A digital management system and method for a prediction model of Fusarium head blight of wheat

Abstract

The present invention discloses a digital management system and method for a prediction model of wheat scab, which relates to the field of digital big data technology. The present invention collects the meteorological environment data of wheat growth in the observation area within a period through a meteorological collection device, and collects the wheat images in the observation area within a period by using an image collection device; based on the collected environmental data, a data matrix of wheat growth environmental data within the corresponding period is constructed with the collection time point as the data label; the diseased ear rate of the wheat environmental data in the current period is analyzed in combination with the historical wheat disease lesion environmental data; the abnormal pixel points are located by combining the collected images with the chromaticity information of the pixel points of the historical wheat images; the wheat area is divided comprehensively according to the abnormal pixel points of the current wheat images, and the lesion characteristics are analyzed in combination with the divided sub-regions, and the diseased sub-regions are predicted; the diseased ear rate analysis and area division information of the current wheat growth area are visually output, and the predicted diseased areas are warned.

Description

Technical Field

[0001] The present invention relates to the technical field of digital big data, and specifically to a digital management system and method for a wheat scab prediction model. Background Art

[0002] The prediction of wheat scab utilizes multiple factors such as meteorological data, crop growth conditions, and historical disease occurrence data, and through methods such as statistical analysis and machine learning, an analysis model is established to predict the occurrence trend and severity of wheat scab; by analyzing the data, early warning of the disease occurrence is carried out to reduce the disease loss.

[0003] During the current wheat growth cycle, there is a possibility of being affected by scab in multiple stages. Precise prediction of the occurrence of scab during the wheat growth cycle is an important means to protect wheat growth; in the current mode, the detection of scab in most wheat is mostly based on manual experience and full-cycle monitoring for manual judgment. However, this detection method has a large error in accuracy and cannot accurately locate large-scale wheat production areas, with a large prediction range error and accuracy error. Summary of the Invention

[0004] The purpose of the present invention is to provide a digital management system and method for a wheat scab prediction model to solve the problems raised in the prior art.

[0005] To achieve the above purpose, the present invention provides the following technical solutions:

[0006] A digital management method for a wheat scab prediction model, the method comprising the following steps:

[0007] S100. Collect the meteorological environment data of wheat growth in the observation area during the period through a meteorological collection device, and simultaneously collect the wheat images in the observation area during the period by using an image collection device;

[0008] S200. Based on the collected environmental data, construct a wheat growth environment data matrix for the corresponding period with the corresponding collection time point as the data label; analyze the diseased ear rate of the current period wheat environmental data by combining the historical wheat historical lesion environmental data.

[0009] S300. Perform abnormal pixel point positioning processing on the collected images by combining the chromaticity information of the pixel points of the historical wheat images; comprehensively divide the wheat area based on the abnormal pixel points of the current wheat images, and combine the divided sub-regions to perform lesion feature analysis and predict the diseased sub-regions.

[0010] S400. Visualize and output the diseased ear rate analysis and regional division information of the current wheat growth area, and warn the predicted diseased areas.

[0011] The specific steps for the S100 to collect the meteorological environment data of wheat growth in the observation area during a period through a meteorological collection device and simultaneously collect the wheat images in the observation area through an image collection device are as follows:

[0012] S101. Collect the real-time growth environment data of wheat through the meteorological equipment of the wheat growth area device; obtain the growth environment data of wheat within the corresponding period by setting the collection period; transmit the collected data to the analysis end through regional networking; the meteorological equipment includes a temperature sensor, a humidity sensor, a light sensor, etc.; the wheat growth environment includes temperature, humidity, light, wind force, etc.

[0013] S102. Conduct full-area image collection of the wheat growth area through an image collection device, and remotely transmit the collected image data to the analysis end through an airborne communication device; the image collection device is a high-precision image sensor.

[0014] The specific steps for the S200 to construct a wheat growth environment data matrix within a corresponding period with the corresponding collection time point as the data label based on the collected environment data and analyze the diseased ear rate of the current period wheat environment data by combining the historical wheat historical lesion environment data are as follows:

[0015] S201. Retrieve the wheat growth environment data collected during the period, classify and process the corresponding collection environment data types by determining the collection time point; construct a corresponding environmental type data set for the environmental data collected at each time point; use the collection time point as the matrix data label, and overall plan the environmental type data sets at each time point collected during the period in chronological order to construct a periodic wheat growth environment data matrix.

[0016] S202. Retrieve the growth environment data matrix of wheat in the current observation area, extract the environmental data corresponding to each environmental type in the environmental data sets at each time point in the matrix, and conduct a periodic comprehensive analysis of the environmental data of each type. The calculation formula is

[0017]

[0018] where PVE(n:T) is the periodic equilibrium value of the environmental data with type number n within the corresponding period T; M(t∈T) is the number of time points t within the period T; t∈T is the time point t within the period T; H(n:t) is the environmental data value of type number n at each time point t within the corresponding period T.

[0019] Based on the periodic equilibrium values of the environmental data of each type within the period, analyze the scab diseased ear rate of the observed wheat within the period. The calculation formula is

[0020]

[0021] Where YDP(T) is the scab ear rate of wheat observed within period T; Fn is the scab impact coefficient of environmental data corresponding to type number n; by retrieving historical environmental data of wheat scab, and taking the period equilibrium value of each type of environmental data in the current period as a reference, retrieving the scab ear rate data of wheat in the corresponding periods of historical same environmental data; by performing error analysis on the retrieved historical scab ear rate data and the scab ear rate in the current period, if |YDP(T) - YDP(AL)| > E, then re-collect the current observed wheat environmental data and re-perform the above environmental data analysis steps; where YDP(AL) is the mean value of historical scab ear rates of wheat under the same environmental data in the current period; E is the error control parameter;

[0022] If |YDP(T) - YDP(AL)| ≤ E, then analyze the comprehensive impact ratio of the diseased ear rate of the current observed wheat, and its calculation formula is

[0023]

[0024] Where, POI(T) is the comprehensive impact ratio of the diseased ear rate of wheat observed in the current period; YDP(Amax) and YDP(Amin) are respectively the maximum value and the minimum value of the historical scab ear rate data of wheat under the same environmental data in the current period.

