165 results about "Crop mapping" patented technology

The invention discloses a close planting farmland growth vigor assessment method and system based on image processing, and relates to the field of agricultural information, and the method comprises the following steps: S1, multi-source data collection and preprocessing; s2, improving image segmentation, and extracting crop features; and S3, multi-dimensional growth vigor evaluation. According to the method, the field block level, the plant level and the whole growth period are covered through multi-source data collection, a generative adversarial network is used for repairing and shielding the plant image and restoring complete form information, the segmentation problem in a close planting scene is solved, the accuracy of close planting crop image analysis is improved, accurate registration of multi-modal data is achieved by means of feature point matching, and the accuracy of close planting crop image analysis is improved. The graph neural network optimizes image segmentation, effectively distinguishes overlapped leaves and stalks, deeply fuses multi-modal features and dynamically selects a fusion strategy, improves feature distinguishability, constructs a dynamic adaptive evaluation model, improves generalization ability and evaluation precision, identifies and intervenes abnormities in real time, and improves crop anti-risk ability and yield prediction accuracy.

The invention relates to the technical field of agricultural environment intelligent regulation and control, in particular to an agricultural environment intelligent regulation and control method and system based on a cloud platform, and the method comprises the steps: collecting farmland environment parameters and crop image data through a multi-source sensor, filtering the farmland environment parameters and crop image data through an edge computing node, and transmitting the filtered farmland environment parameters and crop image data to the cloud platform; preprocessing the farmland environment parameters and the crop image data, and extracting time domain and space domain features; an agricultural environment state index prediction model is constructed based on a multi-head self-attention encoder, training of the agricultural environment state index prediction model is completed through historical data, the agricultural environment state index prediction model is deployed as a cloud service API, farmland environment parameters and crop image data input in real time are processed online, a control instruction is generated according to a prediction result, and the control instruction is sent to the cloud service API. Accurate and intelligent regulation and control of the agricultural environment are completed through Internet of Things protocol issuing equipment. Through dynamic feature modeling and multi-index collaborative optimization, the precision of agricultural environment regulation and control and the resource utilization efficiency are remarkably improved.

The invention relates to the technical field of agricultural intelligent monitoring, in particular to a crop growth detection system and method based on machine vision, and the system comprises an image collection module, a morphological feature capture module, a growth trend judgment module, an abnormal region marking module and a state information output module. According to the method, crop images are collected through a multi-angle camera, leaf contours, stem bending and plant spacing are extracted, multi-dimensional modeling of morphology is realized, structural change identification is enhanced, time sequence comparison of key morphological characteristics is realized, identification precision and time efficiency are improved, leaf and stem change trends are continuously analyzed, offset is quantified, and growth abnormity is early warned in advance; health degradation identification is combined with fluctuation area marking, dynamic monitoring is achieved, high-risk positioning is carried out on a time overlapping area, time-space locking is enhanced, precise management is assisted, the process is from image analysis to abnormal focusing, an information reasoning chain is established, and the monitoring precision and response efficiency are remarkably improved.

The invention relates to the technical field of water and fertilizer application evaluation, and particularly discloses an intelligent evaluation method for water and fertilizer utilization efficiency and an operation system, and the method comprises the steps: obtaining a crop image of a crop corresponding to a fertilization parameter; identifying the crop image, outputting a growth evaluation report, and constructing a mapping relation from the fertilization parameters to the growth evaluation report for evaluating the fertilization parameters; fertilization parameters are recorded, meanwhile, crop images are obtained through a visual monitoring point, the crop images are recognized, the growth conditions before and after fertilization are determined, then the growth difference before and after fertilization is determined, and the mapping relation from the fertilization parameters to the growth difference is constructed; the growth difference of the fertilization parameters can be predicted before the fertilization behavior, and the prediction accuracy is high.

