KR20210069987A - Crop growth diagnosis system and method using image - Google Patents

Abstract

The present invention relates to a crop growth diagnosis system using an image, including: an image processing unit for calculating size information of a crop from a crop image; a database unit for storing the size information and the weight information of the crop; and a relationship model producing unit for calculating a relationship model by analyzing the relationship between the size information and the weight information. Accordingly, the present invention can calculate the size and weight information of the crop only from the crop image, and can diagnose and predict a current growth stage.

Description

Crop growth diagnosis system and method using video images {CROP GROWTH DIAGNOSIS SYSTEM AND METHOD USING IMAGE}

The present invention relates to a system and method for diagnosing the growth stage of a crop, and more particularly, to a system and method for diagnosing the growth of a crop using an image image.

In the process of highly industrialization, agriculture had difficulties in increasing production efficiency due to the decrease of the rural population and the aging of the population. However, recent advances in information and communication technologies (ICT) such as artificial intelligence (AI), Internet of Things (IoT), cloud computing, and big data are affecting rural areas and agriculture.

Smart farm technology refers to an intelligent farm that can properly maintain and manage the growing environment of crops and livestock remotely and automatically by integrating ICT technology with the existing agricultural industry. The smart farm creates an optimal growth environment based on data on crop growth information and environmental information, and can improve the productivity and quality of agricultural products with less labor, energy, and nutrients than before. In addition, smart farm can be applied to the entire process of the agricultural industry from production to distribution and sales, improving productivity, efficiency, and quality, while reducing labor costs, which is a technology that can create high added value. Therefore, smart farm technology is an important means for increasing agricultural productivity and securing competitiveness, and various projects are being promoted in all fields of agriculture, such as production and distribution of crops and agricultural management, using smart farm technology.

Since smart farms remotely check and control the growing environment of crops, technology for monitoring crops, measuring changes in the growth of crops, and observing the growth process to analyze the status of crops are required. In particular, in order to improve the actual productivity of crops, an automated system that can effectively collect, analyze and manage biometric information related to the entire life cycle of each farmhouse's vegetables from planting to shipment is required.

However, the existing crop image processing technology is at a rudimentary stage at home and abroad, and there is currently no system capable of predicting crop growth changes or diagnosing crop conditions by collecting measured crop data.

Accordingly, the present applicant has devised a system that can increase the productivity of a crop when applied to an actual agricultural field by collecting, analyzing, and managing biometric information according to the entire life cycle of a crop.

An object of the present invention is to provide a system and method for diagnosing crop growth using an image image, in which size information of a crop can be extracted only from an image image taken of a crop, and weight information can be grasped through the size information.

In order to achieve the above object, the present invention provides a crop growth diagnosis system using an image image, comprising: an image processing unit for calculating size information of the crop from the image image of the crop; a database unit for storing the size information and the weight information of the crop; and a relationship model generator configured to calculate a relationship model by analyzing the relationship between the size information and the weight information.

Preferably, the video image includes a length measuring means as a reference for calculating the size information of the crop, and the length measuring means may be photographed together with the crop.

Preferably, the size information includes the horizontal and vertical lengths of the crop, and the relationship model generator sets the size information as an independent variable and sets the weight information as a dependent variable to perform multiple regression analysis. A regression model can be calculated.

Preferably, the image processing unit includes: a preprocessing module for removing noise from the image image; a transform module for Fourier transforming the video image from which noise has been removed; And it may include a size calculation module for calculating the size information of the crop from the converted image image.

In addition, in order to achieve the above object, the present invention provides a method for diagnosing crop growth using an image image, (a) storing size and weight data of a plurality of crops, and multiplying the relationship between size and weight of crops with pre-stored data Regression analysis to generate a relationship model to calculate a multiple regression model; (b) a photographing step of photographing a crop to be irradiated to generate a video image; (c) an image processing step of pre-processing the video image and performing a Fourier transform; (d) a calculation step of calculating the size information of the crop from the converted video image; (e) an output step of inputting the size information into the multiple regression model to calculate weight information for the size information of the research target crop;

According to the present invention, since the image image is obtained by photographing a crop using a general camera, shooting and image acquisition are easy, and information on the size and weight of the crop is obtained only by inputting the image image into a pre-designed model and system. It is possible to easily and accurately predict information according to the growth stage of crops. In addition, it is possible to diagnose and predict the current growth stage of the crop taken through the relationship model calculated by analyzing the relationship between the size and weight of the crop. It has the advantage of being able to produce crops.

