 are synchronized by a unified clock and then input into the data processing and analysis unit 401 via a data bus to calculate the incident angle. Reflection angle It also corrects geometric errors under different viewpoints. Through this correction process, the consistency of inversion results can be guaranteed under multi-angle imaging conditions.

[0039] Furthermore, the data processing and analysis unit 401 is used to realize real-time calculations of image acquisition and preprocessing, feature extraction and polarization calculation, optical reflection modeling and humidity inversion.

[0040] Specifically, the data processing and analysis unit 401 uses an FPGA programmable gate array or an embedded GPU platform as its core.

[0041] Specifically, the calculation method of the data processing and analysis unit 401 includes steps S11 to S14.

[0042] In step S11, the image acquisition and preprocessing involves filtering and denoising the four-channel polarized image acquired by the polarization camera 203, correcting geometric distortion, and spatially registering it to ensure that the pixel positions of each channel are consistent.

[0043] Step S12, the feature extraction and polarization calculation, are specifically based on the formula: , Calculate the degree of polarization and polarization angle After polarization characteristic map, incident angle Light intensity The subsequent model is passed in synchronously.

[0044] Step S13, the optical reflection modeling is based on Fresnel reflection theory to establish a reflectivity model: , The incident angle obtained by the attitude measurement unit 301 and direction of light The polarization reflectivity is corrected.

[0045] Step S14, when the soil moisture content As the surface water film thickness and refractive index n increase, the difference between parallel and perpendicular reflectivity decreases, thus... The decline was addressed by establishing an empirical model using calibration experiments. , in, , , To calibrate the obtained regression parameters, the system calculates them in real time. Inversely calculate pixel-level water content Generate a humidity distribution matrix and statistical image.

[0046] Furthermore, the display and output module 501 outputs the calculated humidity distribution results to the display terminal via the communication interface, generating a color heat map and overlaying geographic coordinate information. Simultaneously, the result data is uploaded to a host computer or agricultural cloud platform via the wireless communication module, enabling remote monitoring and multi-regional data fusion. The system supports automatic storage and historical trend querying, and can be used for humidity time-series analysis and regional soil moisture diagnosis.

[0047] Furthermore, to achieve high-precision humidity inversion, this invention introduces a multi-dimensional joint calibration and adaptive optimization mechanism before system deployment and during operation. The specific steps of this mechanism include S21~S22: Step S21: Construction of multidimensional calibration experiment.

[0048] For example, representative soil samples were selected from the experimental site, and their moisture content was controlled. The variation is graded within the range of 0% to 40%, and multiple incident angles are set for each sample. , direction of light polarization and band parameters.

[0049] The spectral parameters can be selected using a multi-band LED light source in the range of 400~1000 mm, and polarization image sequences are acquired synchronously. This is achieved by analyzing different incident geometries, illumination conditions, and moisture states. The mean and spectral reflectance are registered to form a high-dimensional calibration dataset. : , in, The wavelength is the center wavelength of the light source.

[0050] Step S22, data-driven regression modeling.

[0051] Specifically, based on the calibrated dataset, a regularized nonlinear regression or lightweight machine learning model is used for fitting. With moisture content The non-linear relationship between them.

[0052] From the above, we can conclude that the model expression can be formalized as: , in, This is the model parameter set, which is automatically optimized during the training process.

[0053] Specifically, this model is better than the traditional exponential fit. It can better describe the nonlinear saturation characteristics of high humidity and extremely low humidity regions.

[0054] like Figure 2 The present invention provides a non-contact soil moisture detection method based on polarization imaging, comprising steps S31-S34. Step S31, Construction and acquisition of polarization imaging unit; Specifically, a polarization camera mounted on an adjustable gimbal simultaneously acquires images in four polarization directions.

[0055] Step S32, configuration and compensation of the illumination control unit; Specifically, the illumination control unit records the incident light intensity in real time. The incident angle and polarization direction are transmitted to the data processing and analysis unit 401 via a data bus.

[0056] Step S33, Integration and calibration of the attitude measurement unit; Specifically, the attitude measurement unit records in real time the pitch, roll, and yaw angle information of the polarization camera 203, the distance from the lens to the ground surface measured by the laser rangefinder 303, and the theoretical coordinates recorded by the positioning module. d It is then transmitted to the data processing and analysis unit 401 via the data bus.

[0057] Step S34, Construction of the data processing and analysis unit; Specifically, step S34 in this embodiment includes steps S341 to S345: Step S341, Image preprocessing; Specifically, the acquired polarization images are filtered for noise reduction, geometric distortion correction, and spatial registration to ensure that the pixel positions of each channel are consistent.