The invention relates to a multi-modal fusion corn harvester header height adaptive adjustment method and system, and belongs to the technical field of automation control, the method comprises the following steps: using a binocular camera to collect crop images, and preprocessing the crop images to obtain a depth map; the image is input into a YOLOV8n model, and coordinates of a center point at the bottom of a bounding box of the corncob are obtained through model operation; performing plane fitting on the depth map to obtain a ground height; calculating a corn ear height initial value based on the ground height and the center point coordinate; vibration data of working of the harvester is collected to perform vibration compensation on the initial height value to obtain a height correction value; and dynamically adjusting the header height of the harvester by adopting a dynamic fuzzy PID algorithm based on the correction value. According to the method, the target detection model and the dynamic fuzzy PID algorithm are combined, so that the problems of low visual detection precision and control response lag of existing agricultural machinery in a complex field scene are solved, and crop loss in harvesting operation is effectively reduced.

The present invention belongs to the technical field of crop mapping based on remote-sensing images, and relates to a method and system for crop mapping across large regions with low sample dependence. The method includes: acquiring remote sensing data, ground sample data, meteorological data, soil data, establishing geographically divided crop planting regions; establishing key growth period model libraries corresponding to individual crop regions; constructing machine learning models based on a plurality of machine learning algorithms, to obtain machine learning crop extraction models; selecting an optimal machine learning crop extraction model; acquiring a spatial crop distribution base map; performing product correction based on the disaster information; and acquiring a regional crop map using a target crop extraction model adapted for the disaster response. The present invention, achieve high-accuracy and large-scale crop mapping, and reduce the crop sample dependence of crop mapping.

The invention relates to a crop irrigation demand prediction method based on an intelligent integrated prefabricated pump station, and belongs to the technical field of artificial intelligence. The method comprises the following steps: collecting a crop image in a prefabricated pump station irrigation scene, obtaining auxiliary data, calculating a daily accumulated reference evapotranspiration value, and obtaining a time-synchronized multi-modal data set; labeling and dividing the multi-modal data set; constructing an irrigation demand prediction model, wherein the irrigation demand prediction model comprises a spectral reflectivity enhancement module, a double-flow separation convolution module, a cross gating fusion module, a region sensitive pyramid module and a coupling prediction head module; inputting data in the divided training set into the model, and training the model through weighted total loss to obtain a trained model; inputting a to-be-detected image into the trained model to obtain a soil water content and transpiration prediction result; and performing irrigation decision based on the soil water content and transpiration prediction result. The accuracy of irrigation prediction can be improved.

The invention relates to a field water and fertilizer intelligent control method based on the Internet of Things, and the method comprises the steps: carrying out the crop type recognition and growth stage recognition through collecting crop images, soil water and fertilizer data and meteorological prediction data, and generating an initial water and fertilizer control scheme of a current time node; a soil water and fertilizer prediction model is created to predict the soil water and fertilizer change trend in a specified time period, the initial water and fertilizer control scheme of the current time node is adjusted according to the predicted soil water and fertilizer change trend, growth cycle characteristics corresponding to crop types are obtained based on big data, and a dynamic water and fertilizer control scheme is formulated. A water and fertilizer adjusting control scheme is executed at a current time node, a dynamic water and fertilizer control scheme is executed at a specified time node, the field water and fertilizer intelligent control method integrating Internet of Things sensing, intelligent identification, trend prediction, dynamic planning and closed-loop feedback is realized, the whole process intelligence from data acquisition to scheme execution is realized, and the water and fertilizer utilization efficiency is improved. The agricultural production cost is reduced.

