A hybrid wavelength weed control method based on target feature analysis

Abstract

The application discloses a mixed wavelength weeding control method based on target feature analysis, and relates to the technical field of photoelectric weeding control, which comprises the following steps: collecting target plant multi-source data of a region to be treated and generating a target feature vector; establishing a target plant tissue state model according to the target feature vector; generating a candidate mixed wavelength combination according to a preset wavelength library, inputting the target plant tissue state model, and obtaining an action process prediction result of each region; analyzing growth point inactivation requirements and crop safety constraints of a growth point region based on the action process prediction result, and obtaining inactivation judgment conditions; screening wavelength combinations, action target points and energy parameters according to the inactivation judgment conditions, and forming a wavelength control parameter group; irradiating according to the wavelength control parameter group, collecting temperature response, spectral response and morphological response, and generating a corrected wavelength control parameter group according to response deviation.

Description

Technical Field

[0001] This invention relates to the field of photoelectric weed control technology, specifically a hybrid wavelength weed control method based on target feature analysis. Background Technology

[0002] In recent years, photoelectric control technologies for precision weed control in farmland have gradually evolved from simple target recognition to an integrated approach encompassing "recognition-location-energy application." Multispectral imaging, depth sensing, and active light source illumination are being used to distinguish crops from weeds and to acquire plant phenotypes, reflectivity, and spatial location. Simultaneously, lasers, intense pulsed light, and multi-wavelength light sources are increasingly being used for non-contact weed control, with research focus shifting from fixed-power ablation to energy control based on plant tissue characteristics, water content, target site, and crop proximity.

[0003] Existing light-based weed control methods typically translate weed identification results directly into irradiation locations and fixed irradiation parameters, lacking a calculable intermediate model between target identification features and wavelength selection, as well as heat dose distribution. Different weed individuals exhibit variations in light absorption, scattering, water content, thermal conductivity, and vaporization consumption across their leaves, stems, stem bases, and growing points. Determining irradiation parameters solely based on target frames, segmented regions, or category labels can easily lead to problems such as leaf surface carbonization without growth point inactivation, energy consumption due to water phase transitions, and heat diffusion to neighboring crops. While some multi-wavelength weed control schemes introduce different wavelengths, they often focus on wavelength superposition or empirical parameter selection, failing to establish tissue state models based on the target plant's spectral characteristics, geometric structure, water content, and growing point exposure status. They also lack predictions for energy deposition, phase transition consumption, and heat diffusion processes. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by this invention is that existing mixed wavelength light energy weed control methods have the following problems: target identification features are disconnected from energy action parameters, lack prediction of energy deposition, phase transition consumption and heat diffusion in plant tissue regions, and are difficult to balance growth point inactivation and safety constraints of adjacent crops. It is also difficult to dynamically generate and correct mixed wavelength control parameters based on the characteristics of target plants.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a mixed-wavelength weed control method based on target feature analysis, comprising: collecting multi-source data of target plants in the area to be treated and generating target feature vectors; establishing a target plant tissue state model based on the target feature vectors; generating candidate mixed wavelength combinations according to a preset wavelength library and inputting them into the target plant tissue state model to obtain prediction results of the action process in each region; analyzing the growth point inactivation requirements and crop safety constraints in the growth point region based on the prediction results of the action process to obtain inactivation determination conditions; screening wavelength combinations, target points, and energy parameters according to the inactivation determination conditions to form a wavelength control parameter set; implementing irradiation according to the wavelength control parameter set, collecting temperature response, spectral response, and morphological response, and generating a corrected wavelength control parameter set based on the response deviation.

[0007] As a preferred embodiment of the hybrid wavelength weed control method based on target feature analysis described in this invention, the generation of the target feature vector includes: a data acquisition process using the same candidate plant target within the treatment area as the object, acquiring visible light images, selected band reflectance data, and depth data, and converting the visible light images, selected band reflectance data, and depth data to the same target coordinate system; separating candidate plant targets based on color differences, band reflectance differences, and height differences; determining candidate weed targets based on the positional relationship between the candidate plant targets and the crop protection area determined by the crop row centerline; extracting spectral features, geometric structure features, water content features, and growth point exposure features from the candidate weed targets, and associating them according to the same target number to generate a target feature vector.

[0008] As a preferred embodiment of the hybrid wavelength weed control method based on target feature analysis described in this invention, the target plant tissue state model includes: using a target feature vector as input, determining the leaf region, stem region, stem base region, and growth point region of the target plant based on the geometric structural features in the target feature vector; matching the equivalent absorption degree, equivalent scattering degree, and equivalent penetration depth of each tissue region of the target plant under corresponding candidate wavelengths based on the spectral features and water content features in the target feature vector; matching the equivalent specific heat, equivalent thermal conductivity, and phase change consumption parameters of each tissue region of the target plant based on the geometric structural features and water content features in the target feature vector; and setting the shading attenuation relationship based on the growth point exposure state characteristics to generate a target plant tissue state model containing tissue regions, optical parameters, thermal parameters, and growth point locations.

[0009] As a preferred embodiment of the hybrid wavelength weed control method based on target feature analysis described in this invention, the method for obtaining the predicted action process results for each region includes: pre-screening a preset wavelength library according to a target plant tissue state model to obtain an effective wavelength set; selecting two different wavelengths from the preset wavelength library and generating candidate hybrid wavelength combinations according to the input power range, irradiation duration range, pulse width range, and spot size range; inputting the candidate hybrid wavelength combinations into the target plant tissue state model, and obtaining the energy deposition distribution of each region based on the absorption degree, penetration depth, incident angle, and spot coverage relationship of different wavelengths in each tissue region; when the water content characteristics indicate that the corresponding region has entered the water vaporization stage, the corresponding part of the input energy is included in the phase change consumption; after deducting the phase change consumption, the temperature change process formed by the diffusion of remaining heat to the growth point region and adjacent crop regions is predicted according to the spatial relationship between tissue regions and thermal conductivity parameters, and the temperature change process is used as the heat diffusion result to generate the predicted action process results.

[0010] As a preferred embodiment of the hybrid wavelength weed control method based on target feature analysis described in this invention, the method for obtaining inactivation determination conditions includes: extracting temperature changes, effective heat dose, and sustained temperature rise time in the growing point region based on the prediction results of the action process; obtaining the basic inactivation temperature threshold, basic duration threshold, and basic heat dose threshold in the growing point region according to the preset tissue inactivation benchmark parameters corresponding to the target plant; correcting the basic duration threshold and basic heat dose threshold according to the water content, growing point exposure degree, and stem base size of the target plant to obtain the duration threshold and heat dose threshold, and using the basic inactivation temperature threshold as the inactivation temperature threshold; determining that the growing point inactivation requirement is met when the predicted heat dose in the growing point region reaches the heat dose threshold, the sustained temperature rise time reaches the duration threshold, and the highest predicted temperature reaches the inactivation temperature threshold; extracting the temperature change process and heat load change process of the adjacent crop region based on the prediction results of the action process; determining that the adjacent crop region meets the crop safety constraint when the temperature of the adjacent crop region does not exceed the crop safety temperature threshold and the heat load of the adjacent crop region does not exceed the crop safety heat load threshold; and generating inactivation determination conditions based on the determination results of the growing point inactivation requirement and the determination results of the crop safety constraint.

[0011] As a preferred embodiment of the hybrid wavelength weed control method based on target feature analysis described in this invention, the formation of the wavelength control parameter set includes: excluding candidate hybrid wavelength combinations that do not simultaneously meet the growth point inactivation requirements and crop safety constraints based on inactivation determination conditions; selecting wavelength combinations from the remaining candidate hybrid wavelength combinations that meet the heat dose requirements and whose total input energy meets preset energy limits; determining the target point based on the location of the growth point region; determining the input power, irradiation duration, pulse width, and number of irradiation repetitions based on the difference between the predicted heat dose and the heat dose threshold in the growth point region; and combining the wavelength combination, target point, input power, irradiation duration, pulse width, and number of irradiation repetitions to form the wavelength control parameter set.