[0025] The S300 performs abnormal pixel point positioning processing by combining the acquired image with the chromaticity information of pixel points in the historical wheat image; comprehensively divides the wheat area based on the abnormal pixel points in the current wheat image, and combines the divided sub-areas to perform lesion feature analysis. The specific steps for predicting the diseased sub-areas are as follows:

[0026] S301. Retrieve the growth image data of wheat in the area within the period, and obtain the chromaticity data of each pixel point in the image through image local window magnification processing; the chromaticity data is the RGB three-channel intensity value corresponding to the pixel point; establish a coordinate system to obtain the coordinates of each pixel point in the image, and through constructing a data association channel, associate the chromaticity data of the corresponding pixel point with the coordinate data of the corresponding pixel point; the association processing is to obtain the corresponding chromaticity data by retrieving the coordinate information corresponding to the pixel point; construct the corresponding chromaticity feature vector based on the chromaticity data of each pixel point in the coordinate system, and perform screening analysis by combining the feature vectors of each pixel point with the standard chromaticity feature vector of historical healthy wheat, and its calculation formula is

[0027]

[0028] Where, D[(x,y):HE] refers to the similarity between the chromaticity feature vector of the pixel point with coordinates (x,y) and the standard chromaticity feature vector of historical healthy wheat. is the chromaticity feature vector of the pixel point corresponding to the coordinates (x, y); is the standard chromaticity feature vector of historical healthy wheat; and respectively correspond to the norms of the chromaticity feature vector of the pixel point corresponding to the coordinates (x, y) and the standard chromaticity feature vector of historical healthy wheat; where the dot product of vectors is used for the calculation between vectors; the chromaticity feature vector of the corresponding coordinates is where I R , I G , I B respectively correspond to the chromaticity intensity values of the R, G, and B channels of the coordinate pixel points; by setting the health parameter Dp, if D[(x, y):HE] ≤ Dp, it is determined that the pixel point corresponding to the current coordinate is a normal pixel point; if D[(x, y):HE] > Dp, it is determined that the pixel point corresponding to the current coordinate is an abnormal pixel point; the abnormal pixel points are marked explicitly, and the normal pixel points are screened out;

[0029] S302. Based on the wheat area coordinate image data after pixel point screening processing, equal area division is performed, and labels are assigned to each sub-region; by analyzing the similarity between the abnormal feature points in each sub-region and the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat, the calculation formula is

[0030]

[0031] where, D[(x, y) j :SC] refers to the similarity between the chromaticity feature vector of the pixel point corresponding to the coordinates (x, y) in the corresponding sub-region number j and the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat; is the chromaticity feature vector of the pixel point with coordinates (x, y) in the corresponding sub-region number j, is the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat, is the norm of the chromaticity feature vector of the pixel point with coordinates (x, y) in the corresponding sub-region number j, is the norm of the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat;

[0032] Combining the similarity analysis data between the chromaticity feature vectors corresponding to the abnormal pixel points in each sub-region and the standard chromaticity feature vector of historical healthy wheat and the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat, the Fusarium head blight abnormal index of the wheat in the sub-region is analyzed, and the calculation formula is

[0033]

[0034] where, GI jis the scab abnormality index of wheat in the sub-region corresponding to the serial number j; f is the activation function; U j is the number of abnormal points in the sub-region numbered j; U a is the number of abnormal points in the entire region of the wheat image; D[(x,y) j :HE] is the similarity between the chromaticity feature vector of the pixel point with coordinates (x, y) in the sub-region corresponding to the serial number j and the standard chromaticity feature vector of historical healthy wheat; b is the bias parameter; (x, y) ∈ j are the coordinates of each pixel point in the sub-region corresponding to the serial number j; by introducing the comparison index GI(th), if GI j <GI(th), it is determined that there is no risk of scab occurrence in the wheat in the current sub-region corresponding to the serial number j; if GI j ≥GI(th), it is determined that there is a risk of scab occurrence in the wheat in the current sub-region corresponding to the serial number j; if there are multiple sub-regions with a risk of scab occurrence, the monitoring priorities are allocated according to the scab abnormality indices of the corresponding sub-regions from large to small.

[0035] The S400 visually outputs the analysis of the diseased ear rate and the regional division information of the current wheat growth area, and the specific steps for warning the predicted diseased area are as follows:

[0036] S401. Visually output the analysis data of the diseased ear rate, the analysis data of the pixel points of the regional image, and the regional division data of the current observed wheat growth area;

[0037] S402. Give a warning feedback to the sub-regions with a risk of scab occurrence, and monitor according to the priorities of each sub-region.

[0038] A digital management system for a wheat scab prediction model, the system includes a regional data acquisition module, a periodic diseased ear rate analysis module, a regional abnormality index analysis module, and a prediction output module;

[0039] The regional data acquisition module collects the meteorological environment data of wheat growth in the observation area during the period through meteorological acquisition equipment, and simultaneously collects the wheat images in the observation area during the period by using image acquisition equipment; the periodic diseased ear rate analysis module constructs a data matrix of wheat growth environment data in the corresponding period with the corresponding acquisition time point as the data label based on the collected environment data; analyzes the diseased ear rate of the current period wheat environment data by combining the historical wheat historical lesion environment data; the regional abnormality index analysis module locates the abnormal pixel points by combining the collected image with the chromaticity information of the pixel points of the historical wheat image; comprehensively divides the wheat area according to the abnormal pixel points of the current wheat image, and combines the divided sub-regions to analyze the lesion characteristics and predict the diseased sub-regions; the prediction output module visually outputs the analysis of the diseased ear rate and the regional division information of the current wheat growth area, and warns the predicted diseased area.