The invention relates to the technical field of fertilization and irrigation regulation and control, and discloses a fertilization and irrigation two-way regulation and control system under water demand and fertilizer demand coupling modeling, and the system constructs the fertilization and irrigation two-way regulation and control system based on the water demand and fertilizer demand coupling modeling. The multi-source data acquisition module is used for acquiring soil, crop growth period and meteorological data through a layered soil sensor, crop image acquisition equipment and a meteorological acquisition point; the water and fertilizer scheme generation module calculates irrigation amount, fertilization type and dosage based on data, and adjusts intervals in combination with temperature; the operation execution module controls the operation of the irrigation and fertilization equipment and records the state; and the data correction module compares the sensor with the sample data to generate a correction scheme. The system realizes accurate regulation and control through a closed loop of data acquisition, scheme generation, execution and correction, simultaneously covers equipment state monitoring and early warning, and ensures efficient fertilization and irrigation to adapt to crop requirements.

The invention relates to the technical field of biological breeding, in particular to a biological breeding management method and system based on big data. The method comprises the following steps: acquiring original germplasm data; multi-source heterogeneous data integration is carried out on the original germplasm data, the data integration comprises genotype sequencing integration and historical phenotype integration, and integrated germplasm data is generated and stored in a database in a distributed mode; obtaining original growth data, performing data preprocessing, generating standardized germplasm resource comprehensive data including a crop image feature set and an environment response parameter set, and synchronizing the data to the database; and performing whole genome selection prediction on the integrated germplasm data based on the standardized germplasm resource comprehensive data to generate a genotype-phenotype prediction model. By integrating germplasm resources, field management, breeding management, phenotype management, genotype management and whole genome selection, the accuracy and adjustment flexibility of biological breeding management are improved.

The invention relates to the technical field of agricultural irrigation, in particular to an intelligent irrigation control system and method based on an intelligent crop growth model. The method comprises the steps of collecting temperature, humidity, illumination and soil moisture data of a crop growth environment, and taking the data as input of a machine learning model; an LSTM machine learning algorithm is utilized to construct a crop growth model used for predicting the crop growth trend and calculating the irrigation water demand; in the crop growth process, crop image data are acquired in stages and compared with a standard growth image, similarity and a correction coefficient are calculated, and the predicted growth trend is corrected in real time; according to the corrected growth trend, the irrigation water demand is dynamically adjusted, and closed-loop feedback control is formed; according to the corrected irrigation water demand, the irrigation equipment is accurately controlled for irrigation, and intelligent monitoring of crop growth and efficient utilization of water resources are achieved.

The invention provides a multi-modal agricultural technology question and answer method and system, and belongs to the technical field of artificial intelligence, and the method comprises the steps: extracting a text feature vector, an image feature vector, a time sequence feature vector and a spatio-temporal context feature vector according to a query text and a crop image; calculating an initial fusion feature according to the spatio-temporal context feature vector, the image feature vector and the time sequence feature vector; when the query text contains a semantic entity of a preset type, calculating to obtain a final fusion feature according to the text feature vector and the initial fusion feature; splicing the final fusion feature and the text feature vector, and mapping the spliced vector to a space-time knowledge graph for reasoning to obtain a causal reasoning path; and generating a question and answer result according to the causal reasoning path. According to the method, deep alignment of time and space and semantics is carried out on the multi-modal agricultural data, and causal reasoning is carried out in combination with the knowledge graph with time and space constraints, so that the accuracy of a question and answer result is remarkably improved.

The invention provides an agricultural greenhouse digital cockpit monitoring method and system based on Unity 3D and AIoT, and relates to the technical field of agricultural information, and the system realizes accurate monitoring and regulation of a greenhouse environment through multi-modal data acquisition, intelligent decision and virtual-real interaction. The method comprises the following steps: constructing a three-dimensional virtual model simulation environment; the Internet of Things nodes collect data such as temperature, humidity and illumination, and the inspection vehicle obtains crop images and preprocesses the crop images; a YOLOv8 model library and an OfficientDet-D7 model library are adopted to recognize diseases and insect pests, and the accuracy is improved by combining multi-model voting; a regulation and control strategy is generated based on the crop growth model, and remote control and optimization are performed through a VR cockpit. The system comprises an environment sensing unit, a mobile inspection device, a cloud platform, a VR interaction terminal and a regulation and control execution system. The Internet of Things node communicates with the cloud, the inspection vehicle performs full-ridge inspection, the cloud supports disease and insect pest recognition and strategy generation, and the VR cockpit provides immersive interaction. According to the method, multi-source data, edge intelligence and reinforcement learning are fused, and the data credibility and the recognition accuracy are improved.

