Portable diagnosis device and method for relative index of crop canopy water physiological condition
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-14
- Publication Date
- 2026-08-11
AI Technical Summary
传统单点测量法（如叶温枪）无法代表完整冠层且受角度干扰大；无人机遥感虽然覆盖广，但存在成本高、易受大气吸收干扰、像素混合严重等问题
[0008]与现有技术中相比，本发明的有益效果在于，以可伸缩支架为基础，集成水平指示模块的手持端，以及集成气温、热红外与可见光传感器的传感器端，通过平行光轴设计与水平指示模块协同确立垂直向下（Nadir View）的标准化采集模式，利用“支架+水准泡”实现物理层面的垂直观测，有效消除了传感器倾角引入的阴影和背景误差，保证了高分辨率多模态数据的空间匹配精度。
Smart Images
Figure CN122545394A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a portable diagnostic device and method for the relative index of crop canopy water physiological status in the fields of smart agriculture, agricultural remote sensing and vegetation physiological monitoring. Background Technology
[0002] Crop water physiological status is a core indicator reflecting its health and the degree of drought stress. Traditional single-point measurement methods (such as leaf temperature guns) cannot represent the entire canopy and are greatly affected by angle interference; although UAV remote sensing has wide coverage, it suffers from high cost, susceptibility to atmospheric absorption interference, and severe pixel mixing. Currently, when calculating the relative water index, traditional methods rely excessively on complex theoretical meteorological formulas or additional dry / wet reference boards placed in the field, limiting the portability and diagnostic accuracy of field operations. Summary of the Invention
[0003] The purpose of this invention is to provide a portable diagnostic device and method for the relative index of crop canopy water physiological status. It aims to achieve standardized data collection through hardware structure and to achieve high-precision real-time diagnosis of water physiological status without the need for external reference through an algorithm of "multi-location sampling and full-sample driving".
[0004] To achieve the above objectives, the present invention provides a portable diagnostic device for the relative index of crop canopy water physiological status, including a retractable connecting bracket. One end of the retractable connecting bracket is a handheld control end, on which a level indicator module for indicating the horizontal status of the device is installed. The other end of the retractable connecting bracket is a sensor end, on which an air temperature sensor, a full-radiation thermometric thermal infrared sensor and a visible light image sensor are integrated.
[0005] As a further improvement of the present invention, the optical axes of the lenses of the thermal infrared sensor and the visible light image sensor are parallel, and during measurement, the handheld control end is adjusted by the level indicator module to keep the support in a horizontal state so as to obtain vertically downward crop canopy information.
[0006] As a further improvement of the present invention, the handheld control terminal also integrates a data acquisition and processing host and a power module.
[0007] As a further improvement of the present invention, the level indicator module is a spirit level.
[0008] Compared with the prior art, the beneficial effects of the present invention are that, based on a retractable bracket, a handheld end integrating a horizontal indicator module and a sensor end integrating air temperature, thermal infrared and visible light sensors are established through a parallel optical axis design and in collaboration with the horizontal indicator module to establish a standardized vertical downward (Nadir View) acquisition mode. The "bracket + bubble level" is used to achieve vertical observation at the physical level, effectively eliminating shadows and background errors introduced by the sensor tilt angle, and ensuring the spatial matching accuracy of high-resolution multimodal data.
[0009] To achieve the above objectives, the present invention also provides a method for diagnosing the relative index of crop canopy water physiological status, comprising the following steps: S1. Data Acquisition: Keep the support horizontal at multiple spatial locations within the area to be measured, and simultaneously acquire ambient air temperature Ta, RGB images, and thermal infrared images Tc; S2, Feature Calculation: The data acquisition and processing host calculates the normalized greenness index (NGI) and real-time temperature difference (Ts) for all pixels; S3, Canopy Mask Segmentation: The data acquisition and processing host generates a binary canopy mask image through threshold segmentation using NGI; S4. Boundary Fitting: The upper boundary Tmax and lower boundary Tmin of the scatter plot are nonlinearly fitted using a quadratic equation to obtain the dry limit function Tmax and the wet limit function Tmin. S5. Index Calculation and Output: Based on the dry limit function and the wet limit function, calculate the crop water physiological status relative index CWCI pixel by pixel, and output a high-resolution diagnostic map after background removal.
