Regeneration rice yield prediction method, device, equipment and medium

Abstract

The invention discloses a regenerated rice yield prediction method, device and equipment and a medium. The method comprises the following steps: acquiring a current first time period hyperspectral image and a current first time period color image of regenerated rice in a target area; according to the hyperspectral image in the current first time period, calculating the surface reflectance of the hyperspectral image in the current first time period to obtain a vegetation parameter corresponding to the hyperspectral image in the current first time period; determining morphological characteristics of the regenerated rice according to the color image in the current first time period; according to the morphological characteristics and a morphological characteristic library, comparing rice varieties in the morphological characteristics and the morphological characteristic library, and determining the variety type of the regenerated rice; and inputting the variety type and the vegetation parameters into a pre-trained regenerated rice yield prediction model to obtain the predicted yield of the regenerated rice in the current second time period. According to the embodiment of the invention, the efficiency and accuracy of regenerated rice yield prediction can be improved.

Description

Technical Field

[0001] This invention relates to the field of crop management technology, and in particular to a method, apparatus, equipment and medium for predicting the yield of regenerated rice. Background Technology

[0002] With the development of agricultural modernization, crop yields can be predicted. By predicting the yield of ratooning rice, management measures can be adjusted in a timely manner.

[0003] Currently, existing technologies typically rely on human observation and historical experience to predict the yield of regenerated rice.

[0004] However, predicting the yield of regenerated rice using the above methods has drawbacks such as high labor costs, low efficiency, and insufficient accuracy. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for predicting the yield of regenerated rice. The embodiments of this invention can improve the efficiency and accuracy of predicting the yield of regenerated rice.

[0006] In a first aspect, embodiments of the present invention provide a method for predicting the yield of regenerated rice, the method comprising:

[0007] Acquire the current first time period hyperspectral image and the current first time period color image of the regenerated rice in the target area; the current first time period hyperspectral image includes at least one of the following: greening stage image, tillering stage image, jointing stage image and grain filling stage image;

[0008] Based on the hyperspectral image of the current first time period, calculate the surface reflectance of the hyperspectral image of the current first time period, and obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period;

[0009] Based on the color images of the first time period, determine the morphological characteristics of the regenerated rice;

[0010] Based on morphological characteristics and a morphological characteristic database, the rice varieties in the database are compared to determine the variety type of ratooning rice.

[0011] The variety type and vegetation parameters are input into the pre-trained ratooning rice yield prediction model to obtain the predicted yield of ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

[0012] Secondly, embodiments of the present invention also provide a device for predicting the yield of regenerated rice, the device comprising:

[0013] The image acquisition module is used to acquire a hyperspectral image and a color image of the current first time period of regenerated rice in the target area; the hyperspectral image of the current first time period includes at least one of the following: image of the greening stage, image of the tillering stage, image of the jointing stage, and image of the grain-filling stage;

[0014] The vegetation parameter acquisition module is used to calculate the surface reflectance of the hyperspectral image of the current first time period based on the hyperspectral image of the current first time period, and obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period.

[0015] The morphological feature determination module is used to determine the morphological features of regenerated rice based on the color image of the current first time period.

[0016] The variety type determination module is used to compare rice varieties with morphological characteristics and morphological characteristic databases to determine the variety type of ratooning rice.

[0017] The yield prediction module is used to input variety type and vegetation parameters into the pre-trained ratooning rice yield prediction model to obtain the predicted yield of ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

[0018] Thirdly, embodiments of the present invention also provide a device for predicting the yield of regenerated rice, the device comprising:

[0019] At least one processor; and

[0020] A memory that is communicatively connected to at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to perform the method for predicting the yield of regenerated rice according to any embodiment of the present invention.

[0022] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions for causing a processor to execute the method for predicting the yield of regenerated rice according to any embodiment of the present invention.

[0023] The technical solution of this invention, by acquiring hyperspectral and color images of a target area over a specific time period, enables comprehensive observation of the growth status of ratooning rice, providing a data foundation for subsequent analysis of vegetation parameters and morphological characteristics. By collecting hyperspectral images at different key growth stages, continuous monitoring of the ratooning rice's growth process from early to late stages is achieved, facilitating the capture of its spectral variation characteristics. Calculating surface reflectance from hyperspectral images and further extracting vegetation parameters allows for the quantification of key indicators reflecting the rice's physiological state from a spectral perspective. Analyzing color images to determine rice morphological characteristics supplements morphological and structural features beyond spectral information, further enhancing the accuracy of variety identification and yield prediction. By inputting the identified variety type and vegetation parameters into a yield prediction model, the future yield of ratooning rice in a specific area can be predicted. This solution addresses the shortcomings of relying on manual experience for predicting ratooning rice yield, which suffers from high labor costs, low efficiency, and insufficient accuracy. It improves the efficiency and accuracy of ratooning rice yield prediction.