The invention relates to the field of intelligent greenhouse management, and particularly discloses an intelligent greenhouse management system based on computer vision and deep learning, and the system comprises a data collection module which obtains environment data and soil data in a greenhouse in real time; the image acquisition module is used for acquiring crop growth images; the core controller preprocesses the data acquired by the data acquisition module and the image acquisition module and uploads the data to the cloud database; the computer vision and deep learning analysis module is used for identifying crop types and pest and disease damage conditions based on the crop images; the water-fertilizer control unit comprises an intelligent control module, a water-fertilizer all-in-one machine and a drip irrigation system; the user interaction module provides remote monitoring and control functions for a user; according to the system and the method, the condition of economic loss caused by the fact that crop diseases and insect pests cannot be correctly treated due to insufficient experience in traditional agriculture is overcome, environmental visualization is realized, a farmer can conveniently monitor the greenhouse in real time, precise irrigation and fertilization are realized, the agricultural production efficiency is improved, and resource waste is reduced.

The invention relates to an insect pest detection processing method, system and device, and belongs to the field of insect pest detection.The method comprises the steps that a crop image in a target area is collected, black spot recognition and malformation recognition are conducted on the crop image, and black spot parameters and malformation parameters are obtained; according to the black spot parameter and the malformation parameter, calculating to obtain a first insect pest parameter, and configuring an insect body detection range and an insect body detection scale; collecting a detection image set, and carrying out insect body recognition to obtain an insect body recognition result; and processing according to the insect body identification result to obtain a second insect pest parameter, and calculating to obtain an insect pest parameter as an insect pest detection processing result in combination with the first insect pest parameter. According to the method, the technical problems of inaccurate detection and low treatment efficiency caused by the fact that pest distribution characteristics are not considered in mangosteen pest detection in the prior art are solved, and the technical effects of analyzing the pest distribution characteristics through crop pest symptoms, optimizing a detection strategy and improving pest detection accuracy and treatment efficiency are achieved.

The invention provides an unmanned aerial vehicle pesticide spraying area recognition method and system combined with image recognition, and relates to the technical field of agricultural unmanned aerial vehicle operation, and the method comprises the steps: firstly obtaining an image set which is collected by an unmanned aerial vehicle and comprises farmland scene images at different time periods, and carrying out the region semantic association analysis to generate a farmland region semantic association graph; next, an image recognition model is called to perform growth state recognition on the crop image part, and a spraying area adaptation model is constructed in combination with the semantic association graph; then, based on the spraying area adaptation model, evaluating a spraying demand, dividing farmland sub-areas and generating a spraying area division scheme; determining pesticide spraying parameters of each sub-region according to the spraying region division scheme, and generating an unmanned aerial vehicle pesticide spraying instruction; finally, the instruction is sent to an unmanned aerial vehicle control system, the unmanned aerial vehicle is controlled to execute pesticide spraying operation, and therefore the pesticide spraying accuracy and uniformity can be improved.

The invention provides a navigation line generation method and device, a storage medium and a program product, and relates to the technical field of agricultural production. According to the method, the target area is segmented from the crop image, the target area refers to the crop area and / or the land area, then the reference line of the target area is extracted, the weight coefficient of the reference line is acquired, and the navigation line of the agricultural equipment can be generated according to the weight coefficient and the reference line, so that the navigation line of the agricultural equipment can be generated in the operation process of the agricultural equipment. The navigation line is dynamically generated by integrating the reference line of the target area, so that the accuracy and the real-time performance of agricultural machinery navigation are improved, the artificial interference is reduced, the adaptive capacity of the system to a complex operation environment is enhanced, the operation cost can be effectively saved, and the operation efficiency can be effectively improved.

