KR-102961788-B1 - INTEGRATED CROP DIAGNOSIS SYSTEM BASED ON MULTISPECTRAL VEGETATION INDEX THAT CAN REPLACE HYPERSPECTRAL

KR102961788B1KR 102961788 B1KR102961788 B1KR 102961788B1KR-102961788-B1

Abstract

The crop integrated diagnosis system based on a multispectral vegetation index capable of replacing hyperspectral imaging according to the embodiment captures crops using a multispectral camera to acquire spectral data, generates matched spectral data after image alignment, and calculates a vegetation index by image mapping recognized object data from the spectral image of the matched spectral data and the color image of the matched spectral data. Subsequently, physiological disorders, pests and diseases, and internal quality of the crops are diagnosed through a classification model using the calculated vegetation index, and solutions based on the diagnosis results are provided through an expert system model.

Inventors

이영호
이옥정
류종훈

Assignees

아이티컨버젼스 주식회사

Dates

Publication Date: 20260511
Application Date: 20240229

Claims (11)

In a multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging through a server and client terminal, The above server includes a learning server and a service server, and The above learning server collects multispectral data (102) and hyperspectral data (111), extracts color data (103) from the multispectral data (102), and uses the color data (103) as training data to generate a recognition model (107) that recognizes an analysis target, which is a crop object including leaves and stems in a crop image. A classification model (112) for diagnosing physiological disorders, pests and diseases, and internal quality is generated by using color data matching multispectral data (102) and hyperspectral data (111) as training data, and An expert system model (115) is created to provide solutions for physiological disorders, pests and diseases and internal quality of crops through data distributed by an agricultural institution, and The above service server provides the recognition model (107), classification model (112), and expert system model (115) to the client terminal in response to a request from the client terminal. The above client terminal is, The image is aligned using spectral data (101) of an image captured through a multispectral camera, and the aligned image is processed to generate matching spectral data. The type of image is determined from the matching spectral data (103). If the determined type of image is a color image, an object is recognized and first object data is generated. An additional object is recognized through a recognition model to generate second object data, which is the recognized additional object. The spectral image of the second object data is mapped to the matching spectral data. Calculating a vegetation index from the above mapping results, and determining physiological disorders, pests and diseases, and internal quality of crops using a classification model based on the vegetation index, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 1, the above learning server Color data is extracted by combining spectral data including blue, green, and red from matched spectral data, and Using the above extracted color data as training data for a recognition model that performs crop object recognition, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 2, the above learning server Train a recognition model with training data including color data, and The above recognition model is A multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging that recognizes crop objects in color images.
In paragraph 1, the above learning server A matching process for aligning pixel coordinate systems between multiple aligned spectral band images, performing feature point extraction, feature point matching, and geometric transformation processes, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In Article 1, The above recognition model (107) recognizes an analysis target, which is a crop object including leaves and stems in a crop image, through color data, and The above classification model (112) diagnoses physiological disorders, pests and diseases, and internal quality of crops through vegetation indices calculated from the result of image mapping of object data and spectral images, and The above expert system model (115) is characterized by providing solutions for physiological disorders, pests and diseases, and internal quality of crops. Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 1, the vegetation index is Containing at least one of NDVI (Normalized Difference Vegetation Index), NDRE (Normalized Difference Red Edge), GNDVI (Green Normalized Difference Vegetation Index), AVI (Atmospherically Resistant Vegetation Index), SAVI (Soil-Adjusted Vegetation Index), ARVI (Atmospherically Resistant Vegetation Index), and SIPI (Structure Insensitive Pigment Index). Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 1, the client terminal Calculating a pixel-by-pixel vegetation index by analyzing the data resulting from image mapping of object data and spectral images according to wavelength and spectrum, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 7, the above client terminal Color images are generated from red, green, and blue spectral images from matched spectral data, crop types are identified and object data is extracted from the generated color data using a pre-trained recognition model, vegetation index images are calculated from the matched spectral data, and the vegetation index images are mapped to the regions of interest of the objects and analyzed through a classification model. A multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral analysis, for diagnosing physiological disorders, pests and diseases, and internal quality of crops.
In paragraph 1, the above classification model is Diagnosing physiological disorders, pests and diseases, and internal quality based on the type of calculated vegetation index, the range including the calculated vegetation index, and the vegetation index value, and predicting whether physiological disorders will manifest, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 8, the above classification model is Calculate wavelength-specific reflectance and absorption rates of multispectral data, and compare the calculated reflectance and absorption rates with those corresponding to physiological disorders to diagnose physiological disorders, pests and diseases, and internal quality of crops. Generates a reflectance graph according to wavelength of multispectral data, and diagnoses physiological disorders, pests and diseases, and internal quality of crops by comparing the graph with a crop physiological disorder reference. The above crop physiological disorder reference is Includes wavelength-dependent reflectance graphs of multispectral data for each type of physiological disorder, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.
In paragraph 1, the client terminal When a physiological disorder of a crop is diagnosed through multispectral data, at least one of an environmental solution, a quality solution, a pest control solution, and a harvest time solution is generated through an expert system model according to the type of the diagnosed physiological disorder, Multispectral vegetation index-based integrated crop diagnostic system capable of replacing hyperspectral imaging.