[0024] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 A flowchart illustrating a method for predicting the yield of regenerated rice provided in an embodiment of the present invention;

[0027] Figure 2 A flowchart illustrating a method for predicting the yield of regenerated rice provided in an embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of the structure of a regenerated rice yield prediction device provided in an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram of a device for predicting the yield of regenerated rice, provided in an embodiment of the present invention. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0032] The acquisition, storage, and application of hyperspectral images, color images, and morphological feature libraries involved in the technical solutions of this invention comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0033] Figure 1 This is a flowchart illustrating a method for predicting the yield of ratooned rice according to an embodiment of the present invention. This embodiment is applicable to predicting the yield of already harvested rice. The method can be executed by a ratooned rice yield prediction device, which can be implemented in hardware and / or software.

[0034] See Figure 1 The method for predicting the yield of regenerated rice shown includes:

[0035] S101. Obtain the current first time period hyperspectral image and the current first time period color image of the regenerated rice in the target area; the current first time period hyperspectral image includes at least one of the following: greening stage image, tillering stage image, jointing stage image and grain filling stage image.

[0036] The target area refers to the specific agricultural area where the yield prediction for ratooning rice needs to be performed. The target area can include either a field-level rice planting area or a farm-level rice planting area. The target area defines the spatial geographic scope of image acquisition and prediction results, giving the collected data and calculation results a clear spatial orientation, facilitating field-level or region-level ratooning rice prediction. For example, a rice planting field (200 meters long and 50 meters wide) can be used as the target area to predict the yield of ratooning rice in that field.

[0037] Among them, regenerated rice can refer to a second or third season of rice crop that is grown naturally from the original rice stubble without being replanted after the harvest season.

[0038] The "current first time period" can refer to the time period within the current planting and growth cycle of the ratooning rice. Within this current first time period, the ratooning rice is in the development stage. For example, if the maturity and harvest date of the ratooning rice is November 1st of a certain year, and the growth cycle of the ratooning rice is from June to October, then the first time period can refer to the time period from June to October.

[0039] Hyperspectral images refer to images where each pixel contains a spectral curve, which represents the spectral characteristics of that pixel across multiple bands. Hyperspectral images can provide the spectral response of leaves of crops such as regenerated rice in different bands, and can be used to calculate vegetation parameters such as chlorophyll content, leaf water content, and nitrogen content. Optionally, the band collection range of the hyperspectral image can be set to the range of 400 nm to 1700 nm. 400 nm to 700 nm belongs to visible light, which reflects the light absorption state of regenerated rice; 700 nm to 1300 nm belongs to near-infrared light, which reflects the light scattering state of regenerated rice; and 1300 nm to 1700 nm belongs to short-wave infrared, which reflects the water status of regenerated rice.

[0040] In this context, a color image refers to an image composed of three RGB (Red, Green, Blue) channels. Color images are captured using a regular visible light camera and are used to display the morphological appearance information of regenerated rice in a target area. This morphological appearance information includes color, shape, and structure. Color images provide intuitive visual information and can extract the morphological features of regenerated rice.

[0041] The images mentioned here refer to those taken during the greening-up stage of ratooned rice. The greening-up stage refers to the period after rice seedlings have been transplanted or emerged, during which they resume growth and begin to turn green after a period of recovery.

[0042] Among them, the tillering stage image refers to a hyperspectral image taken of ratooned rice during the tillering stage. The tillering stage is the stage when lateral branches begin to differentiate from the base of the main stem of rice.

[0043] Among them, the jointing stage image refers to a hyperspectral image taken when ratooning rice enters the jointing stage. The jointing stage is the stage in which the rice stem elongates rapidly, and the internodes begin to differentiate and grow taller.

[0044] Among them, the grain-filling period image refers to a hyperspectral image taken during the grain-filling stage of ratooning rice. The grain-filling period is the period after rice flowering and pollination, during which the grains gradually fill out and starch and nutrients accumulate.

[0045] Specifically, images of ratooning rice can be taken at the greening stage, tillering stage, jointing stage, and grain-filling stage to obtain images of ratooning rice at these stages.

[0046] S102. Based on the hyperspectral image of the current first time period, calculate the surface reflectance of the hyperspectral image of the current first time period to obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period.

[0047] Surface reflectivity refers to the ability of ratooned rice to reflect incident solar radiation at a specific wavelength. It is typically expressed as the ratio of reflected radiation to incident radiation. Surface reflectivity reflects the reflection characteristics of ratooned rice to electromagnetic waves of different wavelengths. It is unaffected by atmospheric interference and solar angle, and serves as fundamental data for extracting vegetation parameters from ratooned rice.