[0010] As a further improvement to the present invention, the specific content of S1 is as follows. The operator holds a retractable support and randomly selects N sampling points within the target field. By observing the bubble level at the handheld end, the operator adjusts the retractable support to a horizontal position so that the optical axis of the sensor at the end of the support is vertically downward. The sensor simultaneously acquires N sets of data, each set containing an RGB image, a resampled and aligned thermal infrared image Tc, and the real-time ambient temperature Ta, and transmits them to the data acquisition and processing host.
[0011] As a further improvement to the present invention, S2 is specifically described below. The data acquisition and processing host uses visible light images to calculate all canopy pixels and extract their normalized greenness index NGI=g / (r+g+b) and real-time temperature difference value Ts = Tc - Ta; where r, g and b are the DN values of each channel of the RGB image acquired by the camera.
[0012] As a further improvement to the present invention, S3 is specifically described below. From all canopy pixels, a preset number of sample pixels are randomly selected, and the (NGI,Ts) values of these points are projected onto a two-dimensional coordinate system. A global two-dimensional scatter plot is constructed with NGI as the horizontal axis and temperature difference (Tc-Ta) as the vertical axis. Since the transpiration and cooling capacity of crops under different greenness is different, the scatter plot shows obvious envelope characteristics.
[0013] As a further improvement to the present invention, S4 is specifically described below. The pixels collected from the images are used to form a scatter plot. Then, a quadratic equation is used to fit the scatter plot to derive the upper limit function Tmax and the lower limit function Tmin. These represent the relationship between temperature and greenness under extreme dry and wet conditions of vegetation during the current data collection task. Tmax represents the theoretical upper limit of extreme water scarcity, and Tmin represents the theoretical lower limit of sufficient water. Dry limit function: Tmax = a1 × NGI 2 + b1 × NGI + c1; Wet limiting function: Tmin = a² × NGI 2 + b2× NGI + c2; Where a1, b1, and c1 are the parameters fitted to the upper boundary of the scatter points, and a2, b2, and c2 are the parameters fitted to the lower boundary of the scatter points.
[0014] As a further improvement to the present invention, S5 is specifically described below. For any pixel in the sampled image, its NGI value is read, and the corresponding boundary reference Tmax and Tmin images are obtained by substituting them into Tmax and Tmin. Then, based on the corresponding thermal infrared canopy temperature image and air temperature, the CWCI value of the pixel is calculated using the following formula: CWCI = ((Tc - Ta) - Tmin) / (Tmax - Tmin) Among them, a CWCI close to 1 indicates that the point is under severe drought stress, while a CWCI close to 0 indicates sufficient water and healthy physiological state; The final result is a high-resolution CWCI mask image that not only filters out the soil background but also preserves the subtle moisture differences between leaves within the canopy, providing data support for precision irrigation.
[0015] Compared with existing technologies, the advantages of this invention lie in its use of a canopy mask generated from visible light images and a normalized greenness index. Based on image data obtained from all sampling locations in this measurement task, pixels are randomly selected within the canopy mask to establish a greenness-temperature difference (Tc-Ta) scatter plot. A quadratic equation is then used to fit the dry / wet limit boundary function in real time. Finally, the relative index of crop water physiological status is calculated pixel-by-pixel, and a high-resolution diagnostic map is output. This invention requires no external dry / wet reference board or complex meteorological formulas and possesses advantages such as low cost, portability, standardization, and data-driven adaptability. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the overall structure of the device of the present invention.
[0017] Figure 2 This is a schematic diagram of the diagnostic method of the present invention.
[0018] Figure 3 This is the original RGB visible light image of the crop canopy acquired in this invention.