[0012] As a preferred embodiment of the hybrid wavelength weed control method based on target feature analysis described in this invention, the step of generating a corrected wavelength control parameter set based on the response deviation includes: controlling the corresponding wavelength to irradiate the target plant according to the determined target point and energy parameters based on the wavelength control parameter set; collecting the temperature response of the growing point area and adjacent crop areas during irradiation; collecting the reflection change of the selected band and the morphological change of the irradiated area after irradiation; comparing the collected temperature response with the predicted result of the action process to obtain the response deviation; calculating the actual heat dose based on the temperature response of the growing point area, and obtaining the heat dose deviation based on the difference between the actual heat dose and the predicted heat dose of the growing point area; when the actual heat dose is lower than the predicted heat dose of the growing point area, correcting the energy parameters according to the adjustment priority of irradiation duration, number of irradiation repetitions, and input power; when the temperature response of the adjacent crop area exceeds the crop response temperature threshold, reducing the input power, shortening the irradiation duration, and reducing the number of irradiation repetitions; when the spectral response and morphological response of the adjacent crop area exceed the corresponding response anomaly threshold, stopping irradiation, and generating a corrected wavelength control parameter set based on the energy parameter adjustment result and the irradiation stop result.

[0013] As a preferred embodiment of the hybrid wavelength weed control system based on target feature analysis described in this invention, it includes: a feature acquisition module, a tissue modeling module, a process prediction module, an inactivation determination module, a wavelength control module, and a response correction module; the feature acquisition module is used to acquire multi-source data of target plants in the area to be treated and generate target feature vectors; the tissue modeling module is used to establish a target plant tissue state model based on the target feature vectors; the process prediction module is used to generate candidate hybrid wavelength combinations based on a preset wavelength library and input them into the target plant tissue state model to obtain the predicted process results for each area; the inactivation determination module is used to analyze the growth point inactivation requirements and crop safety constraints in the growth point area based on the predicted process results to obtain inactivation determination conditions; the wavelength control module is used to screen wavelength combinations, target points, and energy parameters according to the inactivation determination conditions to form a wavelength control parameter set; the response correction module is used to implement irradiation according to the wavelength control parameter set, acquire temperature response, spectral response, and morphological response, and generate a corrected wavelength control parameter set based on the response deviation.

[0014] A computer device includes a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement a hybrid wavelength weed control method based on target feature analysis.

[0015] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a hybrid wavelength weed control method based on target feature analysis.

[0016] The beneficial effects of this invention are as follows: The mixed-wavelength weed control method based on target feature analysis provided by this invention transforms the spectral characteristics, geometric structural characteristics, water content characteristics, and growth point exposure characteristics of the target plant into a tissue state model of the target plant. This model is then used to predict the energy deposition, phase transition consumption, and heat diffusion processes in each tissue region under the action of mixed wavelengths. The predicted heat dose, sustained temperature rise, and maximum predicted temperature of the growth point region, along with the heat load and safe temperature of adjacent crop regions, are used as judgment conditions to screen wavelength combinations, target points, and energy parameters. A corrected wavelength control parameter set is generated based on the actual temperature response, spectral response, and morphological response. This solves the problems in traditional weed control, such as the disconnect between identification results and energy control, the inability of leaf surface damage to represent growth point inactivation, the distortion of effective heat estimation due to phase transition consumption of water-containing tissues, and the difficulty in pre-constraining the heat risk of adjacent crops. It can predict the action process of different mixed wavelength combinations before irradiation, correct energy parameters based on response deviations during irradiation, reduce energy or stop irradiation when abnormal crop responses occur, improve the consistency of growth point inactivation, reduce ineffective energy input, and reduce the risk of accidental damage to adjacent crops. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the 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.

[0018] Figure 1 The present invention provides an overall flowchart of a hybrid wavelength weed control method based on target feature analysis.

[0019] Figure 2 A schematic diagram of a computer device provided by the present invention. Detailed Implementation

[0020] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. 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 protection scope of the present invention.

[0021] Reference Figure 1 As an embodiment of the present invention, a hybrid wavelength weed control method based on target feature analysis is provided, comprising:

[0022] S1: Collect multi-source data of target plants in the area to be processed and generate target feature vectors.

[0023] Furthermore, the generation of target feature vectors includes the following steps: During the data acquisition process, using the same candidate plant target within the processing area as the object, visible light images, selected band reflectance data, and depth data are collected, and these data are converted to the same target coordinate system; candidate plant targets are separated based on color differences, band reflectance differences, and height differences; candidate weed targets are determined based on the positional relationship between the candidate plant targets and the crop protection area determined by the crop row centerline; spectral features, geometric structure features, water content features, and growth point exposure features are extracted from the candidate weed targets, and they are associated according to the same target number to generate target feature vectors.

[0024] It should be noted that one specific scheme for generating the target feature vector includes, in this embodiment, the target plant number being denoted as... The wavelength number is denoted as Candidate mixed wavelength combinations are numbered as follows: The wavelength components in the candidate mixed wavelength combination are denoted as Plant tissue regions are denoted as The set of plant tissue regions is represented as:

[0025]

[0026] in, This represents the set of organizational regions participating in the modeling. Indicates the blade area. Indicates the stem region. Indicates the area at the base of the stem. Indicates the growth point region. Indicates the area adjacent to crops.

[0027] Using the same candidate plant target within the area to be processed as the object, visible light images, selected band reflectance data, and depth data were acquired separately. The coordinate system of the operating platform or the ground coordinate system of the target area was used as the unified target coordinate system, and the three types of data were transformed to the unified target coordinate system. The target pixel coordinates in the visible light image are denoted as follows: The target pixel coordinates in the selected band reflection data are marked as The target pixel coordinates in the depth data are as follows The corresponding transformation relationship is expressed as:

[0028]

[0029]

[0030]

[0031] in, Indicates the candidate plant target number. This indicates a visible light image data source. Indicates the selected band of reflection data source. Indicates a deep data source; , , They represent the numbers respectively. The coordinates of candidate plant targets in visible light images, selected band reflectance data, and depth data; , , These represent the transformation matrices from the visible light image data source, the selected band reflection data source, and the depth data source to the unified target coordinate system, respectively. , , These represent the target coordinates after the three types of data have been transformed into a unified target coordinate system.

[0032] After completing the coordinate transformation, the three types of transformation results are correlated at the target level. The three types of data are identified as corresponding to the same candidate plant target, and unified target coordinates are generated, represented as follows:

[0033]

[0034] in, Indicates the number is The coordinates of the candidate plant targets in the unified target coordinate system; , , These represent the coordinate fusion weights corresponding to visible light image data, selected band reflection data, and depth data, respectively.

[0035] To make candidate plant target separation feasible, this embodiment represents the plant candidate score for each pixel or spatial point as follows:

[0036]

[0037] in, Indicates the first Plant candidate scores for each pixel or spatial point Indicates the first The normalized color difference of each point relative to the ground background Indicates the first The difference in normalized band reflection of each point relative to the ground background Indicates the first Normalized height difference of each point relative to the ground background , , These represent the weights of color difference, band reflectance difference, and height difference, respectively. When plant candidate scores At that time, the first Each point is assigned to the candidate plant target region. The plant candidate score threshold is preferably set to 0.52.

[0038] After the candidate plant targets are separated, candidate weed targets are determined based on the positional relationship between the candidate plant targets and the crop protection area defined by the crop row centerline. Let the crop row centerline be... The shortest distance from the candidate plant target center to the crop row centerline is expressed as:

[0039]

[0040] in, Indicates the number is Candidate plant targets to crop row centerline The shortest distance, This indicates the calculation of the distance from a point to a line. Indicates the number is The positions of candidate plant targets within the same target coordinate system. The half-width of the crop protection area is denoted as... 5mm is preferred. At that time, the number will be The candidate plant targets were identified as candidate weed targets; when At that time, the number will be Candidate plant targets are included in crop protection areas and are not included in high-energy irradiation control processes.