[0048] Vegetation parameters refer to a set of parameters that characterize the physiological and biochemical state, structural state, and growth characteristics of plants. Vegetation parameters may include chlorophyll content, leaf water content, and nitrogen content. Chlorophyll content reflects the photosynthetic capacity of ratooning rice; leaf water content reflects the leaf water content of ratooning rice; and nitrogen content reflects the nutrient sufficiency of ratooning rice.

[0049] S103. Based on the color image of the current first time period, determine the morphological characteristics of the regenerated rice.

[0050] Morphological characteristics refer to the observable physical features of the external structure and shape of ratooning rice. These characteristics can include plant height, leaf shape, leaf color, and panicle shape. Morphological characteristics can be used to distinguish different rice varieties because different varieties have significant differences in external morphology, which is an important basis for identifying rice variety types. For example, ratooning rice variety A has dark green, narrow, and long leaves and a compact panicle shape; ratooning rice variety B has wide, short leaves, light green leaf color, and a loose panicle shape.

[0051] S104. Based on morphological characteristics and the morphological characteristic database, compare the morphological characteristics and the rice varieties in the morphological characteristic database to determine the variety type of regenerated rice.

[0052] The morphological feature library refers to a pre-established set of standard morphological feature information corresponding to a group of rice varieties. The morphological feature library can be used to quickly determine the variety type of a target rice by comparing and matching its morphological features with those of the rice to be identified.

[0053] Among them, variety type can refer to the specific cultivar of ratooning rice. Different varieties of ratooning rice have different regeneration capabilities, nutrient requirements, growth cycles, and stress resistance. The variety type of ratooning rice has a significant impact on yield prediction. Therefore, the variety type of ratooning rice needs to be included as one of the input data.

[0054] S105. Input the variety type and vegetation parameters into the pre-trained ratooning rice yield prediction model to obtain the predicted yield of ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

[0055] The ratooning rice yield prediction model can be defined as a model that takes the variety type and vegetation parameters of ratooning rice as input and outputs the predicted yield of ratooning rice for the current second time period. In a specific example, the ratooning rice yield prediction model extracts features from the variety type and vegetation parameters of ratooning rice, encodes the extracted features, and obtains the predicted yield of ratooning rice for the current second time period.

[0056] The "current second time period" can refer to the period from the start of planting to the maturity of the rice. Within this period, the ratooning rice progresses from its developmental stage to maturity. For example, if the harvest date for ratooning rice is November 1st of a certain year, and the initial developmental stage begins on June 1st, then the second time period could refer to the period from June 1st to November 1st.

[0057] As can be seen, in this embodiment, by acquiring hyperspectral and color images of the target area over a specific time period, a comprehensive observation of the growth status of ratooning rice can be achieved, providing a data foundation for subsequent analysis of vegetation parameters and morphological characteristics. By collecting hyperspectral images at different key growth stages, continuous monitoring of the ratooning rice growth process from early to late stages can be achieved, which is beneficial for capturing the spectral variation characteristics of ratooning rice. By calculating the surface reflectance from hyperspectral images and further extracting vegetation parameters, key indicators reflecting the physiological state of rice can be quantified from a spectral dimension. By analyzing color images to determine the morphological characteristics of rice, morphological and structural features beyond spectral information can be supplemented, further enhancing the accuracy of variety identification and yield prediction. By inputting the identified variety type and vegetation parameters into the yield prediction model, the future yield of ratooning rice in a specific area can be predicted. This solves the problems of high labor costs, low efficiency, and insufficient accuracy in predicting the yield of ratooning rice based on human experience, and can improve the efficiency and accuracy of ratooning rice yield prediction.

[0058] In an optional embodiment, Figure 2 The flowchart of a method for predicting the yield of regenerated rice provided in an embodiment of the present invention refines the process of "calculating the surface reflectance of the hyperspectral image of the current first time period based on the hyperspectral image of the current first time period to obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period" into "performing geometric correction on the hyperspectral image of the current first time period to obtain a standard hyperspectral image; obtaining the shooting parameters of the hyperspectral image of the current first time period; the shooting parameters include at least one of the following: shooting time, solar altitude angle, and atmospheric type; performing atmospheric correction on the standard hyperspectral image based on the shooting parameters to obtain the surface reflectance; and obtaining the vegetation parameters corresponding to the hyperspectral image of the current first time period based on the surface reflectance." This refines the process for predicting the yield of regenerated rice.

[0059] It should be noted that for parts not described in detail in the embodiments of the present invention, please refer to the descriptions in other embodiments.

[0060] See also Figure 2 The method for predicting the yield of ratooned rice shown includes:

[0061] S201. Obtain the current first time period hyperspectral image and the current first time period color image of the regenerated rice in the target area; the current first time period hyperspectral image includes at least one of the following: greening stage image, tillering stage image, jointing stage image and grain filling stage image.

