A rice biomass continuous monitoring method based on unmanned aerial vehicle low-altitude remote sensing image and phenology
By combining low-altitude remote sensing images from UAVs with phenological data, a continuous monitoring model for rice biomass was constructed, which solved the problem of temporal continuity in rice biomass monitoring in traditional methods. This enabled dynamic monitoring and efficient estimation throughout the entire growth period, reduced costs, and enhanced the applicability of the model.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING UNIV OF CIVIL ENG & ARCHITECTURE
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-21
AI Technical Summary
Existing methods for monitoring rice biomass lack temporal continuity. Traditional models are typically developed and tested at specific points in time, making it impossible to achieve continuous monitoring of rice biomass, and are also labor-intensive.
By combining low-altitude remote sensing images and phenological data from UAVs, and by collecting multi-band spectral images and visible light images, spectral preprocessing and vegetation index calculation are performed. By combining ground data and phenological data, a continuous monitoring model for rice biomass is constructed, and continuous monitoring of biomass is achieved by using phenological indicators such as growth accumulated temperature and number of days after sowing.
It enables continuous and dynamic monitoring of rice biomass throughout its entire growth period, improving the temporal completeness and accuracy of the estimation, reducing labor costs, enhancing the universality and robustness of the model, and providing technical support for smart agricultural management.
Smart Images
Figure CN122435487A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of agricultural remote sensing technology, specifically to a method for continuous monitoring of rice biomass based on a combination of UAV low-altitude remote sensing images and phenological phenomena. Background Technology
[0002] Real-time monitoring of aboveground biomass is crucial for understanding crop growth, estimating crop yield, and carbon dynamics. Traditional methods for measuring aboveground biomass are time-consuming and labor-intensive. Non-destructive unmanned aerial vehicle (UAV) remote sensing technology, with its large spatial coverage, has become a promising method for monitoring rice biomass. The most common approach is to extract spectral information from the visible and near-infrared regions of spectral images to generate VIs for aboveground biomass estimation studies. However, existing models are typically developed and tested at specific time points, lacking temporal continuity. Some researchers have demonstrated that combining remote sensing data with crop growth models can improve the accuracy of biomass estimation; however, the evolutionary relationship between rice aboveground fresh biomass (FB) and dry biomass (DB) and growth stage (phenological processes) remains unverified.
[0003] To address the aforementioned issues, there is an urgent need for a continuous monitoring method for rice biomass based on a combination of UAV low-altitude remote sensing images and phenological data, which can solve the problems associated with traditional methods. Summary of the Invention
[0004] To overcome the shortcomings of existing technologies, the purpose of this invention is to provide a method for continuous monitoring of rice biomass based on the combination of UAV low-altitude remote sensing images and phenology.
[0005] To achieve the above objectives, the present invention provides the following solution: This invention provides a method for continuous monitoring of rice biomass based on a combination of low-altitude remote sensing images from unmanned aerial vehicles (UAVs) and phenological data, comprising: Step 1: Acquire multi-band spectral images and visible light images using UAVs; Step 2: Perform spectral preprocessing and vegetation index calculation on the acquired multi-band spectral images and visible light images; Step 3: Collect ground data; Step 4: Collect phenological data; Step 5: Based on the calculated vegetation index and the collected ground data and phenological data, construct a continuous monitoring model for rice biomass and conduct continuous monitoring of rice biomass.
[0006] Preferably, in step 1, the acquisition of multi-band spectral images and visible light images by the UAV specifically includes: Equip drones with multispectral and visible light imaging sensors; Set a fixed flight speed, fixed flight altitude, and fixed overlap for the drone; Multi-band spectral and visible light images of crops in the test area were collected using drones.
[0007] Preferably, the drone acquires multi-band spectral images and visible light images once at each of the crop's growth stages.
[0008] Preferably, in step 2, the acquired multi-band spectral images and visible light images undergo spectral preprocessing and vegetation index calculation, specifically as follows: Preprocess the acquired multi-band spectral images; Preprocess the acquired visible light images; Based on the vegetation index calculation formula, the vegetation index data of crops at different growth stages in the experimental area were calculated using preprocessed multi-band spectral images and visible light images.
[0009] Preferably, the acquired multi-band spectral images are preprocessed, specifically as follows: Acquire multi-band spectral images; The multi-band spectral images are sequentially processed by image alignment, grid creation, and image stitching. The random forest algorithm is used to distinguish rice paddies from the background in the processed image. Based on MATLAB, regions of interest were selected from the multi-band spectral images after differentiation and processing to obtain the average spectral reflectance of the canopy in different cells within the experimental area.