[0041] For candidate weed targets, spectral features, geometric structure features, water content features, and growth point exposure features are extracted. Spectral features can be obtained from the normalized reflectance and band ratios of different bands. Normalized reflectance can be expressed as:

[0042]

[0043] in, Indicates the number is The candidate weed targets are numbered Normalized reflectance in the band, Indicates the number is The candidate weed targets are numbered The measured reflection intensity in the band, Indicates the number is The reference reflection intensity corresponding to the band, This indicates that to prevent small positive numbers with a denominator of zero, the preferred value is... Normalized reflectance is used to calculate spectral and water-state characteristics, reducing the impact of ambient light intensity and sensor gain variations on the judgment results. Water-state characteristics can be obtained by comparing the reflectance between the water-sensitive band and the reference band, expressed as:

[0044]

[0045] in, Indicates the number is The water content characteristics of candidate weed targets, Indicates the number is Candidate weed targets in the reference band Normalized reflectance under the following conditions Indicates the reference band number, Indicates the number is Candidate weed targets in the water-sensitive band Normalized reflectance under the following conditions The number indicating the moisture-sensitive band. This indicates that the calculation result will be limited to the range of 0 to 1. When When indicates a low water content state, The time indicates the water content state, when The time indicates a high water content state. The exposure status of the growth point of candidate weed targets is determined based on the candidate location at the base of the stem, the convergence point of the leaves, the visible area, and the shading ratio, and is represented as:

[0046]

[0047] in, Indicates the number is The growth point exposure status characteristics of candidate weed targets. Indicates the number is The visible area of ​​the candidate region for target growth points of candidate weeds. Indicates the number is The candidate weed targets are estimated based on the size of the stem base and the area of ​​the candidate growing point region. This indicates the occlusion ratio of the candidate region for the growth point. When... When the growth point region is defined as the area that can be directly irradiated; when When the growth point region is defined as the partially exposed region, when... At that time, the growth point region was defined as a low-exposure region.

[0048] The spectral feature vector of the candidate weed target consists of normalized reflectance in multiple selected bands, expressed as:

[0049]

[0050] in, Indicates the number is The spectral feature vectors of candidate weed targets, Indicates the selected band number used to acquire reflection data. Indicates the number of selected bands; Indicates the number is The candidate weed targets in the first Normalized reflectance at a selected band.

[0051] The geometric feature vector of the candidate weed target is calculated from the depth data and the target contour data, and is represented as:

[0052]

[0053] in, Indicates the number is The geometric structure feature vector of the candidate weed target; This indicates the target height obtained based on depth data; This represents the canopy width obtained based on the target contour; This indicates the location of the stem base determined based on the lowest visible point of the target and the connected area of ​​the stem; This indicates the candidate locations of growth points determined based on the position of the stem base and the convergence area of ​​the leaves; This indicates the stem base size obtained by normalizing the width or diameter of the stem base region; This represents the shortest distance from the candidate weed target to the centerline of the crop row. Spectral features, geometric structure features, water content features, and growth point exposure features are associated with the same target number to generate a target feature vector.

[0054]

[0055] in, Indicates the number is The target feature vector of the candidate weed target.

[0056] It should also be noted that by acquiring visible light images, selected band reflectance data, and depth data, and transforming multi-source data to the same target coordinate system, a target feature vector containing spectral features, geometric structure features, water content features, and growth point exposure features is generated. This transforms candidate weed targets from simple image recognition objects into numerical objects that can participate in tissue modeling and photothermal control, and provides a data foundation for distinguishing targets with different water content states, growth point exposure levels, and crop proximity relationships, thereby improving modeling reliability and reducing the risk of false illumination.

[0057] S2: Establish a target plant tissue state model based on the target feature vector.

[0058] Furthermore, the target plant tissue state model includes: using the target feature vector as input, determining the leaf region, stem region, stem base region, and growth point region of the target plant based on the geometric structural features in the target feature vector; matching the equivalent absorption, equivalent scattering, and equivalent penetration depth of each tissue region of the target plant at corresponding candidate wavelengths based on the spectral and water content features in the target feature vector; matching the equivalent specific heat, equivalent thermal conductivity, and phase change consumption parameters of each tissue region of the target plant based on the geometric structural and water content features in the target feature vector; and setting the shading attenuation relationship based on the growth point exposure state characteristics to generate a target plant tissue state model that includes tissue regions, optical parameters, thermal parameters, and growth point locations.

[0059] It should be noted that one specific approach to establishing a target plant tissue state model includes, based on the target feature vector... Geometric structural features Candidate weed targets are divided into tissue regions. Regions whose width is greater than their thickness and extend along the main direction of the leaf are marked as leaf regions within the segmented contours of the candidate weed targets. The region that connects the leaf convergence area and the lowest visible point of the target along the longitudinal direction is marked as the stem region. Mark the stem area within 3–15 mm above the lowest visible point of the target as the stem base area. ; the base of the stem area Between the leaf convergence area and satisfying the characteristics of the growth point exposure state. The local region of visibility calculation is marked as the growth point region. After the organizational regions are divided, for each organizational region... and each candidate wavelength The equivalent absorption degree is calculated and expressed as:

[0060]

[0061] in, Indicates the number is tissue regions in candidate weed targets In candidate wavelength number The equivalent absorption level below; Indicates the organization area In candidate wavelength number The basic absorption level can be obtained by pre-calibration; This represents the coefficient indicating the influence of normalized reflectance on the equivalent absorption level. Indicates the number is tissue regions in candidate weed targets In candidate wavelength number Normalized reflectance; This represents the coefficient indicating the influence of water content on the equivalent absorption rate. Indicates the number is tissue regions in candidate weed targets The regional water content characteristics. Based on each tissue region. and candidate wavelength Calculate the equivalent scattering degree and equivalent penetration depth. The equivalent scattering degree is expressed as:

[0062]

[0063] in, Indicates the number is tissue regions in candidate weed targets In candidate wavelength number The equivalent scattering degree under the following conditions; Indicates the organization area In candidate wavelength number The degree of basic scattering; This represents the coefficient indicating the influence of the incident angle on the degree of scattering; Indicates the number is The incident direction and tissue area of ​​the candidate weed targets The angle between surface normals; The coefficient representing the influence of surface undulations on the degree of scattering; Indicates the number is Candidate weed targets are tissue regions calculated based on depth data. The degree of surface undulation. The equivalent penetration depth is expressed as:

[0064]

[0065] in, Indicates the number is tissue regions in candidate weed targets In candidate wavelength number The equivalent penetration depth below; Indicates the organization area In candidate wavelength number The underlying penetration depth; This represents the adjustment coefficient of water content on penetration depth; Indicates the number is tissue regions in candidate weed targets Regional water content characteristics; Indicates the number is tissue regions in candidate weed targets In candidate wavelength number The equivalent absorption level below; and Representing organizational regions In candidate wavelength number The minimum and maximum allowable penetration depths are specified. Preferably, Take a thickness of 0.05–0.2 mm. Use a range of 0.5–3 mm. This range is used to limit the calculated penetration depth to the plant tissue scale. For each tissue region… Matching thermal parameters, expressed as:

[0066]

[0067]

[0068]

[0069] in, Indicates the number is tissue regions in candidate weed targets The equivalent specific heat parameter, Indicates the organization area Basic specific heat parameters This represents the coefficient indicating the influence of the water content state on the equivalent specific heat parameter. Indicates the number is tissue regions in candidate weed targets The equivalent volume; Indicates the number is tissue regions in candidate weed targets The equivalent thermal conductivity parameters, Indicates the organization area Basic thermal conductivity parameters Indicates the organization area The influence coefficient of geometric connectivity on the equivalent thermal conductivity parameter. This represents the geometric thermal conductivity factor obtained by normalizing the tissue region thickness, tissue region contact area, and stem diameter. Indicates the number is tissue regions in candidate weed targets Phase change consumption parameters, Indicates the organization area The basic parameters of phase change consumption per unit water volume.

[0070] For the growth point region Based on the characteristics of the growth point exposure state Set the occlusion attenuation coefficient, expressed as:

[0071]

[0072] in, Indicates the number is tissue regions in candidate weed targets The occlusion attenuation coefficient; This indicates that the current tissue region is a growth point region; This indicates that the current tissue region is not a growth point region. When... When the growth point region enters direct irradiation control; when At that time, the growth point area enters partial shading control; when At that time, the growth point area enters low exposure control.