[0062] S202. Perform geometric correction on the hyperspectral image of the current first time period to obtain a standard hyperspectral image.

[0063] Geometric correction can refer to the operation of spatially calibrating geometric distortions present in a hyperspectral image; a standard hyperspectral image can be a hyperspectral image that has undergone geometric correction.

[0064] S203. Obtain the shooting parameters of the hyperspectral image of the current first time period; the shooting parameters shall include at least one of the following: shooting time, solar altitude angle and atmospheric type.

[0065] The imaging parameters refer to the environmental and equipment conditions related to the acquisition of hyperspectral images. These parameters can be used for atmospheric correction and image inversion calculations.

[0066] Among them, the shooting time can refer to the precise time point of hyperspectral image acquisition, which usually includes year, month, day, hour, and minute; the solar altitude angle can refer to the angle between the center of the sun and the observation plane, which is usually expressed in degrees; and the atmospheric type can refer to the standardized classification of the comprehensive climatic characteristics of the local atmosphere at the time of image capture, such as aerosol type, humidity, and air pressure.

[0067] S204. Based on the shooting parameters, perform atmospheric correction on the standard hyperspectral image to obtain the surface reflectance.

[0068] Atmospheric correction refers to the process of converting the radiant intensity values ​​received from the sensor into the true surface reflectance. The purpose of atmospheric correction is to eliminate atmospheric interference with the spectral signal, ensuring that the spectral features in the hyperspectral image accurately reflect the inherent properties of the crop, rather than differences caused by atmospheric conditions.

[0069] S205. Based on the surface reflectance, obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period.

[0070] S206. Based on the color image of the current first time period, determine the morphological characteristics of the regenerated rice.

[0071] S207. Based on morphological characteristics and the morphological characteristic database, compare the morphological characteristics and the rice varieties in the morphological characteristic database to determine the variety type of ratooning rice.

[0072] S208. Input the variety type and vegetation parameters into the pre-trained ratooning rice yield prediction model to obtain the predicted yield of ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

[0073] As can be seen, in this embodiment, by performing geometric correction on the hyperspectral image, accurate registration of the image's spatial location can be achieved, thereby ensuring that each pixel in the image corresponds to its actual position on the ground, improving the spatial accuracy of subsequent parameter analysis. By acquiring the shooting parameters of the hyperspectral image, a quantitative description of the external environmental factors during imaging can be achieved, providing the necessary input conditions for subsequent atmospheric correction. By using the shooting parameters for atmospheric correction, the interference of atmospheric scattering and absorption on spectral data can be eliminated, thereby obtaining more realistic and accurate surface reflectance information. By extracting vegetation parameters based on surface reflectance, an objective assessment of the physiological state of regenerated rice can be achieved, providing key input for crop growth analysis and yield prediction.

[0074] In some embodiments, vegetation parameters include at least one of the following: chlorophyll content, leaf water content, and nitrogen content.

[0075] Chlorophyll content refers to the chlorophyll content in the leaves of ratooned rice. Chlorophyll content directly determines the greenness of the leaves and is one of the key indicators for measuring a plant's photosynthetic capacity. Chlorophyll content can be used to assess the physiological health of crops and indirectly reflect the yield of ratooned rice.

[0076] Leaf water content refers to the proportion of water in the total mass of leaf tissue. Plants absorb water through their roots and transpire through their leaves. When leaf water content decreases, it indicates that the ratooning rice is under water stress, which may affect photosynthesis and normal growth, thereby impacting the yield of ratooning rice.

[0077] Nitrogen content refers to the concentration or proportion of nitrogen in regenerated rice. Nitrogen is one of the essential macronutrients for plants, and it is a crucial raw material for the synthesis of proteins, nucleic acids, and chlorophyll. Nitrogen content is highly correlated with chlorophyll content, thus affecting final yield.

[0078] As can be seen, in this embodiment, by introducing chlorophyll content as a vegetation parameter, the photosynthetic capacity of regenerated rice can be effectively characterized, thereby more accurately reflecting its growth vigor and yield potential; by introducing leaf water content as a vegetation parameter, the water status of regenerated rice in a specific period can be effectively assessed, thereby improving the environmental adaptability of yield prediction; by introducing nitrogen content as a vegetation parameter, the nutrient supply level of regenerated rice can be reflected, thereby more scientifically predicting its final yield formation capacity.

[0079] In some embodiments, determining the morphological characteristics of regenerated rice based on the current first time period color image includes:

[0080] Edge detection is performed on the color image of the first time period to obtain the rice paddy area;

[0081] Based on the rice-growing region, determine the leaf and stem characteristics of the ratooned rice.