Description

Integrated Crop Diagnosis System Based on Multispectral Vegetation Index That Can Replace Hyperspectral The present disclosure relates to an integrated crop diagnostic system based on a multispectral vegetation index that can replace hyperspectral analysis. Specifically, it relates to an integrated crop diagnostic system based on a multispectral vegetation index that can replace hyperspectral analysis, which generates a classification model by matching multispectral data with hyperspectral data and performs an integrated diagnosis of the crop's condition using a vegetation index (optical index). An optical index is an index obtained by generating spectra through various combinations, and in the agricultural field, optical indices are also referred to as vegetation indices. Physiological disorders, pests and diseases, and internal quality can be diagnosed based on the vegetation index. Unless otherwise indicated in this specification, the contents described in this section are not prior art for the claims of this application, and are not to be recognized as prior art simply because they are included in this section. Multispectral spectroscopy is a technology that detects light or electromagnetic waves over a broad range of the electromagnetic spectrum. Primarily used in the field of optics, multispectral spectroscopy is a technology that simultaneously detects a wide range of wavelengths rather than specific wavelengths or colors. Specific spectral region techniques are used to detect light of a specific wavelength. However, multispectral technology enables a more detailed understanding of an object's properties by simultaneously detecting and analyzing light of multiple wavelengths. Multispectral technology can be utilized across various wavelength ranges, including infrared spectra, visible light, ultraviolet light, and radio waves. On the other hand, while multispectral imaging can acquire a large amount of spectral information across wavelength bands, it has disadvantages such as the high cost of multispectral equipment, long imaging times, and the requirement of an imaging stage. FIG. 1 is a diagram showing an integrated diagnostic system of a crop integrated diagnostic system based on a multispectral vegetation index that can replace hyperspectral imaging according to an embodiment. FIG. 2 is a drawing showing an example of the installation of a multispectral camera. FIG. 3 is an interface for inputting an image acquired through a multispectral camera in an embodiment. Figure 4 shows an interface representing the multispectral analysis results of a recognized image. FIG. 5 is an image recognition interface for crop object recognition in an embodiment Figure 6 is a diagram showing the pixel analysis results of the band after crop object recognition. FIG. 7 is a drawing showing the crop object extraction results according to an embodiment. Figure 8 is a diagram showing the result of extracting a region of interest from a crop object. FIGS. 9 and 10 are diagrams of the output interface for the multispectral analysis results of a region of interest (ROA) according to an embodiment. FIG. 11 is an interface showing the analysis results of physiological disorders, pests and diseases, and internal quality of crops through multispectral image analysis according to an embodiment. FIG. 12 is a diagram showing the solution extraction results based on the crop diagnosis results according to an embodiment. FIG. 13 is a drawing showing a block diagram of a server according to an embodiment. FIGS. 14 and 15 are drawings for explaining the process of calculating a vegetation index according to an embodiment. FIGS. 16 and 17 are drawings for explaining an image matching process according to an embodiment. FIG. 18 is a drawing showing detailed items of a solution according to an embodiment. Figure 19 is a graph of the reflectance of hyperspectral and multispectral data. Figures 20 and 21 are diagrams comparing the analysis results of hyperspectral data and multispectral data. Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components regardless of drawing symbols are assigned the same reference number, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not inherently possess distinct meanings or roles. Furthermore, in describing embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not