[0019] Figure 4 This is a thermal infrared temperature image of the crop canopy collected according to the present invention.
[0020] Figure 5 This is an image of the Normalized Greenness Index (NGI) calculated according to the present invention.
[0021] Figure 6 The image is a binarized canopy mask generated by threshold segmentation in this invention.
[0022] Figure 7 The two-dimensional scatter plot of NGI versus temperature difference (Tc-Ta) and the fitted boundary curve are constructed for this invention.
[0023] Figure 8 This is the high-resolution canopy water physiological status relative index (CWCI) mask image that is the final output of this invention.
[0024] The components include: 1. Handheld control terminal; 2. Bubble level; 3. Telescopic connecting bracket; 4. Temperature sensor; 5. Thermal infrared sensor; and 6. Visible light image sensor. Detailed Implementation
[0025] The present invention will be further described below with reference to the accompanying drawings: like Figure 1 The portable diagnostic device for the relative index of crop canopy water physiological status shown includes a retractable connecting bracket 3. One end of the retractable connecting bracket 3 is a handheld control terminal 1, on which a horizontal indicator module for indicating the horizontal status of the device is installed. The other end of the retractable connecting bracket 3 is a sensor terminal, on which an air temperature sensor 4, a full-radiation thermometric thermal infrared sensor 5, and a visible light image sensor 6 are integrated.
[0026] The optical axes of the lenses of the thermal infrared sensor 5 and the visible light image sensor 6 are parallel. During measurement, the handheld control terminal 1 is adjusted with the assistance of the level indicator module to keep the support horizontal, thereby acquiring vertically downward crop canopy information. The handheld control terminal 1 also integrates a data acquisition and processing host and a power module. The level indicator module is a bubble level 2.
[0027] like Figure 2-8 The method for diagnosing the relative index of crop canopy water physiological status, as shown, includes the following steps: S1. Data Acquisition: Keep the support horizontal at multiple spatial locations within the area to be measured, and simultaneously acquire ambient air temperature Ta, RGB images, and thermal infrared images Tc; The operator holds a retractable support and randomly selects N sampling points within the target field. By observing the bubble level at the handheld end, the operator adjusts the retractable support to a horizontal position so that the optical axis of the sensor at the end of the support is vertically downward. The sensor simultaneously acquires N sets of data, each set containing an RGB image, a resampled and aligned thermal infrared image Tc, and the real-time ambient temperature Ta, and transmits them to the data acquisition and processing host.
[0028] S2, Feature Calculation: The data acquisition and processing host calculates the normalized greenness index (NGI) and real-time temperature difference (Ts) for all pixels; The data acquisition and processing host uses visible light images to calculate all canopy pixels and extract their normalized greenness index NGI=g / (r+g+b) and real-time temperature difference value Ts = Tc - Ta; where r, g and b are the DN values of each channel of the RGB image acquired by the camera.
[0029] S3, Canopy Mask Segmentation: The data acquisition and processing host generates a binary canopy mask image through threshold segmentation using NGI; From all canopy pixels, a preset number of sample pixels are randomly selected, and the (NGI,Ts) values of these points are projected onto a two-dimensional coordinate system. A global two-dimensional scatter plot is constructed with NGI as the horizontal axis and temperature difference (Tc-Ta) as the vertical axis. Since the transpiration and cooling capacity of crops under different greenness is different, the scatter plot shows obvious envelope characteristics.
[0030] S4. Boundary Fitting: The upper boundary Tmax and lower boundary Tmin of the scatter plot are nonlinearly fitted using a quadratic equation to obtain the dry limit function Tmax and the wet limit function Tmin. The pixels collected from the images are used to form a scatter plot. Then, a quadratic equation is used to fit the scatter plot to derive the upper limit function Tmax and the lower limit function Tmin. These represent the relationship between temperature and greenness under extreme dry and wet conditions of vegetation during the current data collection task. Tmax represents the theoretical upper limit of extreme water scarcity, and Tmin represents the theoretical lower limit of sufficient water. Dry limit function: Tmax = a1 × NGI 2 + b1 × NGI + c1; Wet limiting function: Tmin = a² × NGI 2 + b2× NGI + c2; Where a1, b1, and c1 are the parameters fitted to the upper boundary of the scatter points, and a2, b2, and c2 are the parameters fitted to the lower boundary of the scatter points.