[0010] Preferably, the acquired visible light image is preprocessed, specifically as follows: Acquire visible light images; The images were stitched together to obtain orthophoto stitched images of the entire community at different reproductive stages; Background removal is performed on orthophoto stitched images using the RF algorithm; The DN value and average spectral reflectance of the canopy are obtained by dividing the region of interest in the orthophoto mosaic image after background removal.
[0011] Preferably, in step 3, collecting ground data specifically involves: The ground data includes fresh biomass and dry biomass. Multiple crops are randomly selected from the center of each plot in the experimental area, and their fresh biomass and dry biomass are calculated. The average value is then taken to obtain the ground data for that plot.
[0012] Preferably, in step 4, collecting phenological data specifically includes: The phenological data includes the phenological index, accumulated temperature for growth, and the number of days after sowing. Obtain the daily average temperature and calculate the phenological index of accumulated temperature for growth; The number of days after sowing is obtained by manually recording the time after sowing.
[0013] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects: This invention provides a method for continuous monitoring of rice biomass based on a combination of UAV low-altitude remote sensing images and phenological data. The method includes acquiring multi-band spectral and visible light images using a UAV; performing spectral preprocessing and vegetation index calculation on the acquired multi-band spectral and visible light images; collecting ground data and phenological data; and constructing a continuous monitoring model for rice biomass based on the calculated vegetation indices and the collected ground and phenological data. This invention transforms the discrete values of rice growth timelines obtained from UAV low-altitude remote sensing images into continuous values, thereby enabling continuous estimation of rice biomass. Two selected phenological indices, post-sowing time and accumulated temperature (DFS and GDD), represent the crop's growth scale. By observing changes in biomass and VI at different stages, the coefficients k and b between biomass and different vegetation indices are obtained using an ordinary least squares regression model. Combined with changes in phenological indices, the established rice full-growth-period biomass monitoring model possesses a continuous dynamic curve. Compared to the scattered rice biomass data collected by traditional high-density UAV flights, this method combines phenology with the collection of low-altitude remote sensing images from UAVs, enabling temporal continuity, reducing labor costs, and providing a reference for continuous monitoring of rice biomass. A continuous monitoring model for fresh and dry biomass with multiple VIs and phenological indicators throughout the entire rice growth period was established. Based on R2p and RMSEp, the optimal parameters for monitoring rice biomass at different phenological indices, VIs, and growth stages were selected. The research results of this invention will contribute to the development of methods for continuous monitoring of rice biomass and can be applied to smart agriculture and environmental management. Attached Figure Description
[0014] Figure 1 This is a schematic diagram of the method flow of the present invention. Detailed Implementation
[0015] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.
[0016] The purpose of this invention is to provide a continuous monitoring method for rice biomass based on the combination of UAV low-altitude remote sensing images and phenological data. By fusing UAV low-altitude remote sensing images and phenological data, continuous and dynamic monitoring of rice biomass throughout its entire growth period is achieved, overcoming the limitations of traditional methods that only estimate at specific time points. This method utilizes phenological indicators such as accumulated temperature (GDD) and days after sowing (DFS) to transform discrete remote sensing observations into continuous biomass change curves, significantly improving the temporal completeness and accuracy of the estimation. Simultaneously, by establishing an optimized model with multiple vegetation indices and multiple phenological indicators, this method reduces dependence on external environmental conditions and enhances the model's universality and robustness. Compared with traditional destructive sampling methods, this invention significantly reduces labor and time costs, providing reliable technical support for accurate monitoring of rice growth, yield prediction, and smart agricultural management.
[0017] like Figure 1 As shown, this invention provides a method for continuous monitoring of rice biomass based on a combination of UAV low-altitude remote sensing images and phenological data, comprising: Step 1: Acquire multi-band spectral images and visible light images using UAVs; Step 2: Perform spectral preprocessing and vegetation index calculation on the acquired multi-band spectral images and visible light images; Step 3: Collect ground data; Step 4: Collect phenological data; Step 5: Based on the calculated vegetation index and the collected ground data and phenological data, construct a continuous monitoring model for rice biomass and conduct continuous monitoring of rice biomass.
[0018] In step 1, multi-band spectral images and visible light images are acquired using a drone, specifically as follows: Equip drones with multispectral and visible light imaging sensors; Set a fixed flight speed (e.g., 2.5 m / s), a fixed flight altitude (e.g., 20 m), and a fixed overlap (e.g., 75% in the heading direction and 60% in the lateral direction). Based on multi-band spectral and visible light images of crops in the experimental area collected by drones; With clear skies and no wind, manually set the exposure time, aperture, ISO, and focal length, and keep them consistent throughout the flight; The drone acquires multi-band spectral images and visible light images once at each of the crop's growth stages.