The invention discloses an agricultural greenhouse water and fertilizer management system and method based on video recognition, and relates to the technical field of computer image recognition, and the system comprises a camera which is used for collecting crop images; the environment sensor is used for collecting water and fertilizer environment data of the agricultural greenhouse; the local host is provided with a crop identification model, and the crop identification model is used for identifying the real-time growth state of crops in the crop image; the cloud server is deployed with a growth decision model and a growth regulation model, the growth decision model formulates a growth control strategy according to the types and growth stages of crops in the agricultural greenhouse and the water and fertilizer environment data, the growth regulation model calculates an income value, and a regulation scheme for the growth control strategy is solved by taking the maximum income value as a target; and forming an adjusted growth control strategy. According to the method, the growth state of the crops is considered on the whole instead of achieving the optimal state of some crops, and the benefit of the agricultural greenhouse is improved.

The invention relates to the technical field of intelligent agricultural intelligent irrigation control, and discloses an intelligent irrigation analysis control system based on a crop growth period. The system comprises a crop image sequence, soil humidity profile data and a regional meteorological sequence. And the data analysis module is used for performing multi-stage growth state interpretation on the image sequence, generating a crop structure state profile in each growth period, performing vertical moisture exchange identification on the soil humidity data, and generating a root zone moisture dynamic map. And the habitat mapping and stress evaluation module synchronously executes habitat adaptability mapping and moisture stress risk mapping. And the irrigation decision module corrects the habitat mapping result according to the stress evaluation result and generates an irrigation decision profile. And the control instruction generation module generates a partition valve control instruction in combination with the decision profile and the alternate irrigation rule. According to the invention, deep fusion analysis of the crop growth state, the root zone moisture and the meteorological environment is realized, and the irrigation decision accuracy and the water resource utilization efficiency are improved.

The invention relates to the technical field of vegetable cultivation, and discloses an automatic light supplement control method and system for vegetable cultivation, and the method comprises the steps: determining the growth state of a vegetable crop based on crop image information and a crop growth model, and judging whether to execute a light supplement strategy or not according to a solar altitude, determining a backtracking light supplementing strategy or an initial light supplementing strategy according to the occurrence frequency of the growth state in the historical light supplementing database, determining target light supplementing data of the vegetable crops according to the number of the light supplementing data, and when the initial light supplementing strategy is determined, determining the target light supplementing data of the vegetable crops based on a recurrent neural network model, and analyzing the historical light supplement record to determine a light supplement swing factor of the light supplement lamp, adjusting the target light supplement data based on the light supplement swing factor, and performing light supplement on the vegetable crops according to the adjusted target light supplement data. According to the invention, the reliability of light supplement is ensured through the crop growth model and the light supplement swing factor.

This disclosure relates generally to system and method to determine crop growth stage nutrient deficiencies. Diagnosing correct nutrient deficiencies in the plant is very challenging based on plant image analysis during the cropping season. The method of the present disclosure enables assessing nutrient deficiency using image processing techniques according to current crop growth stage. The method receives from an image capturing device a plurality of crop images of one or more crop fields. Further, trained single shot deep learning network determines a crop growth stage from a plurality of crop growth stages for each crop by extracting a plurality of morphological features. Then, the health state of the crop is determined based on a balanced plant nutrition index (BPNI) value. Further, a plurality of nutrient deficiencies corresponding to the current crop growth stage of the unhealthy crop. Further, a total nutrient deficiency score for deficient nutrients of the unhealthy crop.