[0040] The regional data acquisition module includes a regional environment data acquisition unit and a wheat image data acquisition unit;

[0041] The regional environmental data acquisition unit collects real-time wheat growth environment data through the meteorological equipment of the wheat growth area device; obtains the wheat growth environment data within the corresponding period by setting the acquisition cycle; and transmits the collected data to the analysis end by using the regional networking;

[0042] The wheat image data acquisition unit acquires images of the entire wheat growing area through an image acquisition device, and remotely transmits the acquired image data to an analysis end through an airborne communication device.

[0043] The periodic diseased ear rate analysis module includes an environmental data processing unit and a wheat diseased ear rate analysis unit;

[0044] The environmental data processing unit retrieves the wheat growth environment data collected periodically, and classifies the collected environment data according to the corresponding collected environment data type by determining the collection time point; constructs a corresponding environment type data set for the environmental data collected at each time point; takes the collection time point as a matrix data label, and coordinates the environment type data sets collected at each time point in the period in a chronological order to construct a periodic wheat growth environment data matrix;

[0045] The wheat diseased ear rate analysis unit retrieves the growth environment data matrix of wheat in the current observation area, extracts the environmental data corresponding to each environmental type in the environmental data set at each time point in the matrix, and performs a periodic comprehensive analysis on each type of environmental data; based on the periodic equilibrium value of each type of environmental data in the period, analyzes the fusarium head rate of wheat observed in the period; retrieves the historical environmental data of wheat fusarium head blight, and retrieves the fusarium head rate data of wheat in the corresponding period of the same historical environmental data by referring to the periodic equilibrium value of each type of environmental data corresponding to the current period; and performs an error analysis between the retrieved historical fusarium head rate data and the fusarium head rate of the current period.

[0046] The regional anomaly index analysis module includes an image anomaly point screening unit and a regional anomaly index analysis unit;

[0047] The image anomaly point screening unit retrieves the growth image data of regional wheat within a cycle, and obtains the chromaticity data of each pixel point in the image through magnifying processing of the local image window; the chromaticity data is the RGB three-channel intensity values corresponding to the pixel points; coordinates of each pixel point in the image are obtained by establishing a coordinate system, and through constructing a data association channel, the chromaticity data of the corresponding pixel point is associated with the coordinate data of the corresponding pixel point; based on the chromaticity data corresponding to each pixel point in the coordinate system, the corresponding chromaticity feature vector is constructed, and screening analysis is carried out by combining the feature vectors of each pixel point with the standard chromaticity feature vectors of historical healthy wheat; the health parameters are set to judge the anomalies of each pixel point; the abnormal pixel points are marked explicitly, and the normal pixel points are screened out;

[0048] The regional anomaly index analysis unit conducts equal-area division based on the wheat regional coordinate image data after pixel point screening processing, and assigns labels to each sub-region; by analyzing the similarity between the anomaly feature points in each sub-region and the chromaticity feature vectors of the standard lesion abnormal pixel points of historical scab-infected wheat; combining the similarity analysis data of the chromaticity feature vectors corresponding to the abnormal pixel points in each sub-region with the standard chromaticity feature vectors of historical healthy wheat and the chromaticity feature vectors of the standard lesion abnormal pixel points of historical scab-infected wheat respectively, the scab anomaly index of the wheat in the sub-region is analyzed; by introducing a comparison index, it is judged whether there is a risk of scab occurrence in the wheat in each sub-region; if there are multiple sub-regions with a risk of scab occurrence, the monitoring priorities are assigned according to the scab anomaly index of each sub-region from large to small.

[0049] The prediction output module includes a data visualization output unit and a warning monitoring feedback unit;

[0050] The data visualization output unit visually outputs the analysis data of the diseased ear rate, the analysis data of the regional image pixel points, and the regional division data of the currently observed wheat growth area;

[0051] The warning monitoring feedback unit gives a warning feedback to the sub-regions with a risk of scab occurrence, and conducts monitoring according to the priorities of each sub-region.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] The present invention combines the collection of wheat growth environment data and image data, processes the corresponding environment data to obtain the scab ear rate of the wheat growth cycle, and analyzes the influence ratio in combination with historical data; screens and divides abnormal points in the collection for analysis, and analyzes the scab abnormal index of sub-regions in combination with the analysis of the influence ratio of the scab ear rate; predicts or locates the occurrence of scab in regional wheat accurately by combining the analysis; the present invention greatly improves the prediction accuracy of the occurrence of wheat scab, and can accurately locate the occurrence area of scab in a large range of wheat, improving the current situation of the lag and inaccurate range of traditional scab prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic structural diagram of a digital management system for a wheat scab prediction model of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0056] Embodiment: As Figure 1 shown, the present invention provides a technical solution:

[0057] A digital management method for a wheat scab prediction model, the method includes the following steps:

[0058] S100. Collect the meteorological environment data of wheat growth in the observation area during the period through a meteorological collection device, and synchronously collect the wheat images in the observation area during the period by using an image collection device;

[0059] S200. Based on the collected environment data, construct a wheat growth environment data matrix for the corresponding period with the corresponding collection time point as the data label; analyze the scab ear rate of the current period wheat environment data by combining the historical wheat historical lesion environment data;

[0060] S300. Locate abnormal pixel points by combining the collected images with the chromaticity information of pixel points in historical wheat images; comprehensively divide the wheat area according to the abnormal pixel points of the current wheat image, and combine the divided sub-regions to analyze the lesion characteristics, and predict the lesion sub-regions;

[0061] S400. Visualize and output the analysis of the scab ear rate and the regional division information of the current wheat growth area, and give a warning to the predicted lesion area.

[0062] The specific steps of S100 for collecting the meteorological environment data of wheat growth in the observation area through meteorological collection equipment and synchronously collecting wheat images in the observation area through image collection equipment are as follows:

[0063] S101. Collect the real-time growth environment data of wheat through the meteorological equipment of the wheat growth area device; obtain the growth environment data of wheat in the corresponding period by setting the collection period; transmit the collected data to the analysis end through regional networking;

[0064] S102. Conduct full-area image collection of the wheat growth area through image collection equipment, and remotely transmit the collected image data to the analysis end through airborne communication equipment.