[0031] S5. Index Calculation and Output: Based on the dry limit function and the wet limit function, calculate the crop water physiological status relative index CWCI pixel by pixel, and output a high-resolution diagnostic map after background removal.
[0032] For any pixel in the sampled image, its NGI value is read, and the corresponding boundary reference Tmax and Tmin images are obtained by substituting them into Tmax and Tmin. Then, based on the corresponding thermal infrared canopy temperature image and air temperature, the CWCI value of the pixel is calculated using the following formula: CWCI = ((Tc - Ta) - Tmin) / (Tmax - Tmin) Among them, a CWCI close to 1 indicates that the point is under severe drought stress, while a CWCI close to 0 indicates sufficient water and healthy physiological state; The final result is a high-resolution CWCI mask image that not only filters out the soil background but also preserves the subtle moisture differences between leaves within the canopy, providing data support for precision irrigation.
[0033] In this invention, a telescopic connecting bracket ranging from 0.5 meters to 2.5 meters is used to adapt to crops of different heights; the level monitoring module is a bubble level set at the handheld end of the bracket to assist the operator in adjusting the bracket to a horizontal state in real time; the sensor end integrates an air temperature sensor, a full-radiation thermal infrared sensor, and a high-resolution RGB image sensor, and the optical axis of the thermal infrared sensor is parallel to that of the RGB sensor to ensure that vertical downward (Nadir View) data acquisition is achieved when the bracket is horizontal.
[0034] This embodiment uses the water physiology diagnosis of a cornfield as an example for detailed explanation.
[0035] Step 1: Standardized data collection at multiple points across the entire field The operator holds the device and randomly selects 100 sampling points within the target field. By observing the bubble level at the handheld end, the 0.5-2.5 meter telescopic support is adjusted to a horizontal position, ensuring the sensor's optical axis at the end of the support is vertically downward. The sensor simultaneously acquires 100 sets of data, each set containing one RGB image (e.g., ...). Figure 3 As shown), a resampled and aligned thermal infrared image (such as...) Figure 4 (as shown) and the real-time ambient temperature Ta (the daily average is 38℃).
[0036] Step 2: Feature Calculation The data acquisition and processing host uses visible light images to calculate all canopy pixels, extracting their Normalized Greenness Index (NGI) = g / (r+g+b) and real-time temperature difference value Ts = Tc – Ta, to obtain a Normalized Greenness Index (NGI) image (e.g., ...). Figure 5 (As shown).
[0037] Step 3: Canopy Mask Segmentation: The data acquisition and processing host generates a binary canopy mask image (e.g., using NGI's threshold segmentation) Figure 6 (as shown) A predetermined number of sample pixels are randomly selected from all canopy pixels. The (NGI, Ts) values of these points are projected onto a two-dimensional coordinate system. A global two-dimensional scatter plot is constructed with NGI as the horizontal axis and temperature difference (Tc-Ta) as the vertical axis. Due to the different transpiration and cooling capabilities of crops under different greenness levels, the scatter plot exhibits obvious envelope characteristics. Step 3: Fitting the boundary function of the entire sample The system randomly selected 20,000 sample pixels from all canopy pixels at 100 locations. The (NGI, Ts) values of these 20,000 points were projected onto a two-dimensional coordinate system. Due to the varying transpiration cooling capacity of crops under different greenness levels, the scatter plot exhibited a clear envelope characteristic. The system used a quadratic equation to fit the upper edge (representing the theoretical upper limit of extreme water scarcity) and lower edge (representing the theoretical lower limit of sufficient water) of the scatter plot (e.g., ...). Figure 7 (as shown) Dry limit function: Tmax = -11.71 × NGI 2 + 8.28 × NGI + 5.69 Wet limiting function: Tmin = 3.61 × NGI 2 - 5.53 × NGI + 3.17 Step 4: Calculate the relative index per pixel For any pixel in the 100 images, its NGI value is read and substituted into the two functions mentioned above to obtain the corresponding boundary reference images Tmax and Tmin. Then, based on the corresponding thermal infrared canopy temperature image and air temperature, the CWCI value of the pixel is calculated using the following formula: CWCI = ((Tc - Ta) - Tmin) / (Tmax - Tmin) Wherein, a CWCI close to 1 indicates that the point is under severe drought stress (red area in the figure), and a CWCI close to 0 indicates sufficient water and healthy physiological state (green area in the figure).