1 shows a configuration diagram of a crop growth diagnosis system according to an embodiment of the present invention.
2 is an experimental example of the present invention, showing a response surface for a relationship model calculated by statistical analysis of the size and weight of strawberries.
3 shows a process in which a method for diagnosing crop growth according to an embodiment of the present invention is performed.

Hereinafter, the present invention will be described in detail with reference to the contents described in the accompanying drawings. However, the present invention is not limited or limited by the exemplary embodiments. The same reference numerals provided in the respective drawings indicate members that perform substantially the same functions.

Objects and effects of the present invention can be naturally understood or made clearer by the following description, and the objects and effects of the present invention are not limited only by the following description. In addition, in describing the present invention, when it is determined that a detailed description of a known technology related to the present invention may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted.

1 shows a configuration diagram of a crop growth diagnosis system 1 according to an embodiment of the present invention.

The crop growth diagnosis system 1 can calculate the size of a crop by using an image image of the crop, and by statistically analyzing the calculated size information and weight information of the crop, it is possible to generate an analysis model about the relationship between the amount information. can It is a system that can diagnose the condition according to the growth stage by extracting size and weight information from only the image of the crop through the analysis model. Since the type of crop that can be the target of the crop growth diagnosis system 1 is not limited, any type of crop may be a target if the size and weight can be measured.

In an embodiment of the present invention, strawberry was selected as a diagnostic target, and the experiment was conducted on the most popular strawberry fruit tree.

The crop growth diagnosis system 1 may include an image processing unit 11 , a database unit 13 , and a relationship model generation unit 15 . This configuration is defined for convenience of description and may be implemented without being physically separated. In addition, it may include a photographing means (3), a length measuring means (5), an output unit (17) to diagnose the growth of crops.

The photographing means 3 is a configuration for acquiring an image image of a crop to be diagnosed, and the type is not limited as long as an image capable of measuring the size of the crop can be acquired. The photographing means 3 may photograph the length measuring means 5 and the crop together to generate a video image including both components.

The length measuring means 5 is a standard for calculating crop size information, and the unit length can be set by the user. Since the length measuring means 5 has a role of a comparison target for measuring the size of a crop, it may be a specific tool having one unit length as well as a tool having a plurality of consecutive scales of the corresponding unit length. In the embodiment of the present invention, the unit length was set to 1 cm, and a mask having a width of 6 cm and a length of 6 cm provided with grid scales at intervals of 1 cm was set as the length measuring means (5) and an experiment was conducted.

In addition, the length measuring means 5 may be photographed together with the crop to be photographed in the video image, and may be provided at a position supporting the crop in the vicinity of the crop or between the crop and the bottom surface. The image size of the crop may be different from the initial image by modifying the image in the image image processing step to be described below. However, if the user sets the unit length by using the length measuring means 5 and measures it together with the crop, the size of the crop can be calculated from the modified image thereafter. That is, by comparing the unit length of the initially set length measuring means 5 and the unit length in the deformed image, the size change rate of the crop image can be grasped and this can be used to calculate the size of the crop.

The image processing unit 11 may calculate crop size information from the image image of the crop. In order to calculate the size information from the image image of the crop, it is necessary to calculate the number of pixels per inch (Dots Per Inch; DPI). In calculating the DPI, unit length information of the length measuring means 5 may be used. The image processing unit 11 may automatically count the number of pixels corresponding to the grid of each square of the grid with respect to the image image photographing the crop located on the length measuring means 5 provided in the form of a grid with a unit length, The size information of the crop can be calculated by counting the number of grids corresponding to the horizontal and vertical positions.

The image processing unit 11 may include a preprocessing module 111 , a conversion module 113 , and a size calculation module 115 .

The pre-processing module 111 may remove noise from the video image. Since the object of measurement and diagnosis in the crop growth diagnosis system 1 is the body of the crop, parts other than the body, such as stems or leaves, are judged as noise and separated from the background and removed.

The transform module 113 may Fourier transform the noise-removed video image. When the experiment is conducted using the length measuring means 5 as a measuring tool continuously provided with scales of a certain length, it is difficult to detect the grid scale of the length measuring means 5 depending on the state or condition in which the crop is photographed. have. Therefore, an image can be transformed by applying a two-dimensional Discrete Fourier Transform (DFT).

The size calculation module 115 may calculate crop size information from the converted image image. The crop size information may include a horizontal length (overwidth) and a vertical length (overheight) of the corresponding crop.

In the embodiment of the present experiment, the size of a grid having a unit length of 1 cm can be calculated using Equation 1 below. I(x,y) represents the size of the edge.

[Equation 1]

The database unit 13 may store size information and weight information of a plurality of crops calculated by the image processing unit 11 .