[0065] The specific steps of S200 for constructing a wheat growth environment data matrix in the corresponding period with the corresponding collection time point as the data label based on the collected environment data and analyzing the diseased ear rate of the current period of wheat environment data by combining historical wheat historical lesion environment data are as follows:

[0066] S201. Retrieve the wheat growth environment data collected in the period, classify and process the corresponding collection environment data types by determining the collection time point; construct a corresponding environmental type data set for the environmental data collected at each time point; take the collection time point as the matrix data label, and overall plan the environmental type data sets at each time point collected in the period in chronological order to construct a period wheat growth environment data matrix;

[0067] S202. Retrieve the wheat growth environment data matrix of the current observation area, extract the environmental data corresponding to each environmental type in the environmental data sets at each time point in the matrix, and conduct a comprehensive analysis of each type of environmental data in the period. The calculation formula is

[0068]

[0069] where PVE(n:T) is the period equilibrium value of the environmental data with type number n in the corresponding period T; M(t∈T) is the number of time points t in the period T; t∈T is the time point t in the period T; H(n:t) is the environmental data value of type number n at each time point t in the corresponding period T;

[0070] Based on the period equilibrium values of each type of environmental data in the period, analyze the scab diseased ear rate of the observed wheat in the period. The calculation formula is

[0071]

[0072] Where YDP(T) is the scab ear rate of wheat observed within period T; Fn is the scab disease impact coefficient of environmental data corresponding to type number n; by retrieving historical environmental data of wheat scab, and by taking the period equilibrium value of the corresponding environmental data of each type in the current period as a reference, retrieving the scab ear rate data of wheat in the corresponding periods of the same historical environmental data; by performing error analysis on the retrieved historical scab ear rate data and the scab ear rate in the current period, if |YDP(T) - YDP(AL)| > E, then re-collect the current observed wheat environmental data and re-perform the above environmental data analysis steps; where YDP(AL) is the average value of the historical scab ear rates of wheat under the same environmental data in the current period; E is the error control parameter;

[0073] If |YDP(T) - YDP(AL)| ≤ E, then analyze the comprehensive impact ratio of the diseased ear rate of the currently observed wheat, and its calculation formula is

[0074]

[0075] Where, POI(T) is the comprehensive impact ratio of the diseased ear rate of the currently observed wheat within the current period; YDP(Amax) and YDP(Amin) are respectively the maximum and minimum values of the historical scab ear rate data of wheat under the same environmental data in the current period.

[0076] The step S300 performs abnormal pixel point localization processing on the collected image by combining the chromaticity information of pixel points in the historical wheat image; comprehensively divides the wheat area based on the abnormal pixel points in the current wheat image, and combines the divided sub-areas to perform lesion feature analysis. The specific steps for predicting the diseased sub-areas are as follows:

[0077] S301, retrieve the growth image data of wheat in the area within the period, and obtain the chromaticity data of each pixel point in the image through image local window magnification processing; the chromaticity data is the RGB three-channel intensity values corresponding to the pixel points; obtain the coordinates of each pixel point in the image by establishing a coordinate system, and through constructing a data association channel, associate the chromaticity data of the corresponding pixel point with the coordinate data of the corresponding pixel point; construct the corresponding chromaticity feature vector based on the chromaticity data of each pixel point in the coordinate system, and perform screening analysis by combining the feature vectors of each pixel point with the standard chromaticity feature vector of historical healthy wheat, and its calculation formula is

[0078]

[0079] Where, D[(x,y):HE] refers to the similarity between the chromaticity feature vector of the pixel point with coordinates (x,y) and the standard chromaticity feature vector of historical healthy wheat; is the chromaticity feature vector of the pixel point with coordinates (x,y); is the standard chromaticity feature vector of historical healthy wheat; and are respectively the norms of the chromaticity feature vectors of the pixel points corresponding to the coordinates (x, y) and the standard chromaticity feature vector of historical healthy wheat; the chromaticity feature vector of the corresponding coordinates is where I R , I G , I B are respectively the chromaticity intensity values of the R, G, and B corresponding channels of the coordinate pixel points; by setting the health parameter Dp, if D[(x,y):HE] ≤ Dp, it is determined that the pixel point corresponding to the current coordinate is a normal pixel point; if D[(x,y):HE] > Dp, it is determined that the pixel point corresponding to the current coordinate is an abnormal pixel point; the abnormal pixel points are marked explicitly, and the normal pixel points are screened out;

[0080] S302. Based on the wheat area coordinate image data after pixel point screening processing, equal area division is performed, and labels are assigned to each sub-region; by analyzing the similarity between the abnormal feature points in each sub-region and the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat, the calculation formula is

[0081]

[0082] where D[(x,y) j :SC] refers to the similarity between the chromaticity feature vector of the pixel point corresponding to the coordinates (x, y) in the corresponding sub-region number j and the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat; is the chromaticity feature vector of the pixel point corresponding to the coordinates (x, y) in the corresponding sub-region number j, is the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat, is the norm of the chromaticity feature vector of the pixel point corresponding to the coordinates (x, y) in the corresponding sub-region number j, is the norm of the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat;

[0083] Combining the similarity analysis data of the chromaticity feature vectors corresponding to the abnormal pixel points in each sub-region with the standard chromaticity feature vector of historical healthy wheat and the chromaticity feature vector of the standard lesion abnormal pixel points of historical Fusarium head blight wheat, the Fusarium head blight abnormal index of wheat in the sub-region is analyzed, and the calculation formula is

[0084]

[0085] where GI j is the Fusarium head blight abnormal index of wheat in the corresponding sub-region number j; f is the activation function; U j is the number of abnormal points in the sub-region numbered j; Ua is the number of abnormal points in the whole area of the wheat image; D[(x,y) j :HE] is the similarity between the chromaticity feature vector of the pixel point with coordinates (x,y) in the corresponding sub-region number j and the standard chromaticity feature vector of historical healthy wheat; b is the bias parameter; (x,y)∈j are the coordinates of each pixel point in the sub-region with the corresponding number j; By introducing the contrast index GI(th), if GI j <GI(th), it is determined that there is no risk of Fusarium head blight in the wheat in the current corresponding sub-region number j; if GI j ≥GI(th), it is determined that there is a risk of Fusarium head blight in the wheat in the current corresponding sub-region number j; if there are multiple sub-regions with a risk of Fusarium head blight, the monitoring priorities are assigned according to the descending order of the corresponding Fusarium head blight abnormal indices of each sub-region.