[0082] Based on leaf and stem characteristics, the morphological characteristics of regenerated rice were determined.

[0083] Edge detection refers to image processing techniques that identify the location of the boundaries of regenerated rice in a color image. The boundaries of regenerated rice are typically regions in a color image where grayscale values ​​or colors change significantly. Edge detection of regenerated rice can be achieved by calculating the gradient changes of pixels in the color image and identifying boundary lines with obvious grayscale or color changes.

[0084] In this context, the rice region refers to the image area in a color image that contains regenerated rice plants, as determined by edge detection. Identifying the rice region helps to focus the analysis target, eliminate background interference, and improve the accuracy of leaf and stem feature extraction.

[0085] Leaf characteristics refer to the morphological and structural features of the leaves of ratooned rice plants. Leaf characteristics can include leaf length, leaf width, leaf shape, and leaf number. Leaf characteristics reflect the crop's growth status, photosynthetic capacity, and health condition, thus helping to identify ratooned rice varieties.

[0086] Stem characteristics refer to the morphological features of the stem structure of ratooning rice. Stem characteristics can include stem length and stem radius. Stem characteristics reflect the tillering ability and lodging resistance of ratooning rice, thus helping to identify the variety type of ratooning rice.

[0087] As can be seen, in this embodiment, by performing edge detection on the color image, the region containing rice plants can be separated from the background, thus providing a precise target range for subsequent feature extraction; by extracting leaf and stem features based on the rice region, targeted analysis of plant structure information can be achieved, which helps to identify the variety type of regenerated rice.

[0088] In some embodiments, leaf characteristics include at least one of the following: leaf length, leaf width, leaf shape, and number of leaves; stem characteristics include at least one of the following: stem length and stem radius.

[0089] As can be seen, in this embodiment, by refining leaf and stem features into multiple specific parameters, the morphological differences of rice plants can be extracted more fully, thereby helping to improve the accuracy of variety identification of regenerated rice.

[0090] In some embodiments, based on morphological characteristics and a morphological characteristic library, the morphological characteristics and rice varieties in the morphological characteristic library are compared to determine the variety type of ratooning rice, including:

[0091] Based on morphological features, construct morphological feature vectors;

[0092] The similarity comparison results are obtained by comparing the morphological feature vectors with the standard feature vectors corresponding to each rice variety in the morphological feature library.

[0093] Based on the similarity comparison results, the variety type of regenerated rice was determined.

[0094] In this context, morphological feature vectors refer to ordered numerical values ​​representing the feature values ​​of regenerated rice across multiple morphological dimensions. These vectors transform the multiple morphological features of regenerated rice into mathematically meaningful vector forms, facilitating computer recognition and processing. In one example, each morphological feature, after being numericalized, is arranged in a predetermined order to form a multidimensional morphological feature vector. For instance, considering six morphological dimensions, the features of a regenerated rice plant can be represented as a 6-dimensional vector [30, 1.2, 2, 18, 60, 0.5], where each value represents a different feature attribute.

[0095] In this context, a standard feature vector can refer to an ordered set of numerical values ​​in a morphological feature library used to represent the standard features of a known regenerated rice variety across multiple morphological dimensions. Each standard feature vector represents the standard morphological features of a variety.

[0096] The similarity comparison result refers to the comparison of the similarity between the morphological feature vector of the rice to be tested and each standard feature vector. The similarity between the morphological feature vector of the rice and each standard feature vector can be measured by the distance or angle between the vectors.

[0097] As can be seen, in this embodiment, by constructing morphological features into morphological feature vectors, a structured expression of multidimensional morphological information of regenerated rice plants can be achieved, which facilitates subsequent unified data processing and similarity analysis. By comparing the similarity between the morphological feature vectors and standard feature vectors, the differences between regenerated rice samples and known varieties at the morphological level can be assessed, thereby providing a quantitative basis for variety identification. By determining the variety type based on the similarity comparison results, the variety of regenerated rice can be accurately identified.

[0098] In some embodiments, the rice yield prediction model can be trained in the following ways:

[0099] Obtain the historical true yield of regenerated rice in the first time period;

[0100] Acquire historical hyperspectral images and historical color images of regrowth rice in the target area for the second historical time period; the historical first time period includes the historical second time period, and the historical second time period is within the historical first time period; the historical second time period hyperspectral images include at least one of the following: historical images of the greening stage, historical images of the tillering stage, historical images of the jointing stage, and historical images of the grain-filling stage;

[0101] Based on the historical hyperspectral images of the second historical time period, the historical surface reflectance of the historical hyperspectral images of the second historical time period is calculated to obtain the historical vegetation parameters corresponding to the historical hyperspectral images.