The invention relates to the field of crop harvesting path planning, in particular to a harvesting path planning method and system for an unmanned harvester, and the method comprises the steps: obtaining real-time RGB images of crops, collecting historical mature and immature crop images, comparing the RGB channel feature differences of the historical mature and immature crop images, and determining the harvesting path of the unmanned harvester according to the RGB channel feature differences of the historical mature and immature crop images. Selecting a channel with the maximum difference as a maturity index, and further dividing the real-time RGB image of the crop to obtain a mature area and an immature area; and determining all candidate loading and unloading points, for each candidate loading and unloading point, calculating the sum of the path lengths from all the mature areas to the loading and unloading points under the set harvesting sequence, selecting the loading and unloading point with the minimum sum of the path lengths and the harvesting sequence to construct an optimal function, and solving to obtain an optimal harvesting path. Harvesting is performed in a mature area in a targeted mode, time and resources are prevented from being wasted in an immature area, and therefore the overall harvesting efficiency is improved.

The invention discloses a farmland irrigation control system, and the system comprises an acquisition module which records a crop image of a target crop, introduces the image into a database, and obtains a humidity threshold value of the target crop and soil data of target soil; the analysis module is used for extracting a first shape feature and a color feature of the crop image, analyzing the first shape feature and the color feature to obtain the rhizome depth of the target crop, and analyzing the soil data to obtain the threshold level of the target soil; the decision-making module analyzes the depth of the rhizome to obtain a humidity monitoring range, the acquisition module detects humidity data of the humidity monitoring range, adjusts a humidity threshold value in combination with a threshold value level, generates a first difference value and makes an irrigation decision; according to the method, a double analysis strategy of appearance features and color features is adopted, the growth cycle is accurately identified, the problem of irrigation depth misalignment caused by erroneous judgment in the growth stage in a traditional method is thoroughly solved, and the effect of improving the irrigation depth matching degree is achieved.

The invention relates to the technical field of agricultural pest recognition, and provides an agricultural pest pattern recognition method and system based on computer vision, and the method comprises the steps: receiving a crop image of a to-be-diagnosed crop in a field, carrying out the preliminary diagnosis of the crop image, obtaining a preliminary diagnosis result, and recognizing key visual features supporting the preliminary diagnosis; generating a visually similar disease candidate set according to the preliminary diagnosis result; performing counter-example comparative analysis on each disease in the disease candidate set to obtain excluded similar diseases and generate exclusion reasons of the excluded similar diseases; providing and displaying a preliminary diagnosis result and typical case reference pictures of the eliminated similar diseases for agronomists; and integrating the preliminary diagnosis result, the key visual features, the elimination reason and the typical case reference picture, generating a structured diagnosis report, and outputting the structured diagnosis report to realize agricultural pest pattern recognition. The method has the effect of improving the accuracy and practicability of agricultural pest pattern recognition.

The invention relates to the technical field of agricultural automation equipment, and discloses a water and fertilizer integrated irrigation device and method based on the Internet of Things and an intelligent strategy, and the device comprises a crop and environment sensing unit which obtains crop image information and environment parameters; the intelligent decision-making unit is used for generating a target irrigation strategy and fertilization concentration; the source water management unit is used for analyzing a water source before irrigation and providing water quality background information; the water-fertilizer synthesis and control unit is used for dynamically synthesizing and outputting a nutrient solution in real time by combining the target concentration and the water quality background; and the cloud management and optimization unit is used for collecting data of each unit and is used for storage, remote monitoring and management. According to the invention, an on-line spectrum analysis technology and a high-frequency pulse microfluidic technology are coupled, a real-time closed-loop control system for directly measuring chemical components is constructed, and high consistency of a digital decision instruction and the chemical components of a physically output nutrient solution is achieved.