[0086] The S400 visually outputs the analysis of the diseased ear rate and the regional division information of the current wheat growth area, and the specific steps of warning the predicted diseased area are as follows:

[0087] S401. Visually output the analysis data of the diseased ear rate, the analysis data of the pixel points of the regional image, and the regional division data of the current observed wheat growth area;

[0088] S402. Give a warning feedback to the sub-regions with a risk of Fusarium head blight and monitor according to the priorities of each sub-region.

[0089] A digital management system for a wheat Fusarium head blight prediction model, the system includes a regional data acquisition module, a periodic diseased ear rate analysis module, a regional abnormal index analysis module, and a prediction output module;

[0090] The regional data acquisition module collects the meteorological environment data of wheat growth in the observation area during the period through meteorological acquisition equipment, and simultaneously collects the wheat images in the observation area during the period by using image acquisition equipment; the periodic diseased ear rate analysis module constructs a data matrix of wheat growth environment data in the corresponding period with the corresponding acquisition time point as the data label based on the collected environment data; analyzes the diseased ear rate of the current period wheat environment data by combining the historical wheat historical lesion environment data; the regional abnormal index analysis module locates the abnormal pixel points by combining the collected image with the chromaticity information of the pixel points of the historical wheat image; comprehensively divides the wheat area according to the abnormal pixel points of the current wheat image, and combines the divided sub-regions to analyze the lesion characteristics and predict the diseased sub-regions; the prediction output module visually outputs the analysis of the diseased ear rate and the regional division information of the current wheat growth area, and warns the predicted diseased area.

[0091] The regional data acquisition module includes a regional environment data acquisition unit and a wheat image data acquisition unit;

[0092] The regional environmental data acquisition unit collects real-time wheat growth environment data through the meteorological equipment of the wheat growth area device; obtains the wheat growth environment data within the corresponding period by setting the acquisition cycle; and transmits the collected data to the analysis end by using the regional networking;

[0093] The wheat image data acquisition unit acquires images of the entire wheat growing area through an image acquisition device, and remotely transmits the acquired image data to an analysis end through an airborne communication device.

[0094] The periodic diseased ear rate analysis module includes an environmental data processing unit and a wheat diseased ear rate analysis unit;

[0095] The environmental data processing unit retrieves the wheat growth environment data collected periodically, and classifies the collected environment data according to the corresponding collected environment data type by determining the collection time point; constructs a corresponding environment type data set for the environmental data collected at each time point; takes the collection time point as a matrix data label, and coordinates the environment type data sets collected at each time point in the period in time order to construct a periodic wheat growth environment data matrix;

[0096] The wheat diseased ear rate analysis unit retrieves the growth environment data matrix of wheat in the current observation area, extracts the environmental data corresponding to each environmental type in the environmental data set at each time point in the matrix, and performs a periodic comprehensive analysis on each type of environmental data; based on the periodic equilibrium value of each type of environmental data in the period, analyzes the fusarium head rate of wheat observed in the period; retrieves the historical environmental data of wheat fusarium head blight, and retrieves the fusarium head rate data of wheat in the corresponding period of the same historical environmental data by referring to the periodic equilibrium value of each type of environmental data corresponding to the current period; and performs an error analysis between the retrieved historical fusarium head rate data and the fusarium head rate of the current period.

[0097] The regional anomaly index analysis module includes an image anomaly point screening unit and a regional anomaly index analysis unit;

[0098] The image abnormal point screening unit retrieves the growth image data of regional wheat in the cycle, and obtains the chromaticity data of each pixel in the image by magnifying the local window of the image; the chromaticity data is the RGB three-channel intensity value corresponding to the pixel; the coordinates of each pixel in the image are obtained by establishing a coordinate system, and the chromaticity data of the corresponding pixel is associated with the coordinate data of the corresponding pixel by constructing a data association channel; the corresponding chromaticity feature vector is constructed based on the chromaticity data corresponding to each pixel in the coordinate system, and the feature vector of each pixel is combined with the standard chromaticity feature vector of historical healthy wheat for screening and analysis; each pixel is judged to be abnormal by setting health parameters; the abnormal pixels are explicitly marked, and the normal pixels are screened out;

[0099] The regional abnormality index analysis unit divides the wheat region coordinate image data into equal regions based on the pixel point screening process, and assigns labels to each sub-region; analyzes the similarity between the abnormal feature points in each sub-region and the chromaticity feature vectors of the standard lesion abnormal pixel points of the wheat with historical ergot; analyzes the ergot abnormality index of the wheat in the sub-region based on the similarity analysis data of the chromaticity feature vectors corresponding to the abnormal pixel points in each sub-region with the standard chromaticity feature vectors of the historical healthy wheat and the chromaticity feature vectors of the standard lesion abnormal pixel points of the wheat with historical ergot; judges whether the wheat in each sub-region has the risk of ergot by introducing a comparison index; if there are multiple sub-regions with the risk of ergot, the monitoring priority is allocated from large to small according to the ergot abnormality index corresponding to each sub-region.