The relationship model generator 15 may calculate a size-weight relationship model of a crop by analyzing the relationship between size information and weight information according to growth stages. The relationship model generator 15 may set the size information of the crop as an independent variable and set the weight information as a dependent variable to statistically analyze the relationship between the size and weight. In this case, since the size information has two or more independent variables including the horizontal length and the vertical length, multiple regression analysis can be performed. The relationship model generator 15 may calculate a multiple regression model through multiple regression analysis. The multiple regression equation is as Equation 2 below.

[Equation 2]

는 y절편,

는 회귀 계수를 의미한다.x is the independent variable, y is the dependent variable,

is the y-intercept,

is the regression coefficient.

In the embodiment of the present invention, the input variables are given by the over-height and over-width and two interactions of the strawberry, and the weight of the strawberry is used as the output variable. And in order to estimate the regression coefficient of the multiple regression model, the method of minimizing the sum of squares of the error of Equation 3 was used.

[Equation 3]

The output unit 17 may input the size information calculated by the image processing unit 11 into the multiple regression model to calculate weight information as a dependent variable on the size of the crop to be investigated.

2 shows a response surface for a multiple regression model calculated by statistical analysis of the size and weight of strawberries as an experimental example of the present invention. In an embodiment of the present invention, a multiple regression model capable of predicting weight was calculated as in Equation 4 below using the overheight, overwidth, and interaction of strawberries.

[Equation 4]

When the calculated coefficient of determination of the multiple regression model of Equation 4 ^{is tested, R 2} is 0.98771, which is much higher than the general reference value (0.65). Therefore, it can be confirmed that the multiple regression model according to the embodiment of the present invention fits the response values well.

3 shows a process in which a method for diagnosing crop growth according to an embodiment of the present invention is performed. (a) of generating a relational model, (b) of taking a video image, (c) of image processing, (d) of calculating size information, (e) of calculating weight information. can

Step (a) refers to a step of storing the size and weight data of a plurality of crops, and calculating a multiple regression model by multiple regression analysis of the relationship between the size and weight of previously stored crops. Step (b) refers to a step of generating a video image by photographing a crop to be investigated. Step (c) refers to an image processing step of pre-processing the video image and performing a Fourier transform. Step (d) refers to a step of calculating crop size information from the converted video image. Step (e) refers to the step of inputting size information into the multiple regression model and calculating weight information for the size information of the research target crop. Steps (a) to (e) show the process performed in the embodiment of the crop growth diagnosis system 1 described above, and since the meaning and significance of each step has been described above, redundant descriptions will be omitted.

Although the present invention has been described in detail through representative embodiments above, those of ordinary skill in the art to which the present invention pertains will understand that various modifications are possible within the limits without departing from the scope of the present invention with respect to the above-described embodiments. will be. Therefore, the scope of the present invention should not be limited to the described embodiments and should be defined by all changes or modifications derived from the claims and equivalent concepts as well as the claims to be described later.

1: Crop Growth Diagnostic System
3: means of shooting
5: Length measuring means
11: image processing unit
111: preprocessing module
113: conversion module
115: size calculation module
13: database part
15: Relational model generation unit
17: output unit

Claims (5)

In the crop growth diagnosis system using an image image,
an image processing unit for calculating size information of the crop from the image image of the crop;
a database unit for storing the size information and the weight information of the crop; and
and a relationship model generator for calculating a relationship model by analyzing the relationship between the size information and the weight information.
The method of claim 1,
The video image is
It includes a length measurement means as a reference for calculating the size information of the crop,
The length measuring means is a crop growth diagnosis system, characterized in that photographed together with the crop.
The method of claim 1,
The size information is
Including the horizontal and vertical length of the crop,
The relationship model generation unit,
Crop growth diagnosis system, characterized in that the multiple regression model is calculated by setting the size information as an independent variable and performing multiple regression analysis by setting the weight information as a dependent variable.
The method of claim 1,
The image processing unit,
a pre-processing module for removing noise from the video image;
a transform module for Fourier transforming the video image from which noise has been removed; and
Crop growth diagnosis system, characterized in that it comprises a size calculation module for calculating the size information of the crop from the converted image image.
In the method for diagnosing crop growth using an image image,
(a) a relationship model generation step of storing the size and weight data of a plurality of crops, and calculating a multiple regression model by multiple regression analysis of the relationship between the size and weight of crops with the previously stored data;
(b) a photographing step of generating a video image by photographing a crop to be investigated;
(c) an image processing step of pre-processing the video image and performing a Fourier transform;
(d) a calculation step of calculating the size information of the crop from the converted video image;
(e) an output step of inputting the size information into the multiple regression model to calculate weight information for the size information of the research target crop;

Images