[0073] After completing the tissue region division, optical parameter matching, thermal parameter matching, and growth point shading attenuation calculation, the target weeds were numbered according to the same candidate weed target number. By performing correlation, a target plant tissue state model is formed, which is represented as follows:

[0074]

[0075] in, Indicates the number is The target plant tissue state model for candidate weed targets. Indicates the number is The location of the growth point area of ​​the candidate weed target in the same target coordinate system.

[0076] It should also be noted that by establishing a target plant tissue state model, the entire weed target is transformed into a photothermal model with tissue-region differences. By dividing the plant into leaf, stem, stem base, and growth point regions, and matching the equivalent absorption, equivalent scattering, equivalent penetration depth, equivalent specific heat, equivalent thermal conductivity, and phase transition consumption parameters for each region at candidate wavelengths, the differences in absorption, temperature rise, and heat transfer of different tissues to mixed wavelengths are transformed into calculable parameters. A shading attenuation relationship is set according to the growth point exposure status to avoid misclassifying low-exposure growth points as fully irradiable objects, providing a model basis for energy deposition and heat diffusion prediction.

[0077] S3: Generate candidate mixed wavelength combinations based on the preset wavelength library, and input them into the target plant tissue state model to obtain the prediction results of the action process in each region.

[0078] Furthermore, obtaining the predicted effects of each region includes: pre-screening a preset wavelength library based on the target plant tissue state model to obtain an effective wavelength set; selecting two different wavelengths from the preset wavelength library and generating candidate mixed wavelength combinations according to the input power range, irradiation duration range, pulse width range, and spot size range; inputting the candidate mixed wavelength combinations into the target plant tissue state model, and obtaining the energy deposition distribution of each region based on the absorption degree, penetration depth, incident angle, and spot coverage relationship of different wavelengths in each tissue region; when the water content characteristics indicate that the corresponding region has entered the water vaporization stage, the corresponding part of the input energy is included in the phase change consumption; after deducting the phase change consumption, the temperature change process formed by the diffusion of the remaining heat to the growth point region and adjacent crop regions is predicted based on the spatial relationship between tissue regions and thermal conductivity parameters, and the temperature change process is used as the heat diffusion result to generate the predicted effects of the effects.

[0079] It should be noted that one specific scheme for obtaining prediction results of energy deposition, phase transition consumption, and heat diffusion processes in various regions includes, before generating candidate mixed wavelength combinations, first using a target plant tissue state model. For the preset wavelength library The candidate wavelengths are pre-screened. Specifically, for each candidate wavelength... The single-wavelength adaptation score is calculated and expressed as follows:

[0080]

[0081] in, Indicates the number is Candidate weed targets for candidate wavelengths Single-wavelength adaptation score; , , These represent the growing point region, stem base region, and adjacent crop region at candidate wavelengths, respectively. The equivalent absorption level below; , They represent the numbers respectively. Candidate weed targets were located in the growing point region and the stem base region at candidate wavelengths. The equivalent scattering degree under the following conditions; , , These represent the weights of depositional contribution from the growing point region, heat transfer contribution from the stem base region, and absorption risk from neighboring crops, respectively. This score prioritizes wavelengths with higher potential for action at the growing point or stem base, while reducing absorption risk from neighboring crops. The effective wavelength set is selected based on the single-wavelength fit score, as shown below:

[0082]

[0083] in, Indicates the number is The set of effective wavelengths corresponding to the target plants. This represents the single-wavelength adaptation scoring threshold. When the number of wavelengths in the effective wavelength set meets the candidate combination requirements, [the threshold is set]. Two different wavelengths are selected to generate candidate mixed wavelength combinations. When the number of wavelengths in the effective wavelength set is insufficient, candidate wavelengths are added according to the single wavelength adaptation score from high to low until the candidate combination requirements are met. Thus, while maintaining the effectiveness of mixed wavelengths, the number of invalid wavelength combinations entering the subsequent energy deposition, phase change consumption, and thermal diffusion prediction processes is reduced, thereby reducing the computational load and improving the feasibility of real-time control.

[0084] With a preset wavelength library and target plant tissue state model As a common input, from the target plant tissue state model Read the collection of tissue regions Equivalent absorption degree Equivalent scattering degree Equivalent penetration depth Equivalent specific heat parameters Equivalent thermal conductivity parameters Phase change consumption parameters shading attenuation coefficient and the location of the growth point region This serves as the calculation parameter for predicting the action process of candidate mixed wavelength combinations. Each selectable wavelength is bound to the corresponding input power range, illumination duration range, pulse width range, and spot size range, so that each generated candidate mixed wavelength combination has a definite wavelength value and an executable energy parameter range. The preset wavelength library is represented as follows:

[0085]

[0086] in, Indicates a preset wavelength library; This indicates the total number of candidate wavelengths in the preset wavelength library; Indicates the number is Candidate wavelengths; and Representing candidate wavelengths Minimum and maximum input power; and Representing candidate wavelengths The minimum and maximum irradiation durations; and Representing candidate wavelengths The minimum pulse width and the maximum pulse width; and Representing candidate wavelengths The minimum and maximum spot sizes. From the preset wavelength library. Two different wavelengths are selected and combined to generate a candidate mixed wavelength combination, represented as:

[0087]

[0088] in, Indicates the number is The candidate weed target corresponding to the first A number of candidate mixed wavelength combinations; Indicates the candidate mixed wavelength combination number; Indicates the wavelength component number in the candidate mixed wavelength combination. ; Indicates the first Among the candidate mixed wavelength combinations, the first... The wavelength number selected for each wavelength component; Indicates wavelength number The corresponding actual wavelength; , , , They represent the first Among the candidate mixed wavelength combinations, the first... Input power, illumination duration, pulse width, and spot size for each wavelength component; This indicates that the two wavelength components in the same candidate mixed wavelength combination use different wavelengths. (The last phrase, "candidate mixed wavelength combination," appears to be an error and is left untranslated.) Input target plant tissue state model Then, calculate the tissue region. In the The first candidate mixed wavelength combination The incident energy obtained at each wavelength component is expressed as:

[0089]

[0090] in, Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination Incident energy obtained at each wavelength component; Indicates the first The first candidate mixed wavelength combination Each wavelength component corresponds to the input power; Indicates the first The first candidate mixed wavelength combination The corresponding irradiation duration for each wavelength component; Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination The coverage ratio for each wavelength component ranges from 0 to 1, where 0 indicates no coverage and 1 indicates complete coverage. Indicates the number is The first candidate weed target The first candidate mixed wavelength combination Incident direction and tissue region under each wavelength component The angle between surface normals. The energy deposition amount is calculated based on the equivalent absorption, equivalent scattering, equivalent penetration depth, and shading attenuation coefficient, expressed as:

[0091]

[0092] in, Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination Energy deposition at each wavelength component; Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination Incident energy at each wavelength component; Indicates the number is tissue regions in candidate weed targets In wavelength number The equivalent absorption level below; Indicates the number is Candidate weed targets in tissue regions The equivalent thickness; Indicates the number is tissue regions in candidate weed targets In wavelength number The equivalent scattering degree under the following conditions; Indicates the number is tissue regions in candidate weed targets The equivalent thickness; Indicates the number is tissue regions in candidate weed targets In wavelength number The equivalent penetration depth below. To determine the tissue region. To determine whether the water vaporization stage has begun, this embodiment uses a dual threshold method based on predicted temperature and water content, expressed as follows:

[0093]

[0094] in, Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination Whether the water vaporization stage has been entered under each wavelength component is determined by a value of 1, which indicates that the water vaporization stage has been entered, and a value of 0 indicates that the water vaporization stage has not been entered. This indicates an indicator function that takes the value 1 when the condition is true and 0 when the condition is false. Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination Each wavelength component in the tissue region The predicted temperature; Indicates the organization area The vaporization judgment temperature threshold; Indicates the organization area The vaporization water content threshold. Preferably, Take 95 degrees Celsius. Take 0.65.