[0100] The prediction output module includes a data visualization output unit and a warning monitoring feedback unit;

[0101] The data visualization output unit visualizes and outputs the diseased ear rate analysis data of the currently observed wheat growth area, the regional image pixel point analysis data, and the regional division data;

[0102] The warning monitoring feedback unit provides warning feedback to sub-areas with the risk of fusarium smut, and monitors each sub-area according to its priority;

[0103] In the example:

[0104] A certain wheat planting base needs to monitor the wheat growing area for fusarium head blight, and is equipped with the digital management system for the wheat fusarium head blight prediction model of the present invention; the base collects the real-time growth environment data of wheat through the meteorological equipment of the wheat growing area device; obtains the growth environment data of wheat in the corresponding period by setting the collection cycle; transmits the collected data to the analysis end by using the regional networking; collects the whole area image of the wheat growing area by the image collection device, and transmits the collected image data to the analysis end by remote data transmission through the airborne communication equipment;

[0105] Retrieve the wheat growth environment data collected periodically, and classify the collected environment data types by determining the collection time point; construct the corresponding environment type data set for the environmental data collected at each time point; use the collection time point as the matrix data label, coordinate the environment type data sets collected at each time point in the period in chronological order, and construct a periodic wheat growth environment data matrix; retrieve the wheat growth environment data matrix in the current observation area, extract the environmental data corresponding to each environment type in the environmental data set at each time point in the matrix, and conduct a periodic comprehensive analysis of each type of environmental data. The calculation formula is:

[0106]

[0107] Based on the periodic equilibrium value of various types of environmental data within a period, analyze the scab ear rate of the observed wheat within the period. The calculation formula is

[0108]

[0109] By retrieving the historical environmental data of wheat scab, with reference to the periodic equilibrium value of the corresponding various types of environmental data in the current period, retrieve the scab ear rate data of wheat in the corresponding periods of the same historical environmental data; by performing error analysis on the retrieved historical scab ear rate data and the scab ear rate in the current period, if |YDP(T) - YDP(AL)| > E, then re-collect the environmental data of the currently observed wheat and re-perform the above environmental data analysis steps;

[0110] If |YDP(T) - YDP(AL)| ≤ E, then analyze the comprehensive influence ratio of the disease ear rate of the currently observed wheat. The calculation formula is

[0111]

[0112] Retrieve the growth image data of regional wheat within a period, and obtain the chromaticity data of each pixel point in the image through image local window magnification processing; the chromaticity data is the RGB three-channel intensity value corresponding to the pixel point; obtain the coordinates of each pixel point in the image by establishing a coordinate system, and through constructing a data association channel, associate the chromaticity data of the corresponding pixel point with the coordinate data of the corresponding pixel point; construct the corresponding chromaticity feature vector based on the chromaticity data of each pixel point in the coordinate system, and perform screening analysis in combination with the feature vector of each pixel point and the standard chromaticity feature vector of historical healthy wheat. The calculation formula is

[0113]

[0114] The chromaticity feature vector corresponding to the coordinate is By setting the health parameter Dp, if D[(x,y):HE] ≤ Dp, then determine that the pixel point corresponding to the current coordinate is a normal pixel point; if D[(x,y):HE] > Dp, then determine that the pixel point corresponding to the current coordinate is an abnormal pixel point; prominently mark the abnormal pixel points and screen out the normal pixel points;

[0115] Based on the wheat area coordinate image data after pixel point screening processing, perform equal area division and assign labels to each sub-region; by analyzing the similarity between the abnormal feature points in each sub-region and the chromaticity feature vector of the standard lesion abnormal pixel points of historical scabbed wheat. The calculation formula is

[0116]

[0117] Combined with the similarity analysis data of the chromaticity feature vectors corresponding to the abnormal pixel points in each sub-region respectively with the standard chromaticity feature vectors of historical healthy wheat and the chromaticity feature vectors of the standard lesion abnormal pixel points of historical scabbed wheat, the scab abnormal index of wheat in the sub-region is analyzed, and its calculation formula is

[0118]

[0119] By introducing the comparison index GI(th), if GI j < GI(th), it is determined that there is no risk of scab occurrence in the wheat in the current sub-region with the corresponding number j; if GI j ≥ GI(th), it is determined that there is a risk of scab occurrence in the wheat in the current sub-region with the corresponding number j; if there are multiple sub-regions with a risk of scab occurrence, the monitoring priorities are allocated according to the scab abnormal indices of each sub-region from large to small;

[0120] Visualize the analysis data of the diseased ear rate, the analysis data of the regional image pixel points, and the regional division data of the currently observed wheat growth area; give a warning feedback to the sub-regions with a risk of scab occurrence, and monitor according to the priorities of each sub-region.

[0121] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present invention. Any reference signs in the claims should not be regarded as limiting the claimed rights.