[0058] Step S342, polarization characteristic calculation; Specifically, the degree of polarization is calculated based on the four polarization directions set by the polarization camera 203. and polarization angle .

[0059] Step S343, optical reflection modeling; Specifically, based on the polarization characteristic map and the incident angle Illumination parameters and degree of polarization and polarization angle Establish a reflectivity model.

[0060] Step S344: Humidity inversion and result output; Specifically, an experimental model was established using calibration experiments to calculate the humidity distribution matrix and statistical images.

[0061] Step S35, Display and Output; Specifically, it displays a humidity distribution map in real time and uploads the results to a host computer or other monitoring platform via a wireless link.

[0062] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A non-contact soil moisture detection system based on polarization imaging, characterized in that: The illumination control unit (101), the polarization imaging unit (201) connected to the illumination control unit via a data bus, the attitude measurement unit (301) connected to the polarization imaging unit via a data bus, the data processing and analysis unit (401) connected to the polarization imaging unit and the attitude measurement unit via a data bus respectively, and the display and output module (501) connected to the data processing and analysis unit via a data bus. The illumination control unit (101) records the incident light intensity, incident angle and incident polarization direction, and provides stable illumination under low light conditions; The polarization imaging unit (201) uses a polarization camera to simultaneously acquire four polarization angle reflection images for calculating the degree of polarization and polarization angle distribution; The attitude measurement unit (301) is used to record the camera attitude angle and distance from the ground surface in real time, to correct the error caused by the change of viewing angle, and to record the geographic coordinates of the soil moisture measurement, supporting the subsequent large-scale soil moisture distribution fusion. The data processing and analysis unit (401) is used for image preprocessing, polarization feature extraction, optical reflection modeling, humidity inversion, and real-time output of humidity estimates. The display and output module (501) is used to display the humidity distribution map in real time and upload the results to the monitoring platform via a wireless link.

2. The non-contact soil moisture detection system based on polarization imaging according to claim 1, characterized in that: The polarization imaging unit (201) includes an adjustable gimbal and a polarization camera with multi-channel polarization synchronous acquisition function.

3. The non-contact soil moisture detection system based on polarization imaging according to claim 1, characterized in that: The polarization camera has four polarization angles set to 0°, 45°, 90°, and 135°, respectively.

4. The non-contact soil moisture detection system and method based on polarization imaging according to claim 1, characterized in that: The polarization camera is mounted on an adjustable gimbal, which is used to precisely control the incident angle of the polarization camera.

5. The non-contact soil moisture detection system based on polarization imaging according to claim 1, characterized in that: The illumination control unit (101) includes a natural light monitoring module and a controllable polarization light source module.

6. The non-contact soil moisture detection system based on polarization imaging according to claim 5, characterized in that... The natural light monitoring module is used to record the incident light intensity, incident angle, and polarization direction in real time.

7. The non-contact soil moisture detection system based on polarization imaging according to claim 5, characterized in that: The controllable polarization light source module is used to automatically adjust the output power and polarization direction to maintain the stability of the reflected signal.

8. The non-contact soil moisture detection system based on polarization imaging according to claim 1, characterized in that: The polarization camera, adjustable gimbal, and data processing unit are set in the same reference coordinate system to maintain consistency between optical and attitude measurements.

9. A non-contact soil moisture detection method based on polarization imaging, characterized in that, Includes the following steps: Step S31: Construction and acquisition of polarization imaging unit, synchronous imaging is achieved by using optical beam splitter or polarizer array, and time series is added after image acquisition; Step S32: Configure and compensate the illumination control unit to achieve consistent illumination conditions in different environments; Step S33: Integration and calibration of the attitude measurement unit, unifying the clock for attitude parameters and distance parameters, calculating geometric errors under different viewpoints, and ensuring the consistency of the inversion results; Step S34: The data processing and analysis unit is constructed to realize real-time calculation of image acquisition, feature extraction, reflection modeling and humidity inversion; Step S35, Display and Output: The calculated humidity distribution result is output to the display terminal via the communication interface, generating a color heat map and overlaying geographic coordinate information.

10. The non-contact soil moisture detection method based on polarization imaging according to claim 1, characterized in that: The construction of the data processing and analysis unit in step S34 specifically includes the following steps: Step S341: Image preprocessing to ensure consistent pixel positions across the four channels; Step S342, polarization characteristic calculation, calculate the degree of polarization. and polarization angle ; Step S343: Optical reflection modeling, fusion of incident angles And the direction of illumination, to correct the polarization reflectivity; Step S344, Humidity inversion and result output, based on real-time calculations. Inversely calculate pixel-level water content Generate a humidity distribution matrix and statistical image.

Images