[0102] Based on the color images of the second historical time period, the historical morphological characteristics of regenerated rice were determined;

[0103] Based on historical morphological characteristics and morphological characteristic databases, the rice varieties in the historical morphological characteristics and morphological characteristic databases are compared to determine the historical variety types of ratooning rice.

[0104] By inputting historical variety types and historical vegetation parameters into the ratooning rice yield prediction model, the historical predicted yield of ratooning rice is obtained.

[0105] The parameters of the rice prediction model are adjusted based on the difference between historical predicted yields and actual yields.

[0106] The actual yield refers to the yield per unit area measured through actual agricultural operations such as harvesting, threshing, and weighing after the entire growth cycle of ratooning rice. The actual yield is used to compare with historical predicted yields, and the parameters of the ratooning rice prediction model are adjusted according to the differences to obtain a pre-trained ratooning rice prediction model.

[0107] As can be seen, in this embodiment, by obtaining the actual yield of regenerated rice over a historical period, a reliable target value can be provided for training the prediction model, thereby achieving accurate supervised training of the yield prediction model; by using the error between the prediction result and the actual yield to adjust the model parameters, the model can be continuously optimized, thereby improving the prediction accuracy and robustness of the model in practical applications.

[0108] Figure 3 A schematic diagram of a device for predicting the yield of regenerated rice is provided in an embodiment of the invention. This embodiment of the invention is applicable to the prediction of yield for harvested rice. The device can execute a method for predicting the yield of regenerated rice and can be implemented in hardware and / or software.

[0109] See Figure 3The illustrated device for predicting the yield of regenerated rice includes: an image acquisition module 301, a vegetation parameter acquisition module 302, a morphological feature determination module 303, a variety type determination module 304, and a yield prediction module 305.

[0110] Image acquisition module 301 is used to acquire a hyperspectral image and a color image of the current first time period of regenerated rice in the target area; the hyperspectral image of the current first time period includes at least one of the following: image of the greening stage, image of the tillering stage, image of the jointing stage, and image of the grain-filling stage;

[0111] The vegetation parameter acquisition module 302 is used to calculate the surface reflectance of the hyperspectral image of the current first time period based on the hyperspectral image of the current first time period, and obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period.

[0112] The morphological feature determination module 303 is used to determine the morphological features of the regenerated rice based on the color image of the current first time period.

[0113] The variety type determination module 304 is used to compare the morphological characteristics and the morphological characteristic library with the rice varieties in the morphological characteristics library to determine the variety type of the ratooning rice.

[0114] The yield prediction module 305 is used to input the variety type and vegetation parameters into the pre-trained ratooning rice yield prediction model to obtain the predicted yield of ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

[0115] The technical solution of this invention, by acquiring hyperspectral and color images of a target area over a specific time period, enables comprehensive observation of the growth status of ratooning rice, providing a data foundation for subsequent analysis of vegetation parameters and morphological characteristics. By collecting hyperspectral images at different key growth stages, continuous monitoring of the ratooning rice's growth process from early to late stages is achieved, facilitating the capture of its spectral variation characteristics. Calculating surface reflectance from hyperspectral images and further extracting vegetation parameters allows for the quantification of key indicators reflecting the rice's physiological state from a spectral perspective. Analyzing color images to determine rice morphological characteristics supplements morphological and structural features beyond spectral information, further enhancing the accuracy of variety identification and yield prediction. By inputting the identified variety type and vegetation parameters into a yield prediction model, the future yield of ratooning rice in a specific area can be predicted. This solution addresses the shortcomings of relying on manual experience for predicting ratooning rice yield, which suffers from high labor costs, low efficiency, and insufficient accuracy. It improves the efficiency and accuracy of ratooning rice yield prediction.

[0116] In some embodiments, in calculating the surface reflectance of the hyperspectral image of the current first time period based on the hyperspectral image of the current first time period, and obtaining the vegetation parameters corresponding to the hyperspectral image of the current first time period, the vegetation parameter acquisition module 302 is specifically used for:

[0117] Geometric correction is performed on the hyperspectral image of the current first time period to obtain a standard hyperspectral image;

[0118] Obtain the capture parameters of the hyperspectral image for the current first time period; the capture parameters should include at least one of the following: capture time, solar altitude angle, and atmospheric type;

[0119] Based on the shooting parameters, atmospheric correction is performed on the standard hyperspectral image to obtain the surface reflectance;

[0120] Based on the surface reflectance, the vegetation parameters corresponding to the hyperspectral image of the first time period are obtained.

[0121] In some embodiments, vegetation parameters include at least one of the following: chlorophyll content, leaf water content, and nitrogen content.