[0019] In step 2, spectral preprocessing and vegetation index calculation are performed on the acquired multi-band spectral images and visible light images, specifically as follows: Preprocess the acquired multi-band spectral images; Preprocess the acquired visible light images; Based on the vegetation index calculation formula, the data of multiple vegetation indices (VI) at different growth stages of crops in the experimental area were calculated using preprocessed multi-band spectral images and visible light images.
[0020] The acquired multi-band spectral images are preprocessed as follows: Acquire multi-band spectral images; Multispectral images of different reproductive stages acquired by UAVs were stitched together using Agisoft Photoscan software to obtain multispectral orthophotos. The specific steps included image alignment, grid creation, and image stitching. The random forest algorithm is used to distinguish rice paddies from the background in the processed image. Based on MATLAB, regions of interest were selected from the multi-band spectral images after differentiation and processing to obtain the average spectral reflectance of the canopy in different cells within the experimental area.
[0021] The acquired visible light images are preprocessed as follows: Acquire visible light images; The images were stitched together to obtain orthophoto stitched images of the entire community at different reproductive stages; Background removal is performed on orthophoto stitched images using the RF algorithm; The DN value and average spectral reflectance of the canopy are obtained by dividing the region of interest in the orthophoto mosaic image after background removal.
[0022] Obtain data on multiple vegetation indices (VI) at different growth stages; Based on the average spectral reflectance of the canopy and the vegetation index calculation formula, various canopy vegetation indices for different growth stages can be calculated. For example, the normalized difference vegetation index (NDVI), ratio vegetation index (RVI), ratio shortwave index (RSI), normalized difference red edge (NDRE), and modified simple ratio (MSR) can be calculated from the average spectral reflectance of the canopy obtained from multi-band spectral images. The average spectral reflectance of the canopy obtained from visible light images can be used to calculate the Red-Green Ratio Index (RGRI), Excess Green (EXG), Visible-band Difference Vegetation Index (VDVI), Visible Atmospherically Resistant Index (VARI), and Normalized Green-Red Difference Index (NGRDI), among others.
[0023] Step 3 involves collecting ground data, specifically: Rice biomass is divided into two types: fresh biomass (FB) and dry biomass (DB). FB and DB data were obtained using a destructive sampling method. Five rice plants were randomly selected from the center of each plot in the experimental field, and the average value was used to represent the aboveground biomass level of that plot, thus eliminating the edge effect of crop growth. Fresh and dry biomass were measured using an electronic balance (±0.01 g). Biomass is calculated based on weight per unit area (kg / m²). 2 According to calculations, the ground data acquisition frequency and the UAV low-altitude remote sensing image acquisition frequency are the same.
[0024] In step 4, phenological data is collected, specifically as follows: The phenological data includes the phenological index, accumulated temperature for growth, and the number of days after sowing. Based on publicly available data from the China Meteorological Data Service Center (CMDC, http: / / data.cma.cn), the daily maximum and minimum temperatures for different years at each study location were statistically analyzed. The average of the maximum and minimum temperatures was used to calculate the daily average temperature. The first phenological index, Growing Degree Days (GDD), was then calculated using the following formula: GDD = (T) max +T min ) / 2-T base (1) In the formula, T max The highest temperature of the day, T min The lowest temperature of the day, T base The reference temperature is generally 8℃ for rice crops.
[0025] The number of days from sowing (DFS) is obtained by manually recording the time after sowing.
[0026] In step 5, based on the calculated vegetation indices for different growth stages and the collected ground and phenological data, a continuous monitoring model for rice biomass is constructed to conduct continuous monitoring of rice biomass. Specifically: Because rice accumulates biomass in different organs at different growth stages, resulting in different spectral responses, this study observes changes in rice biomass and VI (vital energy) at different growth stages. First, Pearson correlation analysis was performed on aboveground dry and fresh biomass with various VIs to screen VIs with high correlation to biomass (r>0.5) and establish a continuous monitoring model for rice biomass. Then, a least squares regression model was established between the screened VIs from multiple growth stages (n) and aboveground dry and fresh biomass, extracting the slope and intercept, denoted as k. n and b n For k in multiple reproductive stages n and b n Multiple regression equations, including linear, power, exponential, and logarithmic forms, were used to establish a fitted regression equation with phenology (x). The regression equation with the best fit was selected to record the optimal regression parameters. Finally, a multi-layered fusion model for continuous monitoring of rice biomass was constructed to monitor the aboveground dry and fresh biomass of rice throughout the entire growth period.