[0095] When the water content characteristics indicate that the corresponding region has entered the water vaporization stage, the target plant tissue state model... Reading tissue regions Phase change consumption parameters This is taken as the upper limit of energy consumption available for water vaporization or dehydration in this tissue region, and the energy consumption for phase change is expressed as:

[0096]

[0097] in, Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination Phase transition energy consumption at each wavelength component Indicates the number is Candidate weed targets in tissue regions The phase transition consumption parameter. Since areas with higher water content in plant tissues preferentially consume a portion of energy for water vaporization or dehydration, the remaining heat is obtained by subtracting the phase transition consumption energy from the energy deposition, expressed as:

[0098]

[0099] in, Indicates the number is tissue regions in candidate weed targets In the The first candidate mixed wavelength combination The remaining heat after deducting the phase transition loss at each wavelength component. The remaining heat... As input for thermal diffusion prediction, after obtaining the remaining heat, the thermal conductivity between different tissue regions is determined. To characterize the thermal conductivity limitation between adjacent tissues, the equivalent thermal conductivity parameters between tissue regions are first calculated, and then a discrete temperature update is performed. The equivalent thermal conductivity parameters are expressed as:

[0100]

[0101] in, Indicates the number is tissue regions in candidate weed targets With adjacent tissue regions The equivalent thermal conductivity parameters between them; Indicates and organizes the area Index of adjacent tissue regions with heat conduction relationships; Indicates the organization area The equivalent thermal conductivity parameters; Indicates the number is Adjacent tissue regions in candidate weed targets The equivalent thermal conductivity parameters.

[0102] When performing heat diffusion prediction, the target plant tissue state model is used. Reading tissue regions Equivalent specific heat parameters Equivalent thermal conductivity parameters and read adjacent tissue regions Equivalent thermal conductivity parameters Based on the target plant tissue state model The set of organizational regions in Based on the spatial location of each organization's regions, determine the set of adjacent regions. and organizational areas With adjacent tissue regions Equivalent distance between In determining the remaining heat and equivalent thermal conductivity Then, the temperature changes in each tissue region are updated using discrete time steps, as shown below:

[0103]

[0104] in, Indicates the number is tissue regions in candidate weed targets In the Predicted temperature at the next time step under the action of candidate mixed wavelength combinations; This represents the time variable in the thermal diffusion prediction process; Indicates the number is tissue regions in candidate weed targets In the The predicted temperature at the current time step under the action of candidate mixed wavelength combinations; This represents the time step for discrete computation, preferably between 0.01 and 0.10 s; Indicates and organizes the area A set of adjacent regions that have a thermal conductivity relationship; Indicates the number is tissue regions in candidate weed targets The equivalent specific heat parameter; Indicates the number is tissue regions in candidate weed targets With adjacent tissue regions The square of the equivalent distance between them; Indicates the number is tissue regions in candidate weed targets In the The predicted temperature at the current time step under the action of a combination of candidate mixed wavelengths.

[0105] After completing the energy deposition calculations, phase change consumption deductions, and thermal diffusion temperature updates, the different calculation results were assigned to the same candidate weed target number. Numbering of the same candidate mixed wavelength combination The results are summarized to form a unified prediction of the action process, which is expressed as follows:

[0106]

[0107] in, Indicates the number is The candidate weed targets in the first Prediction results of the action process under candidate mixed wavelength combinations; Indicates the growth point region The Chinese number is The candidate weed targets in the first Predicted temperature change process under candidate mixed wavelength combinations; Indicates adjacent crop areas The Chinese number is The candidate weed targets in the first Predicted temperature change process under candidate mixed wavelength combinations.

[0108] It should also be noted that candidate mixed wavelength combinations are generated by using a preset wavelength library and a target plant tissue state model, and the energy deposition, phase transition consumption, and heat diffusion processes in each tissue region are predicted, transforming the selectable light source parameters into evaluable results of the action process before irradiation. The energy deposition amount is calculated by combining the wavelength parameter range, tissue absorption degree, scattering degree, penetration depth, incident angle, and light spot coverage relationship, and the phase transition consumption during water vaporization is subtracted to obtain the remaining heat involved in heating and diffusion. Temperature changes in the growing point region and adjacent crop regions are obtained through heat diffusion prediction, thus providing a basis for inactivation determination and crop safety assessment.

[0109] S4: Based on the prediction results of the action process, analyze the growth point inactivation requirements and crop safety constraints in the growth point region to obtain the inactivation judgment conditions.

[0110] Furthermore, the inactivation determination criteria include: extracting temperature changes, effective heat dose, and duration of sustained temperature rise in the growth point region based on the predicted action process; obtaining the baseline inactivation temperature threshold, baseline duration threshold, and baseline heat dose threshold for the growth point region based on the preset tissue inactivation baseline parameters corresponding to the target plant; correcting the baseline duration threshold and baseline heat dose threshold based on the target plant's water content, growth point exposure level, and stem base size to obtain the duration threshold and heat dose threshold, and using the baseline inactivation temperature threshold as the inactivation temperature threshold; determining that the growth point inactivation requirement is met when the predicted heat dose in the growth point region reaches the heat dose threshold, the duration of sustained temperature rise reaches the duration threshold, and the highest predicted temperature reaches the inactivation temperature threshold; extracting the temperature change process and heat load change process of the adjacent crop region based on the predicted action process; determining that the adjacent crop region meets the crop safety constraint when the temperature of the adjacent crop region does not exceed the crop safety temperature threshold and the heat load of the adjacent crop region does not exceed the crop safety heat load threshold; and generating inactivation determination criteria based on the determination results of the growth point inactivation requirement and the crop safety constraint.

[0111] It should be noted that one specific approach to obtaining the inactivation criteria includes predicting the outcome from the action process. Extracting the growth point region Predicting temperature change process Based on the predicted temperature change process, the predicted heat dose in the growth point region is calculated and expressed as:

[0112]

[0113] in, Indicates the number is The target plant in the Predicted thermal dose in the growth point region under candidate mixed wavelength combinations; Indicates the first The total illumination duration of each candidate mixed wavelength combination is determined by the common illumination window of the mixed wavelengths; Represents the integral time element; This represents the reference temperature at which the growth point region participates in the cumulative heat dose. Preferably, Set the temperature to 45 degrees Celsius.

[0114] After obtaining the predicted heat dose in the growth point region, it is determined whether the predicted temperature rise is sufficiently sustained. Specifically, the temperature change process is predicted in the growth point region. In the statistical prediction, the temperature is not lower than the inactivation temperature threshold of the growth point region. The cumulative time is recorded and denoted as the duration of continuous temperature rise in the growth point region, expressed as:

[0115]

[0116] in, Indicates the number is The candidate weed targets in the first Duration of temperature rise in the growth point region under candidate mixed wavelength combinations; Indicates the number is The inactivation temperature threshold for the growing point region of candidate weed targets is preferably 60 degrees Celsius. The inactivation temperature threshold, duration threshold, and heat dose threshold for the growing point region are adjusted based on water content, growing point exposure level, and stem base size, and are expressed as follows:

[0117]

[0118]

[0119]

[0120] in, This indicates the basic inactivation temperature threshold in the growth point region; Indicates the number is The water content characteristics of the growing point region in the target plant; Indicates the number is The normalized value of the stem base size of the target plant; , , These represent the correction coefficients for the inactivation temperature threshold based on water content, degree of shading at the growth point, and stem base size, respectively. Indicates the number is The threshold for the duration of the target plant's growth point region; This represents the threshold for the basic duration of the growth point region; , These represent the correction coefficients for the duration threshold based on the water content and the degree of growth point shading, respectively. Indicates the number is The heat dose threshold of the target plant's growing point region; This indicates the baseline heat dose threshold for the growth point region; , , These represent correction coefficients for the heat dose threshold based on water content, degree of growth point shading, and stem base size, respectively. Preferably, Take 60 degrees Celsius. Take 0.8s.