Claims

1. A digital management method for a wheat scab prediction model, characterized in that: The method comprises the following steps: S100, collecting meteorological environment data of wheat growth in the observation area period by using meteorological collection equipment, and simultaneously collecting wheat images in the observation area period by using image collection equipment; S200, based on the collected environmental data, construct a wheat growth environment data matrix in the corresponding period with the corresponding collection time point as the data label; analyze the diseased ear rate of wheat environmental data in the current period by combining historical wheat disease environment data; S300, performing abnormal pixel location processing by combining the collected image with the pixel chromaticity information of the wheat historical image; dividing the wheat area based on the abnormal pixel points of the current wheat image, and performing pathological feature analysis based on the divided sub-areas to predict the pathological sub-areas; The specific steps of S300 are as follows: S301, retrieve the growth image data of wheat in the region, and obtain the chromaticity data of each pixel in the image by magnifying the local window of the image; the chromaticity data is the RGB three-channel intensity value corresponding to the pixel; obtain the coordinates of each pixel in the image by establishing a coordinate system, and associate the chromaticity data of the corresponding pixel with the coordinate data of the corresponding pixel by constructing a data association channel; construct the corresponding chromaticity feature vector based on the chromaticity data corresponding to each pixel in the coordinate system, and screen and analyze the feature vector of each pixel with the standard chromaticity feature vector of historical healthy wheat, and the calculation formula is: Among them, D[(x,y):HE] refers to the similarity between the chromaticity feature vector of the pixel point with coordinates (x,y) and the standard chromaticity feature vector of historical healthy wheat; is the chromaticity feature vector of the pixel corresponding to the coordinate (x, y); is the standard chromaticity feature vector of historical healthy wheat; and The chromaticity feature vector of the pixel corresponding to the coordinate (x, y) is the modulus of the standard chromaticity feature vector of historical healthy wheat; the chromaticity feature vector of the corresponding coordinate is Among them I R ,I G ,I B The chromaticity intensity values ​​of the R, G, and B channels corresponding to the coordinate pixel points respectively; by setting the health parameter Dp, if D[(x, y):HE]≤Dp, the pixel point corresponding to the current coordinate is judged to be a normal pixel point; if D[(x, y):HE]>Dp, the pixel point corresponding to the current coordinate is judged to be an abnormal pixel point; the abnormal pixel points are explicitly marked, and the normal pixel points are screened out; S302, based on the wheat area coordinate image data after pixel point screening, equal area division is performed, and labels are assigned to each sub-area; by analyzing the similarity of the chromaticity feature vectors of the abnormal feature points in each sub-area with the standard lesion abnormal pixel points of the historical fusarium rust wheat, the calculation formula is: Where D[(x,y) j :SC] refers to the similarity between the chromaticity feature vector of the pixel point with coordinate (x, y) in the corresponding sub-region number j and the chromaticity feature vector of the standard abnormal lesion pixel point of the historical scab disease wheat; is the chromaticity feature vector of the pixel with coordinates (x, y) in the corresponding sub-region number j, is the chromaticity feature vector of the standard abnormal lesion pixel of historical fusarium scab wheat, is the modulus of the chromaticity feature vector of the pixel with coordinate (x, y) in the corresponding sub-region number j, is the modulus of the chromaticity feature vector of the standard abnormal lesion pixel of the historical fusarium head blight wheat; Combining the similarity analysis data of the chromaticity feature vectors corresponding to the abnormal pixels in each sub-region with the standard chromaticity feature vectors of historical healthy wheat and the chromaticity feature vectors of the standard abnormal pixels of lesions in historical fusarium rust wheat, the fusarium rust abnormality index of wheat in the sub-region was analyzed, and the calculation formula was: Among them, GI j is the abnormal index of wheat fusarium wilt in the corresponding sub-region numbered j; f is the activation function; U j is the number of outliers in the sub-region numbered j; U a is the number of abnormal points in the whole area of ​​the wheat image; D[(x,y) j :HE] is the similarity between the chromaticity feature vector of the pixel with coordinate (x, y) in the corresponding sub-region number j and the standard chromaticity feature vector of historical healthy wheat; b is the bias parameter; (x, y)∈j is the coordinate of each pixel in the sub-region number j; by introducing the contrast index GI(th), if GI j <GI(th), then it is judged that there is no risk of fusarium fusarium in the current sub-region corresponding to number j; if GI j ≥GI(th), it is judged that the wheat in the current sub-region corresponding to number j is at risk of fusarium wilt; if there are multiple sub-regions with fusarium wilt risk, the monitoring priority is allocated from large to small according to the corresponding fusarium wilt abnormality index of each sub-region; POI(T) is the comprehensive impact ratio of the diseased ear rate of wheat observed in the current cycle; S400: Visually output the diseased ear rate analysis and regional division information of the current wheat growing area, and issue a warning for the predicted diseased area.

2. A digital management method for wheat scab prediction model according to claim 1, characterized in that: The specific steps of S100 for collecting meteorological environmental data of wheat growth in the observation area period by using meteorological collection equipment and synchronously collecting wheat images in the observation area period by using image collection equipment are as follows: S101, collecting real-time growth environment data of wheat through meteorological equipment of the wheat growth area device; obtaining the growth environment data of wheat in the corresponding period by setting a collection period; and transmitting the collected data to the analysis end by using regional networking; S102, collecting images of the entire wheat growing area through an image collection device, and remotely transmitting the collected image data to an analysis terminal through an airborne communication device.

3. A digital management method for wheat scab prediction model according to claim 2, characterized in that: The S200 constructs a wheat growth environment data matrix in a corresponding period based on the collected environmental data with the corresponding collection time point as the data label; the specific steps of analyzing the diseased ear rate of wheat environmental data in the current period by combining the historical wheat disease environment data are as follows: S201, retrieve the wheat growth environment data collected periodically, and classify the data according to the corresponding collected environment data type by determining the collection time point; and construct a corresponding environment type data set for the environment data collected at each time point; Taking the collection time point as the matrix data label, the environmental type data sets collected at each time point in the cycle are coordinated in chronological order to construct the periodic wheat growth environment data matrix; S202, retrieve the wheat growth environment data matrix of the current observation area, extract the environmental data corresponding to each environmental type in the environmental data set at each time point in the matrix, and perform periodic comprehensive analysis on each type of environmental data. The calculation formula is: Wherein, PVE(n:T) is the period equilibrium value of the environment data with type number n in the corresponding period T; M(t∈T) is the number of time points t in the period T; t∈T is the time point t in the period T; H(n:t) is the value of the environment data with type number n at each time point t in the corresponding period T; Based on the periodic equilibrium value of various types of environmental data within the period, the fusarium head blight rate of wheat observed within the period was analyzed, and the calculation formula is: Wherein, YDP(T) is the fusarium head rate of wheat observed in period T; Fn is the fusarium head rate influence coefficient of the environmental data corresponding to type number n; by retrieving the historical environmental data of wheat fusarium head rate, and taking the period equilibrium value of the corresponding types of environmental data of the current period as a reference, the fusarium head rate data of wheat in the corresponding period with the same historical environmental data are retrieved; by performing error analysis on the retrieved historical fusarium head rate data and the fusarium head rate of the current period, if |YDP(T)-YDP(AL)|>E, the current observed wheat environmental data are recollected and the above environmental data analysis steps are performed again; wherein YDP(AL) is the mean value of the historical fusarium head rate under the same environmental data of wheat in the current period; E is the error control parameter; If |YDP(T)-YDP(AL)|≤E, the comprehensive impact ratio of the diseased ear rate of the currently observed wheat is analyzed, and the calculation formula is: Among them, POI(T) is the comprehensive impact ratio of the diseased ear rate of wheat observed in the current cycle; YDP(Amax) and YDP(Amin) are the maximum and minimum values ​​of the historical fusarium head blight rate data under the same environmental data of wheat in the current cycle, respectively.