[0038] Step 5: Output Results The system ultimately generates 100 high-resolution CWCI mask images (e.g. Figure 8 (As shown in the image). This figure not only filters out the soil background but also preserves the subtle moisture differences between leaves within the canopy, providing data support for precision irrigation.
[0039] This invention establishes a standardized, vertically downward (Nadir View) acquisition mode through a parallel optical axis design and a horizontal indicator module. The diagnostic method utilizes a canopy mask generated from visible light images and a normalized greenness index. Based on image data acquired from all sampling locations in this measurement task, pixels are randomly selected within the canopy mask to create a greenness-temperature difference (Tc-Ta) scatter plot. A quadratic equation is used to fit the dry / wet limit boundary function in real time. Finally, the relative index of crop water physiological status is calculated pixel-by-pixel, and a high-resolution diagnostic image is output.
[0040] This invention overcomes the problems of insufficient data and poor representativeness in a single shooting by aggregating image data from multiple sampling locations for unified boundary fitting, and can accurately capture the physiological extreme values of an entire field under specific meteorological conditions. The algorithm is entirely based on the local data distribution of the image itself, eliminating the need for additional dry / wet reference boards, and automatically reduces the impact of light fluctuations and drastic temperature changes on the calculation baseline, greatly improving field operation efficiency. Utilizing a "support + bubble level" system to achieve vertical observation at the physical level effectively eliminates shadows and background errors introduced by sensor tilt angles, ensuring the spatial matching accuracy of high-resolution multimodal data.
[0041] This invention does not require an external wet / dry reference board or complex meteorological formulas, and has advantages such as low cost, portability, standardization, and data-driven self-adaptation.
[0042] This invention is not limited to the above embodiments. Based on the technical solutions disclosed herein, those skilled in the art can make some substitutions and modifications to some of the technical features without creative effort, and all such substitutions and modifications are within the protection scope of this invention.
Claims
1. A portable diagnostic device for the relative index of crop canopy water physiological status, characterized in that: It includes a retractable connecting bracket, one end of which is a handheld control end, on which a level indicator module for indicating the horizontal status of the device is installed, and the other end of the retractable connecting bracket is a sensor end, which integrates an air temperature sensor, a full-radiation temperature measurement type thermal infrared sensor and a visible light image sensor.
2. The portable diagnostic device for the relative index of crop canopy water physiological status according to claim 1, characterized in that: The optical axes of the lenses of the thermal infrared sensor and the visible light image sensor are parallel, and during measurement, the handheld control end is adjusted with the assistance of the horizontal indicator module to keep the support in a horizontal state in order to obtain vertically downward crop canopy information.
3. The portable diagnostic device for the relative index of crop canopy water physiological status according to claim 2, characterized in that: The handheld control unit also integrates a data acquisition and processing host and a power module.
4. The portable diagnostic device for the relative index of crop canopy water physiological status according to claim 3, characterized in that: The level indicator module is a spirit level.