[0121] Predicting heat load in the growth point region Duration of temperature rise After determining the highest predicted temperature and corresponding threshold, the determination of whether the growth point region meets the inactivation requirements is made, as shown below:

[0122]

[0123] in, Indicates the number is The candidate weed targets in the first Whether the growth point inactivation requirement is met under each candidate mixed wavelength combination is determined by a value of 1 indicating that the requirement is met and a value of 0 indicating that the requirement is not met. This indicates the highest predicted temperature in the growth point region during irradiation. Predicted results from the action process. Extracting neighboring crop regions Predicting temperature change process And calculate the predicted heat load of the adjacent crop area, expressed as:

[0124]

[0125] in, Indicates the number is The target plant corresponds to the adjacent crop area in the first Predicted heat load under candidate mixed wavelength combinations; This indicates the safe reference temperature for the adjacent crop area to participate in the cumulative heat load. Preferably, Take 40 degrees Celsius. The crop safety constraints are determined based on the highest predicted temperature and predicted heat load in neighboring crop areas, and are expressed as follows:

[0126]

[0127] in, Indicates the number is The candidate weed targets in the first Whether the crop safety constraints are met under the candidate mixed wavelength combinations is determined by a value of 1, which indicates that the constraints are met, and a value of 0 indicates that the constraints are not met. This indicates the highest predicted temperature in the adjacent crop area during the irradiation period; Indicates the safe temperature threshold for crops; This indicates the safe heat load threshold for nearby crops. Preferably, Set the temperature to 50 degrees Celsius. Based on the results of the growth point inactivation requirements and crop safety constraints, generate the inactivation criteria:

[0128]

[0129] in, Indicates the number is The candidate weed targets in the first Inactivation criteria under candidate mixed wavelength combinations. When, it indicates the first Several candidate mixed wavelength combinations simultaneously satisfy growth point inactivation requirements and crop safety constraints; when When, it indicates the first The candidate mixed wavelength combinations do not meet the requirements for growth point inactivation and crop safety constraints, and are therefore not considered as feasible candidate combinations.

[0130] It should also be noted that, based on the prediction results of the action process, the determination of whether the growing point region meets the constraints of heat dose, sustained temperature rise, and crop safety achieves a transformation from "predicting temperature and energy processes" to "whether candidate wavelength combinations can be executed." Specifically, by calculating the predicted heat dose in the growing point region, statistically analyzing the sustained temperature rise time to reach the inactivation temperature threshold, and correcting the inactivation temperature threshold, duration threshold, and heat dose threshold based on water content, growing point exposure level, and stem base size, the judgment of weed control effectiveness, which was originally easily misled by leaf surface carbonization or instantaneous temperature peaks, is transformed into a multi-condition judgment object focusing on growing point inactivation.

[0131] S5: Based on the inactivation determination criteria, select wavelength combinations, target points, and energy parameters to form a wavelength control parameter set.

[0132] Furthermore, forming the wavelength control parameter set includes: excluding candidate mixed wavelength combinations that do not simultaneously meet the growth point inactivation requirements and crop safety constraints based on the inactivation determination criteria; selecting wavelength combinations from the remaining candidate mixed wavelength combinations that meet the heat dose requirements and whose total input energy meets the preset energy limit; determining the target point based on the location of the growth point region; determining the input power, irradiation duration, pulse width, and number of irradiation repetitions based on the difference between the predicted heat dose and the heat dose threshold in the growth point region; and combining the wavelength combination, target point, input power, irradiation duration, pulse width, and number of irradiation repetitions to form the wavelength control parameter set.

[0133] It should be noted that one specific scheme for forming a wavelength control parameter set includes, based on the deactivation determination criteria and input energy limitations, selecting executable candidate combinations from candidate mixed wavelength combinations, and forming a wavelength control parameter set. The first... Input power of two wavelength components in a candidate mixed wavelength combination Corresponding irradiation duration Multiply and sum to obtain the total input energy. , Indicates the number is The candidate weed target corresponds to the first The total input energy of the candidate mixed wavelength combinations, ; Indicates the first Among the candidate mixed wavelength combinations, the first... Input power of each wavelength component; Indicates the first Among the candidate mixed wavelength combinations, the first... The irradiation duration for each wavelength component. Based on the inactivation determination criteria and the total input energy limit, a set of executable candidate combinations is formed, represented as:

[0134]

[0135] in, Indicates the number is The set of executable candidate combinations corresponding to the candidate weed targets; Indicates the number is The upper limit of the input energy for candidate weed targets. Upper limit of input energy. It can be set based on the distance between the candidate weed targets and the crop protection area. When When not empty, the target candidate mixed wavelength combination is selected from the set of executable candidate combinations, as follows:

[0136]

[0137] in, Indicates the selected target candidate mixed wavelength combination number; , , These represent the evaluation weights for the total input energy, the predicted crop heat load, and the predicted heat dose deviation at the growing point, respectively. In determining the target candidate mixed wavelength combination Then, from the target plant tissue state model Read the location of the growth point region The closest irradiable location in the stem base region to the growing point region. And combined with the characteristics of the growth point exposure state The target site is identified and represented as follows:

[0138]

[0139] in, Indicates the number is The target location of the candidate weeds; Indicates the number is The location of candidate weed targets in the growth point area; Indicates the number is The candidate weed target is the closest illuminated location in the stem base region to the growing point region. When the growing point is exposed, the characteristics of the state are considered. When the value is not lower than 0.60, the location of the growth point region will be determined. As a target site; when When the value is below 0.60, the irradiated location in the stem base region closest to the growing point region should be selected. As a target site.

[0140] The number of irradiation repetitions is determined based on the thermal dose threshold of the growth point region and the thermal dose that can be generated in a single execution of the target candidate mixed wavelength combination, as expressed as:

[0141]

[0142] in, Indicates the number is The number of irradiation repetitions for candidate weed targets; the number of irradiation repetitions represents the number of times the same candidate mixed wavelength combination is fully executed on the same target point; Indicates the maximum number of permissible irradiation repetitions, preferably ; Indicates the number is The first candidate weed target The predicted thermal dose of the growth point region when a single complete execution of a candidate mixed wavelength combination is performed.

[0143] Selected candidate mixed wavelength combinations Target Input power of each wavelength component in the target candidate mixed wavelength combination irradiation duration Pulse width Spot size and the number of irradiation repetitions Combine to form a wavelength control parameter set , represented as:

[0144]

[0145] in, Indicates the number is Wavelength control parameter set for candidate weed targets; Indicates the number is The first candidate weed target selected A number of candidate mixed wavelength combinations; Indicates the number is The target location of the candidate weeds; , , , They represent the first Among the candidate mixed wavelength combinations, the first... Input power, illumination duration, pulse width, and spot size for each wavelength component.

[0146] It should also be noted that by screening candidate mixed wavelength combinations based on inactivation criteria, and combining the total input energy, predicted crop heat load, and growth point heat dose deviation, target candidate mixed wavelength combinations are determined, forming an executable wavelength control parameter set. The target point is determined based on the growth point exposure status, and the number of irradiation repetitions is determined based on the predicted heat dose generated by a single irradiation. This unifies the wavelength combination, target point, input power, irradiation duration, pulse width, spot size, and number of repetitions into unified control parameters, thereby transforming the prediction results into executable weed control actions.

[0147] S6: Irradiate according to the wavelength control parameter set, collect temperature response, spectral response and morphological response, and generate a corrected wavelength control parameter set based on the response deviation.

[0148] Furthermore, the process of generating a corrected wavelength control parameter set based on the response deviation includes: controlling the corresponding wavelength to irradiate the target plant according to the determined target point and energy parameters based on the wavelength control parameter set; collecting the temperature response of the growing point region and adjacent crop regions during irradiation; collecting the reflection changes of the selected band and the morphological changes of the irradiated area after irradiation; comparing the collected temperature response with the predicted results of the action process to obtain the response deviation; calculating the actual heat dose based on the temperature response of the growing point region, and obtaining the heat dose deviation based on the difference between the actual heat dose and the predicted heat dose of the growing point region; when the actual heat dose is lower than the predicted heat dose of the growing point region, correcting the energy parameters according to the adjustment priority of irradiation duration, number of irradiation repetitions, and input power; when the temperature response of the adjacent crop region exceeds the crop response temperature threshold, reducing the input power, shortening the irradiation duration, and reducing the number of irradiation repetitions; when the spectral response and morphological response of the adjacent crop region exceed the corresponding response anomaly threshold, stopping irradiation, and generating a corrected wavelength control parameter set based on the energy parameter adjustment results and the irradiation cessation results.