The invention discloses a control method and equipment for water-saving irrigation of a mobile device and a medium, and relates to the technical field of intelligent irrigation. The method comprises the following steps: acquiring a crop image, geographic information data and soil moisture content distribution data of a target area, and collecting leaf surface temperature, illumination intensity and wind speed; introducing meteorological prediction data, and generating an irrigation decision matrix based on the crop image, the soil moisture content distribution and the meteorological prediction data; calculating an optimal moving path according to the geographic information and the irrigation decision matrix; according to the optimal moving path, the moving device is controlled to implement hierarchical variable sprinkling irrigation operation, and the moving device is monitored in real time. According to the method, the accuracy and efficiency of agricultural irrigation are remarkably improved through multi-source data fusion, dynamic path planning and intelligent control.

The invention discloses an agricultural machinery operation planning method and system based on an unmanned aerial vehicle, and the method comprises the steps: obtaining a high-resolution remote sensing image of a target farmland region, and carrying out the denoising preprocessing of the remote sensing image through a pre-constructed convolutional neural network model, and obtaining a preprocessed image; generating a preliminary farmland operation flight path planning scheme by adopting a deep reinforcement learning algorithm and combining performance parameters and energy consumption constraint conditions of the unmanned aerial vehicle according to crop growth stage information acquired from the crop image; in the farmland operation process, farmland environment parameter data are collected in real time through an airborne sensor, and the flight attitude and track parameters of the unmanned aerial vehicle are dynamically adjusted by adopting a Kalman filtering algorithm so as to ensure that the collection quality of remote sensing images meets farmland monitoring requirements.

The invention provides a phenotype data measurement method and system in a semi-automatic field scene, and belongs to the field of phenotype data measurement in an outdoor field scene, and the method comprises the steps: collecting a crop image, and obtaining a target phenotype center coordinate point specified by a user according to the crop image; segmenting the crop image and a target phenotype center coordinate point specified by a user through a pre-trained SAM model to obtain a target phenotype mask image; the morphological characteristics of the mask image are obtained by judging the bending characteristics of the mask image, and the morphological characteristics comprise an upright form and a bending form. According to the plant phenotype data calculation method, data collection, calibration and model training are not needed, the method can be widely applied to various plants and scenes, and a large amount of cost caused by data annotation and model training can be avoided.

The invention discloses a crop disease and pest control pesticide spraying method and system based on image recognition, and the method comprises the steps: collecting crop images through carrying multiple cameras by an unmanned aerial vehicle, inputting the crop images into a double-branch deep learning model after preprocessing, and enabling the model to comprise an FCA frequency domain feature extraction network, a DAT-Transform spatial domain feature extraction network and an MSAF fusion module, precise identification of diseases and pests is realized; further integrating real-time environment parameters such as temperature and humidity, wind speed and crop growth stages, constructing a dynamic pesticide amount calculation model, and outputting personalized pesticide spraying amount; the pesticide spraying mechanism is driven by the controller to execute pesticide spraying, effect data is collected after a prevention and treatment period, an optimization recognition model and pesticide amount calculation parameters are fed back, and a recognition-pesticide spraying-feedback-iteration closed-loop prevention and treatment system is formed. The method solves the problems of low recognition precision, rigid pesticide amount regulation and control and lack of a continuous optimization mechanism in a complex environment in the prior art, improves the recognition precision, saves the pesticide amount, and improves the control effect.

The invention discloses an agricultural greenhouse automatic dimming image acquisition method based on visual task feedback, and relates to the field of agricultural data acquisition, and the method comprises the steps: collecting a crop image through a depth camera, carrying out the target detection, extracting an ROI region of a crop, and carrying out the target detection based on an RGB image and a depth image in a detection frame of the crop; the method comprises the steps of calculating a comprehensive image quality index used for representing the overall texture richness and the overall detection reliability of each crop, and then dynamically adjusting the driving current of an LED lamp panel based on the comprehensive image quality index so as to dynamically adjust light. Even if the illumination conditions of different areas of the agricultural greenhouse change and the crops have the phenomena of strong reflection, local shielding and the like, closed-loop dimming can be carried out based on the local image quality of the crops, so that the quality of the collected crop images is improved, and the accuracy and reliability of image analysis are improved.