5. A method for diagnosing the relative index of crop canopy water physiological status, characterized by: The portable diagnostic device for the relative index of crop canopy water physiological status according to any one of claims 1-4 includes the following steps: S1. Data Acquisition: Keep the support horizontal at multiple spatial locations within the area to be measured, and simultaneously acquire ambient air temperature Ta, RGB images, and thermal infrared images Tc; S2, Feature Calculation: The data acquisition and processing host calculates the normalized greenness index (NGI) and real-time temperature difference (Ts) for all pixels; S3, Canopy Mask Segmentation: The data acquisition and processing host generates a binary canopy mask image through threshold segmentation using NGI; S4. Boundary Fitting: The upper boundary Tmax and lower boundary Tmin of the scatter plot are nonlinearly fitted using a quadratic equation to obtain the dry limit function Tmax and the wet limit function Tmin. S5. Index Calculation and Output: Based on the dry limit function and the wet limit function, calculate the crop water physiological status relative index CWCI pixel by pixel, and output a high-resolution diagnostic map after background removal.
6. The method for diagnosing the relative index of crop canopy water physiological status according to claim 5, characterized in that: The specific content of S1 is as follows: The operator holds a retractable support and randomly selects N sampling points in the target field. By observing the bubble level at the handheld end, the operator adjusts the retractable support to a horizontal state so that the optical axis of the sensor at the end of the support is vertically downward. The sensor acquires N sets of data simultaneously. Each set includes an RGB image, a resampled and aligned thermal infrared image Tc, and the real-time ambient temperature Ta, and transmits them to the data acquisition and processing host.
7. The method for diagnosing the relative index of crop canopy water physiological status according to claim 6, characterized in that: The specific details of S2 are as follows. The data acquisition and processing host uses visible light images to calculate all canopy pixels and extract their normalized greenness index NGI=g / (r+g+b) and real-time temperature difference value Ts = Tc - Ta; where r, g and b are the DN values of each channel of the RGB image acquired by the camera.
8. The method for diagnosing the relative index of crop canopy water physiological status according to claim 7, characterized in that: The specific details of S3 are as follows. From all canopy pixels, a preset number of sample pixels are randomly selected, and the (NGI, Ts) values of these points are projected onto a two-dimensional coordinate system. A global two-dimensional scatter plot is constructed with NGI as the horizontal axis and temperature difference (Tc-Ta) as the vertical axis. Since the transpiration and cooling capacity of crops under different greenness is different, the scatter plot shows obvious envelope characteristics.
9. The method for diagnosing the relative index of crop canopy water physiological status according to claim 8, characterized in that: The specific details of S4 are as follows. The pixels collected from the images are used to form a scatter plot. Then, a quadratic equation is used to fit the scatter plot to derive the upper limit function Tmax and the lower limit function Tmin. These represent the relationship between temperature and greenness under extreme dry and wet conditions of vegetation during the current data collection task. Tmax represents the theoretical upper limit of extreme water scarcity, and Tmin represents the theoretical lower limit of sufficient water. Dry limit function: Tmax = a1 × NGI 2 + b1 × NGI + c1; Wet limiting function: Tmin = a² × NGI 2 + b2× NGI + c2; Where a1, b1, and c1 are the parameters fitted to the upper boundary of the scatter points, and a2, b2, and c2 are the parameters fitted to the lower boundary of the scatter points.
10. The method for diagnosing the relative index of crop canopy water physiological status according to claim 9, characterized in that: The specific details of S5 are as follows. For any pixel in the sampled image, its NGI value is read, and the corresponding boundary reference Tmax and Tmin images are obtained by substituting them into Tmax and Tmin. Then, based on the corresponding thermal infrared canopy temperature image and air temperature, the CWCI value of the pixel is calculated using the following formula: CWCI = ((Tc - Ta) - Tmin) / (Tmax - Tmin) Among them, a CWCI close to 1 indicates that the point is under severe drought stress, while a CWCI close to 0 indicates sufficient water and healthy physiological state; The final result is a high-resolution CWCI mask image that not only filters out the soil background but also preserves the subtle moisture differences between leaves within the canopy, providing data support for precision irrigation.