[0149] It should be noted that one scheme for generating a corrected wavelength control parameter set based on the response deviation specifically includes collecting the temperature response of the growing point region and adjacent crop regions during irradiation. The actual heat dose is calculated based on the temperature response of the growing point region, expressed as:

[0150]

[0151] in, Indicates the number is The actual heat dose in the growth point area of ​​the candidate weed target; Indicates the number is The actual observed duration of the candidate weed targets during the irradiation process; Indicates the number is Candidate weed targets in the growing point area at any time The collected temperature response. The response deviation is calculated based on the predicted heat dose and the actual heat dose, and is expressed as:

[0152]

[0153] in, Indicates the number is The response bias of candidate weed targets; Indicates the number is Candidate mixed wavelength combination of candidate weeds Predicted heat dose for the corresponding growth point region; Indicates the number is The candidate weed targets are determined based on the actual heat dose of the growing point region obtained from the actual temperature response. When the actual heat dose is lower than the predicted heat dose, energy compensation is triggered; when When the deviation between the actual and predicted heat doses is within the allowable range, no increase is made to the energy parameters. A 5% deviation band is used to filter temperature acquisition noise and minor natural fluctuations. When the actual heat dose obtained from the temperature response of the growth point region is lower than the predicted heat dose for the growth point region in the prediction results of the action process, the energy parameters are adjusted according to the deviation between the actual and predicted heat doses, prioritizing adjustments based on irradiation duration, number of irradiation repetitions, and input power, as follows:

[0154]

[0155] in, Represents the target candidate mixed wavelength combination The Middle Irradiation duration corrected for each wavelength component; Indicates wavelength number The corresponding maximum irradiation duration, Represents the target candidate mixed wavelength combination The Middle Wavelength number of each wavelength component; Represents the target candidate mixed wavelength combination The Middle Irradiation duration for each wavelength component; This represents the irradiation duration correction factor, preferably 0.3 to 0.8. When and When adjusting the number of irradiation repetitions, it is expressed as:

[0156]

[0157] in, Indicates the number is The number of irradiation repetitions after correction of candidate weed targets; Indicates the maximum number of permissible irradiation repetitions; Indicates the number is Number of irradiation repetitions before correction of candidate weed targets; Indicates the number is Candidate mixed wavelength combination of candidate weeds The predicted heat dose for the growth point region during a single complete execution is determined by... Obtained under single-exposure conditions. When and Adjust the input power as needed:

[0158]

[0159] in, Represents the target candidate mixed wavelength combination The Middle Input power corrected for each wavelength component; Indicates wavelength number The corresponding maximum input power; Represents the target candidate mixed wavelength combination The Middle Input power before correction of each wavelength component; This represents the input power correction factor, preferably 0.1 to 0.4. While providing compensatory irradiation to the target plant, safety adjustments are made based on the temperature response of neighboring crop areas. Specifically, during irradiation, the temperature response of neighboring crop areas is considered. The maximum value is taken as the highest temperature response in the neighboring crop area. , Indicates the number is Candidate weed targets in adjacent crop areas at any time The collected temperature response, Indicates the number is The highest temperature response in the neighboring crop area corresponding to the candidate weed target. Exceeding the crop response temperature threshold At that time, crop safety adjustments are performed. Preferably, Pick This threshold is lower than or equal to the crop safe temperature threshold. This allows for the reduction of energy input in advance when actual crop temperatures approach a risk level. Crop safety regulation is represented as:

[0160]

[0161] in, This represents the crop safety adjustment coefficient, with the preferred value being... After irradiation, the changes in spectral and morphological responses in adjacent crop areas were calculated. The changes in spectral response are expressed as follows:

[0162]

[0163] in, Indicates the number is The spectral response changes of candidate weed targets corresponding to neighboring crop areas; The number after irradiation is indicated. The candidate weed targets correspond to the adjacent crop areas in the selected monitoring bands Average normalized reflectance; The number before irradiation is The candidate weed targets correspond to the adjacent crop areas in the selected monitoring bands Average normalized reflectance; This indicates the selected band number used to monitor crop reflectance anomalies. Morphological response changes are represented as:

[0164]

[0165] in, Indicates the number is The morphological response changes of candidate weed targets in adjacent crop areas; The number after irradiation is indicated. The visible area of ​​the candidate weed target corresponding to the adjacent crop area; The number before irradiation is The candidate weed targets correspond to the visible area of ​​the adjacent crop region. This is determined by analyzing the spectral response changes of the adjacent crop region. and morphological response changes After calculation, it is further determined whether abnormal reactions caused by irradiation have appeared in neighboring crop areas. After irradiation is completed, a stop irradiation indicator is generated based on changes in the spectral and morphological responses of neighboring crop areas, as shown below:

[0166]

[0167] in, Indicates the number is The stop illumination flag corresponding to the candidate weed target has a value of 1 to indicate that illumination has stopped, and a value of 0 to indicate that the stop illumination is not triggered. Indicates the threshold for abnormal spectral response in crop regions; This represents the threshold for abnormal morphological response in crop regions. Preferably, Take 0.14, Take 0.1. After obtaining the stop irradiation sign... Then, the corrected input power Corrected irradiation duration Corrected number of irradiation repetitions And the stop-irradiation flag corresponding to abnormal crop response. By combining these parameters, a modified set of wavelength control parameters is generated, represented as follows:

[0168]

[0169] in, Indicates the number is The corrected wavelength control parameter set corresponding to the candidate weed targets; when At that time, the corrected wavelength control parameter set can be used as the control input for continued irradiation or compensated irradiation; when At that time, the modified wavelength control parameter set carries a stop irradiation flag, causing the execution process to stop applying energy to the corresponding candidate weed target.

[0170] It should also be noted that by implementing irradiation according to the wavelength control parameter set and collecting temperature response, spectral response, and morphological response, predictive control is transformed into response-driven closed-loop correction control. The actual heat dose is calculated based on the temperature response of the growing point area, and the response deviation is obtained by comparing it with the predicted heat dose. When the actual heat dose is insufficient, the energy parameters are corrected according to the adjustment priority of irradiation duration, number of irradiation repetitions, and input power. When the temperature, spectral, or morphological response of adjacent crop areas is abnormal, the energy parameters are reduced or irradiation is stopped. This allows the corrected wavelength control parameter set to adapt to the actual field response, improve inactivation consistency, and reduce the risk of crop damage.

[0171] One embodiment of the present invention provides a hybrid wavelength weed control system based on target feature analysis, including a feature acquisition module, a tissue modeling module, a process prediction module, an inactivation determination module, a wavelength control module, and a response correction module.

[0172] The system comprises the following modules: Feature Acquisition Module (MAC Module) collects multi-source data of target plants in the treatment area and generates target feature vectors; Tissue Modeling Module (TM Module) establishes a target plant tissue state model based on the target feature vectors; Process Prediction Module (MP Module) generates candidate mixed wavelength combinations based on a preset wavelength library and the target plant tissue state model, inputs these combinations into the target plant tissue state model, and obtains prediction results of energy deposition, phase transition consumption, and heat diffusion processes in each region; Inactivation Determination Module (IM Module) determines whether the growth point region meets the constraints of heat dose, continuous temperature rise, and crop safety based on the process prediction results, and obtains inactivation determination conditions; Wavelength Control Module (W Module) selects wavelength combinations, target points, and energy parameters based on the inactivation determination conditions to form a wavelength control parameter set; and Response Correction Module (W Module) implements irradiation according to the wavelength control parameter set, collects temperature response, spectral response, and morphological response, and generates a corrected wavelength control parameter set based on the response deviation.