[0122] In some embodiments, in determining the morphological characteristics of regenerated rice based on the color image of the current first time period, the morphological characteristic determination module 303 is specifically used for:

[0123] Edge detection is performed on the color image of the first time period to obtain the rice paddy area;

[0124] Based on the rice-growing region, determine the leaf and stem characteristics of the ratooned rice.

[0125] Based on leaf and stem characteristics, the morphological characteristics of regenerated rice were determined.

[0126] In some embodiments, leaf characteristics include at least one of the following: leaf length, leaf width, leaf shape, and number of leaves; stem characteristics include at least one of the following: stem length and stem radius.

[0127] In some embodiments, in determining the variety type of ratooning rice by comparing morphological features and rice varieties in the morphological feature library based on morphological features and the morphological feature library, the variety type determination module 304 is specifically used for:

[0128] Based on morphological features, construct morphological feature vectors;

[0129] The similarity comparison results are obtained by comparing the morphological feature vectors with the standard feature vectors corresponding to each rice variety in the morphological feature library.

[0130] Based on the similarity comparison results, the variety type of regenerated rice was determined.

[0131] In some embodiments, the yield prediction module 305 is specifically used for training the regenerated rice yield prediction model in the following ways:

[0132] Obtain the historical true yield of regenerated rice in the first time period;

[0133] Acquire historical hyperspectral images and historical color images of regrowth rice in the target area for the second historical time period; the historical first time period includes the historical second time period, and the historical second time period is within the historical first time period; the historical second time period hyperspectral images include at least one of the following: historical images of the greening stage, historical images of the tillering stage, historical images of the jointing stage, and historical images of the grain-filling stage;

[0134] Based on the historical hyperspectral images of the second historical time period, the historical surface reflectance of the historical hyperspectral images of the second historical time period is calculated to obtain the historical vegetation parameters corresponding to the historical hyperspectral images.

[0135] Based on the color images of the second historical time period, the historical morphological characteristics of regenerated rice were determined;

[0136] Based on historical morphological characteristics and morphological characteristic databases, the rice varieties in the historical morphological characteristics and morphological characteristic databases are compared to determine the historical variety types of ratooning rice.

[0137] By inputting historical variety types and historical vegetation parameters into the ratooning rice yield prediction model, the historical predicted yield of ratooning rice is obtained.

[0138] The parameters of the rice prediction model are adjusted based on the difference between historical predicted yields and actual yields.

[0139] The regenerated rice yield prediction device provided in this embodiment of the invention can execute the regenerated rice yield prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the regenerated rice yield prediction method.

[0140] Figure 4 This is a schematic diagram of a device for predicting the yield of regenerated rice, provided in an embodiment of the present invention.

[0141] like Figure 4As shown, the rice yield prediction device 400 includes at least one processor 401 and a memory, such as a read-only memory (ROM) 402 and a random access memory (RAM) 403, communicatively connected to the at least one processor 401. The memory stores computer programs executable by the at least one processor. The processor 401 can perform various appropriate actions and processes based on the computer program stored in the ROM 402 or loaded from storage unit 408 into the RAM 403. The RAM 403 can also store various programs and data required for the operation of the rice yield prediction device 400. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 408 is also connected to the bus 404.

[0142] Multiple components in the regenerated rice yield prediction device 400 are connected to the I / O interface 405, including: an input unit 406, such as a keyboard, mouse, etc.; an output unit 407, such as various types of displays, speakers, etc.; a storage unit 408, such as a disk, optical disk, etc.; and a communication unit 409, such as a network card, modem, wireless transceiver, etc. The communication unit 409 allows the regenerated rice yield prediction device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0143] Processor 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 401 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 401 performs the various methods and processes described above, such as methods for predicting the yield of regenerated rice.

[0144] In some embodiments, the method for predicting the yield of regenerated rice can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 408. In some embodiments, part or all of the computer program can be loaded and / or installed on the regenerated rice yield prediction device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by processor 401, one or more steps of the regenerated rice yield prediction method described above can be performed. Alternatively, in other embodiments, processor 401 can be configured to perform the regenerated rice yield prediction method by any other suitable means (e.g., by means of firmware).

[0145] Various embodiments of the systems and techniques described above can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0146] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0147] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0148] To provide user interaction, the systems and techniques described herein can be implemented on the operating detection device, which includes: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the rice yield prediction device. Other types of devices can also be used to provide user interaction; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0149] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0150] A computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system. It addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability.