[0027] The basic theoretical framework of the model is shown below: (2) (3) (4) In the formula, k n and b nThe slope and intercept of the ordinary least squares regression model for vegetation indices representing different growth stages and aboveground dry and fresh biomass are respectively given. f(x) represents the fitted regression equation in various forms such as linear, power, exponential and logarithmic. We will select the best regression model as the optimal model, and x corresponds to DFS and GDD, respectively.
[0028] The present invention also includes the following two steps, wherein: Step 6: Evaluate the model The coefficient of determination (R²) is used. 2 The accuracy of the model is evaluated using the root mean square error (RMSE) and the coefficient of determination R. The calculation formulas are shown in the following equations. 2 It can provide a good evaluation of the good fit between predicted and reference values, therefore R 2 A higher R-value indicates a better fit. RMSE is used to evaluate the estimation error between the predicted and reference values; therefore, the closer the RMSE value is to 0, the smaller the model estimation error. 2 R and RMSE are respectively represented by R 2 c and RMSEc are used to denote the prediction set, while R is used to denote the prediction set. 2 p and RMSEp represent: (5) (6) Embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0029] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.
[0030] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.
[0031] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.
[0032] Contents not described in detail in this specification are prior art known to those skilled in the art. It is hereby indicated that the above description is intended to help those skilled in the art understand this invention, but does not limit the scope of protection of this invention. Any equivalent substitutions, modifications, improvements, or simplifications of the above descriptions that do not depart from the essential content of this invention fall within the scope of protection of this invention.
Claims
1. A method for continuous monitoring of rice biomass based on a combination of UAV low-altitude remote sensing images and phenological data, characterized in that, include: Step 1: Acquire multi-band spectral images and visible light images using UAVs; Step 2: Perform spectral preprocessing and vegetation index calculation on the acquired multi-band spectral images and visible light images; Step 3: Collect ground data; Step 4: Collect phenological data; Step 5: Based on the calculated vegetation index and the collected ground data and phenological data, construct a continuous monitoring model for rice biomass and conduct continuous monitoring of rice biomass.
2. The method according to claim 1, characterized in that, In step 1, multi-band spectral images and visible light images are acquired using a drone, specifically as follows: Equip drones with multispectral and visible light imaging sensors; Set a fixed flight speed, fixed flight altitude, and fixed overlap for the drone; Multi-band spectral and visible light images of crops in the test area were collected using drones.
3. The method according to claim 2, characterized in that, The drone acquires multi-band spectral images and visible light images once at each of the crop's growth stages.
4. The method according to claim 3, characterized in that, In step 2, spectral preprocessing and vegetation index calculation are performed on the acquired multi-band spectral images and visible light images, specifically as follows: Preprocess the acquired multi-band spectral images; Preprocess the acquired visible light images; Based on the vegetation index calculation formula, the vegetation index data of crops at different growth stages in the experimental area were calculated using preprocessed multi-band spectral images and visible light images.
5. The method according to claim 4, characterized in that, The acquired multi-band spectral images are preprocessed as follows: Acquire multi-band spectral images; The multi-band spectral images are sequentially processed by image alignment, grid creation, and image stitching. The random forest algorithm is used to distinguish rice paddies from the background in the processed image. Based on MATLAB, regions of interest were selected from the multi-band spectral images after differentiation and processing to obtain the average spectral reflectance of the canopy in different cells within the experimental area.
6. The method according to claim 5, characterized in that, The acquired visible light images are preprocessed as follows: Acquire visible light images; The images were stitched together to obtain orthophoto stitched images of the entire community at different reproductive stages; Background removal is performed on orthophoto stitched images using the RF algorithm; The DN value and average spectral reflectance of the canopy are obtained by dividing the region of interest in the orthophoto mosaic image after background removal.
7. The method according to claim 6, characterized in that, Step 3 involves collecting ground data, specifically: The ground data includes fresh biomass and dry biomass. Multiple crops are randomly selected from the center of each plot in the experimental area, and their fresh biomass and dry biomass are calculated. The average value is then taken to obtain the ground data for that plot.
8. The method according to claim 7, characterized in that, In step 4, phenological data is collected, specifically as follows: The phenological data includes the phenological index, accumulated temperature for growth, and the number of days after sowing. Obtain the daily average temperature and calculate the phenological index of accumulated temperature for growth; The number of days after sowing is obtained by manually recording the time after sowing.