[0173] Reference Figure 2 This embodiment also provides a computer device applicable to the hybrid wavelength weed control method based on target feature analysis, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the hybrid wavelength weed control method based on target feature analysis as proposed in the above embodiment.

[0174] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0175] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the hybrid wavelength weed control method based on target feature analysis as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

Claims

1. A hybrid wavelength weed control method based on target feature analysis, characterized in that, include: Collect multi-source data of target plants in the area to be processed and generate target feature vectors; Establish a target plant tissue state model based on the target feature vector; Candidate mixed wavelength combinations are generated based on a preset wavelength library and input into the target plant tissue state model to obtain the prediction results of the action process in each region; Based on the analysis of the prediction results of the action process, the growth point inactivation requirements and crop safety constraints in the growth point region are obtained to determine the inactivation criteria. Based on the inactivation determination criteria, wavelength combinations, target sites, and energy parameters are selected to form a wavelength control parameter set. Irradiation is performed according to the wavelength control parameter set, and temperature response, spectral response and morphological response are collected. A corrected wavelength control parameter set is generated based on the response deviation.

2. The hybrid wavelength weed control method based on target feature analysis as described in claim 1, characterized in that: The generated target feature vector includes, The data acquisition process takes the same candidate plant target within the area to be processed as the object, and acquires visible light images, selected band reflectance data and depth data, and transforms the visible light images, selected band reflectance data and depth data to the same target coordinate system; Candidate plant targets are separated based on differences in color, band reflectance, and height. Candidate weed targets are determined based on the positional relationship between candidate plant targets and the crop protection area defined by the crop row centerline; Spectral features, geometric features, water content features, and growth point exposure features are extracted from candidate weed targets and associated with them according to the same target number to generate target feature vectors.

3. The hybrid wavelength weed control method based on target feature analysis as described in claim 2, characterized in that: The target plant tissue state model includes, Using the target feature vector as input, the leaf region, stem region, stem base region, and growth point region of the target plant are determined based on the geometric structural features in the target feature vector. Based on the spectral features and water content features in the target feature vector, the equivalent absorption, equivalent scattering and equivalent penetration depth at the corresponding candidate wavelengths are matched for each tissue region of the target plant. Based on the geometric structure features and water content features in the target feature vector, equivalent specific heat, equivalent thermal conductivity and phase change consumption parameters are matched for each tissue region of the target plant. Based on the characteristics of the growth point exposure state, an occlusion attenuation relationship is set to generate a target plant tissue state model that includes tissue regions, optical parameters, thermal parameters, and the location of the growth point.

4. The hybrid wavelength weed control method based on target feature analysis as described in claim 3, characterized in that: The predicted results of the action process in each region include, The preset wavelength library is pre-screened based on the target plant tissue state model to obtain an effective wavelength set; Select two different wavelengths from the preset wavelength library, and generate candidate mixed wavelength combinations according to the input power range, irradiation duration range, pulse width range, and spot size range; The candidate mixed wavelength combination is input into the target plant tissue state model, and the energy deposition distribution of each region is obtained based on the absorption degree, penetration depth, incident angle and light spot coverage of different wavelengths in each tissue region. When the water-bearing state characteristics indicate that the corresponding region has entered the water vaporization stage, the corresponding portion of the input energy will be included in the phase change consumption. After deducting the phase change consumption, the temperature change process caused by the diffusion of remaining heat to the growth point area and adjacent crop areas is predicted based on the spatial relationship between tissue regions and thermal conductivity parameters. The temperature change process is then used as the result of heat diffusion to generate the prediction result of the action process.

5. The hybrid wavelength weed control method based on target feature analysis as described in claim 4, characterized in that: The conditions for obtaining inactivation determination include: Based on the predicted results of the action process, the temperature change, effective heat dose, and duration of continuous temperature rise in the growth point region are extracted. Based on the preset tissue inactivation baseline parameters corresponding to the target plant, the baseline inactivation temperature threshold, baseline duration threshold, and baseline heat dose threshold of the growth point region are obtained. Based on the water content, growth point exposure, and stem base size of the target plant, the baseline duration threshold and baseline heat dose threshold are modified to obtain the duration threshold and heat dose threshold, and the baseline inactivation temperature threshold is used as the inactivation temperature threshold. When the predicted heat dose in the growth point region reaches the heat dose threshold, the sustained temperature rise time reaches the duration threshold, and the highest predicted temperature reaches the inactivation temperature threshold, the growth point inactivation requirement is deemed to be met. Based on the prediction results of the action process, the temperature change process and heat load change process of the adjacent crop area are extracted. When the temperature of the adjacent crop area does not exceed the crop safety temperature threshold and the heat load of the adjacent crop area does not exceed the crop safety heat load threshold, the adjacent crop area is determined to meet the crop safety constraints. Based on the results of the determination of growth point inactivation requirements and crop safety constraints, inactivation determination conditions are generated.

6. The hybrid wavelength weed control method based on target feature analysis as described in claim 5, characterized in that: The wavelength control parameter set includes... Based on the inactivation criteria, candidate mixed wavelength combinations that do not simultaneously meet the growth point inactivation requirements and crop safety constraints are excluded. Select the wavelength combination from the remaining candidate mixed wavelength combinations that meets the thermal dose requirements and whose total input energy meets the preset energy limit; Determine the target point based on the location of the growth point region; Based on the difference between the predicted heat dose and the heat dose threshold in the growth point region, the input power, irradiation duration, pulse width, and number of irradiation repetitions are determined. The wavelength combination, target point, input power, irradiation duration, pulse width, and number of irradiation repetitions are combined to form a wavelength control parameter set.

7. The hybrid wavelength weed control method based on target feature analysis as described in claim 6, characterized in that: The process of generating the corrected wavelength control parameter set based on the response deviation includes... Based on the wavelength control parameter set, the corresponding wavelength is controlled to irradiate the target plant according to the determined target point and energy parameters; During irradiation, temperature responses in the growing point region and adjacent crop areas were collected. After irradiation, changes in reflection in selected bands and morphological changes in the irradiated area were collected. The collected temperature response is compared with the predicted results of the action process to obtain the response deviation; The actual heat dose is calculated based on the temperature response of the growth point region, and the heat dose deviation is obtained based on the difference between the actual heat dose and the predicted heat dose of the growth point region. When the actual heat dose is lower than the predicted heat dose in the growth point region, the energy parameters are corrected according to the adjustment priority of irradiation duration, number of irradiation repetitions, and input power. When the temperature response in the adjacent crop area exceeds the crop response temperature threshold, reduce the input power, shorten the irradiation duration, and reduce the number of irradiation repetitions. When the spectral and morphological responses of neighboring crop areas exceed the corresponding abnormal thresholds, irradiation is stopped, and a corrected set of wavelength control parameters is generated based on the energy parameter adjustment results and the irradiation stop results.

8. A hybrid wavelength weed control system based on target feature analysis, employing the hybrid wavelength weed control method based on target feature analysis as described in any one of claims 1 to 7, characterized in that: It includes a feature acquisition module, an organization modeling module, a process prediction module, an inactivation determination module, a wavelength control module, and a response correction module; The feature acquisition module is used to acquire multi-source data of the target plants in the area to be processed and generate target feature vectors; The tissue modeling module is used to establish a target plant tissue state model based on the target feature vector; The process prediction module is used to generate candidate mixed wavelength combinations based on a preset wavelength library, and input them into the target plant tissue state model to obtain the prediction results of the action process in each region. The inactivation determination module is used to analyze the inactivation requirements of the growth point region and crop safety constraints based on the prediction results of the action process, and to obtain the inactivation determination conditions. The wavelength control module is used to select wavelength combinations, target points and energy parameters according to the inactivation determination conditions to form a wavelength control parameter group. The response correction module is used to implement irradiation according to the wavelength control parameter set, collect temperature response, spectral response and morphological response, and generate a corrected wavelength control parameter set according to the response deviation.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the hybrid wavelength weed control method based on target feature analysis as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the hybrid wavelength weed control method based on target feature analysis as described in any one of claims 1 to 7.

Images