[0151] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0152] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for predicting the yield of regenerated rice, characterized in that, The method includes: Acquire a hyperspectral image and a color image of regenerated rice in the target area for the current first time period; the hyperspectral image for the current first time period includes at least one of the following: images of the greening stage, tillering stage, jointing stage, and grain-filling stage; Based on the hyperspectral image of the current first time period, the surface reflectance of the hyperspectral image of the current first time period is calculated to obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period; Based on the color image of the current first time period, determine the morphological characteristics of the regenerated rice; Based on the morphological features and the morphological feature library, the rice varieties in the morphological features and the morphological feature library are compared to determine the variety type of the regenerated rice; The variety type and vegetation parameters are input into a pre-trained ratooning rice yield prediction model to obtain the predicted yield of the ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

2. The method according to claim 1, characterized in that, The step of calculating the surface reflectance of the hyperspectral image of the current first time period based on the hyperspectral image of the current first time period, and obtaining the vegetation parameters corresponding to the hyperspectral image of the current first time period, includes: Geometric correction is performed on the hyperspectral image of the current first time period to obtain a standard hyperspectral image; Acquire the shooting parameters of the hyperspectral image of the current first time period; the shooting parameters include at least one of the following: shooting time, solar altitude angle, and atmospheric type; Based on the shooting parameters, atmospheric correction is performed on the standard hyperspectral image to obtain the surface reflectance; Based on the surface reflectance, the vegetation parameters corresponding to the hyperspectral image of the current first time period are obtained.

3. The method according to claim 2, characterized in that, The vegetation parameters include at least one of the following: chlorophyll content, leaf water content, and nitrogen content.

4. The method according to claim 1, characterized in that, Determining the morphological characteristics of the regenerated rice based on the color image of the current first time period includes: Edge detection is performed on the color image of the current first time period to obtain the rice paddy area; Based on the rice region, determine the leaf and stem characteristics of the regenerated rice; The morphological characteristics of the regenerated rice are determined based on the leaf characteristics and the stem characteristics.

5. The method according to claim 4, characterized in that, The leaf characteristics include at least one of the following: leaf length, leaf width, leaf shape, and number of leaves; the stem characteristics include at least one of the following: stem length and stem radius.

6. The method according to claim 1, characterized in that, The step of comparing the morphological features and the morphological feature library with the rice varieties in the morphological features and the morphological feature library to determine the variety type of the ratooning rice includes: Based on the morphological features, construct a morphological feature vector; The similarity comparison results are obtained by comparing the morphological feature vector with the standard feature vector corresponding to each rice variety in the morphological feature library. Based on the similarity comparison results, the variety type of the regenerated rice is determined.

7. The method according to claim 1, characterized in that, The regenerated rice yield prediction model can be trained in the following ways: Obtain the historical true yield of the regenerated rice in the first time period; Acquire historical hyperspectral images and historical color images of regenerated rice in the target area for a second historical time period; the historical first time period includes the historical second time period, and the historical second time period is within the historical first time period; the historical second time period hyperspectral images include at least one of the following: historical images of the greening stage, historical images of the tillering stage, historical images of the jointing stage, and historical images of the grain-filling stage; Based on the hyperspectral image of the second historical time period, the historical surface reflectance of the hyperspectral image of the second historical time period is calculated to obtain the historical vegetation parameters corresponding to the historical hyperspectral image. Based on the color images of the second historical time period, the historical morphological characteristics of the regenerated rice were determined; Based on the historical morphological characteristics and the morphological characteristic database, the rice varieties in the historical morphological characteristics and the morphological characteristic database are compared to determine the historical variety type of the ratooning rice. The historical variety type and the historical vegetation parameters are input into the ratooning rice yield prediction model to obtain the historical predicted yield of the ratooning rice. The parameters of the regenerated rice prediction model are adjusted based on the difference between the historical predicted yield and the actual yield.

8. A device for predicting the yield of regenerated rice, characterized in that, include: The image acquisition module is used to acquire a hyperspectral image and a color image of the current first time period of regenerated rice in the target area; the hyperspectral image of the current first time period includes at least one of the following: image of the greening stage, image of the tillering stage, image of the jointing stage, and image of the grain-filling stage; The vegetation parameter acquisition module is used to calculate the surface reflectance of the hyperspectral image of the current first time period based on the hyperspectral image of the current first time period, and obtain the vegetation parameters corresponding to the hyperspectral image of the current first time period. The morphological feature determination module is used to determine the morphological features of the regenerated rice based on the color image of the current first time period. The variety type determination module is used to compare the morphological features and the morphological feature library with the rice varieties in the morphological features and the morphological feature library to determine the variety type of the ratooning rice. The yield prediction module is used to input the variety type and the vegetation parameters into a pre-trained ratooning rice yield prediction model to obtain the predicted yield of the ratooning rice in the current second time period; the current second time period includes the current first time period, and the current first time period is within the current second time period.

9. A device for predicting the yield of regenerated rice, characterized in that, The device for predicting the yield of regenerated rice includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the method for predicting the yield of regenerated rice as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that, when executed by a processor, implement the method for predicting the yield of regenerated rice as described in any one of claims 1-7.

Images