4. A digital management method for wheat scab prediction model according to claim 3, characterized in that: The specific steps of S400 for visually outputting the diseased ear rate analysis and regional division information of the current wheat growing area and warning the predicted diseased area are as follows: S401, visually output the diseased ear rate analysis data of the currently observed wheat growth area, the regional image pixel point analysis data, and the regional division data; S402. Provide warning feedback to sub-areas with risk of ergot disease and conduct monitoring based on the priority of each sub-area.

5. A digital management system for a wheat fusarium head blight prediction model, using a digital management method for a wheat fusarium head blight prediction model as claimed in any one of claims 1 to 4, characterized in that: The system includes a regional data acquisition module, a periodic diseased ear rate analysis module, a regional abnormal index analysis module and a prediction output module; The regional data acquisition module collects meteorological environment data of wheat growth within the observation area period through meteorological acquisition equipment, and simultaneously uses image acquisition equipment to collect wheat images within the observation area period; the periodic diseased ear rate analysis module constructs a wheat growth environment data matrix within the corresponding period based on the collected environment data, with the corresponding collection time point as the data label; the diseased ear rate of wheat environment data in the current period is analyzed by combining historical wheat pathological environment data; the regional abnormal index analysis module performs abnormal pixel location processing by combining the collected image with the chromaticity information of the pixel points of the historical wheat image; the wheat area is divided according to the abnormal pixels of the current wheat image, and the pathological characteristics are analyzed in combination with the divided sub-areas, and the pathological sub-areas are predicted; the prediction output module visualizes the diseased ear rate analysis and regional division information of the current wheat growth area, and warns the predicted pathological area.

6. A digital management system for wheat scab prediction model according to claim 5, characterized in that: The regional data acquisition module includes a regional environment data acquisition unit and a wheat image data acquisition unit; The regional environmental data acquisition unit collects real-time wheat growth environment data through the meteorological equipment of the wheat growth area device; obtains the wheat growth environment data within the corresponding period by setting the acquisition cycle; and transmits the collected data to the analysis end by using the regional networking; The wheat image data acquisition unit acquires images of the entire wheat growing area through an image acquisition device, and remotely transmits the acquired image data to an analysis end through an airborne communication device.

7. A digital management system for wheat scab prediction model according to claim 6, characterized in that: The periodic diseased ear rate analysis module includes an environmental data processing unit and a wheat diseased ear rate analysis unit; The environmental data processing unit retrieves the periodically collected wheat growth environmental data, and classifies and processes the data according to the collected environmental data type by determining the collection time point; Construct corresponding environmental type data sets for environmental data collected at each time point; Taking the collection time point as the matrix data label, the environmental type data sets collected at each time point in the cycle are coordinated in chronological order to construct the periodic wheat growth environment data matrix; The wheat diseased ear rate analysis unit retrieves the growth environment data matrix of wheat in the current observation area, extracts the environmental data corresponding to each environmental type in the environmental data set at each time point in the matrix, and performs a periodic comprehensive analysis on each type of environmental data; based on the periodic equilibrium value of each type of environmental data in the period, analyzes the fusarium head rate of wheat observed in the period; retrieves the historical environmental data of wheat fusarium head blight, and retrieves the fusarium head rate data of wheat in the corresponding period of the same historical environmental data by referring to the periodic equilibrium value of each type of environmental data corresponding to the current period; and performs an error analysis between the retrieved historical fusarium head rate data and the fusarium head rate of the current period.

8. A digital management system for wheat scab prediction model according to claim 7, characterized in that: The regional anomaly index analysis module includes an image anomaly point screening unit and a regional anomaly index analysis unit; The image abnormal point screening unit retrieves the growth image data of wheat in the region within the cycle, and obtains the chromaticity data of each pixel in the image by magnifying the local window of the image; The chromaticity data is the RGB three-channel intensity value corresponding to the pixel point; the coordinates of each pixel point in the image are obtained by establishing a coordinate system, and the chromaticity data of the corresponding pixel point is associated with the coordinate data of the corresponding pixel point by constructing a data association channel; the corresponding chromaticity feature vector is constructed based on the chromaticity data corresponding to each pixel point in the coordinate system, and the feature vector of each pixel point is combined with the standard chromaticity feature vector of historical healthy wheat for screening and analysis; each pixel point is judged to be abnormal by setting health parameters; the abnormal pixel points are explicitly marked, and the normal pixel points are screened out; The regional abnormality index analysis unit divides the wheat region coordinate image data into equal regions based on the pixel point screening process, and assigns labels to each sub-region; analyzes the similarity between the abnormal feature points in each sub-region and the chromaticity feature vectors of the standard abnormal lesion pixel points of the wheat with historical fusarium rust; combines the similarity analysis data of the chromaticity feature vectors corresponding to the abnormal pixels in each sub-region with the standard chromaticity feature vectors of the healthy wheat in history and the chromaticity feature vectors of the standard abnormal lesion pixel points of the wheat with historical fusarium rust, and analyzes the abnormal fusarium rust index of the wheat in the sub-region; and introduces a comparison index to determine whether the wheat in each sub-region has a risk of fusarium rust; If there are multiple sub-regions with the risk of ergot disease, monitoring priorities will be allocated from large to small based on the corresponding ergot disease abnormality index of each sub-region.

9. A digital management system for wheat scab prediction model according to claim 8, characterized in that: The prediction output module includes a data visualization output unit and a warning monitoring feedback unit; The data visualization output unit visualizes and outputs the diseased ear rate analysis data of the currently observed wheat growth area, the regional image pixel point analysis data, and the regional division data; The warning monitoring feedback unit provides warning feedback to sub-areas where there is a risk of ergot disease occurring, and monitors each sub-area according to its priority.

Images