KR-20260067452-A - SYSTEM AND METHOD FOR PREDICTING OPTIMUM DISPLACEMENT TIME AFTER RAISING SEEDLINGS OF ONION BASED ON ARTIFICIAL INTELLIGENCE

KR20260067452AKR 20260067452 AKR20260067452 AKR 20260067452AKR-20260067452-A

Abstract

An artificial intelligence-based system for predicting the optimal planting time for onion seedlings and a prediction method thereof are disclosed. An artificial intelligence-based system for predicting the optimal planting time for onion seedlings and a prediction method thereof include: a data collection unit that collects indoor onion seedling growth image data from an image acquisition device, collects outdoor open-field environment data from various sensors installed outdoors, and collects weather data from a weather agency server; a data preprocessing unit that analyzes the growth state of seedlings from indoor onion seedling growth image data using a Convolutional Neural Network (CNN), normalizes outdoor open-field environment data and weather data, and supplements missing values; and an optimal planting time prediction unit that predicts future weather data from the growth state of seedlings, normalized outdoor open-field environment data and weather data using a Long Short-Term Memory (LSTM) or Recurrent Neural Network (RNN) model, and predicts the optimal planting time from the growth state of seedlings, normalized outdoor open-field environment data and weather data with missing values supplemented, using an eXtreme Gradient Boosting (XGBoost) or Gradient Boosting Machine (GBM).

Inventors

서우종
김현수

Assignees

주식회사 인지솔루션

Dates

Publication Date: 20260513
Application Date: 20241104

Claims (8)

A data collection unit that collects indoor onion seedling growth image data from an image acquisition device, collects outdoor open-field environment data from various sensors installed outdoors, and collects weather data from a weather agency server; A data preprocessing unit that analyzes the growth status of seedlings from the indoor onion seedling growth image data using a CNN (Convolutional Neural Network), normalizes the outdoor open-field environment data and the weather data, and supplements missing values; and An artificial intelligence-based optimal planting time prediction system for onion seedlings, characterized by comprising: an optimal planting time prediction unit that predicts future weather data from the growth state of the seedling, the outdoor open-field environment data normalized and missing values supplemented, and the weather data using an LSTM (Long Short-Term Memory) or RNN (Recurrent Neural Network) model, and predicts the optimal planting time from the growth state of the seedling, the outdoor open-field environment data normalized and missing values supplemented, and the predicted weather data using XGBoost (eXtreme Gradient Boosting) or GBM (Gradient Boosting Machine).
In paragraph 1, An AI-based system for predicting the optimal planting time for onion seedlings, characterized in that the image acquisition device is a time-lapse camera, and the data preprocessing unit receives a time-lapse video of an indoor onion seedling from the data collection unit and analyzes the growth status of the seedling.
In paragraph 2, An AI-based onion seedling optimal planting time prediction system characterized in that the growth state of the seedling is at least one of the growth rate of the seedling's fruit, the flowering rate of the seedling, and the correlation of fruit quality according to the growth rate of the seedling's fruit.
In paragraph 1, It further includes a pest prediction unit that predicts pests from a training dataset obtained through the mapping of the pest occurrence timing and environmental conditions, the aforementioned outdoor open-field environment data and weather data, which have been normalized and have missing values supplemented through Logistic Regression and Support Vector Machine (SVM). The above data collection unit further collects past pest occurrence data from the pest database, and An artificial intelligence-based system for predicting the optimal planting time for onion seedlings, characterized in that the data preprocessing unit generates a learning data set through mapping the pest occurrence time and environmental conditions from the past pest occurrence data.
In paragraph 1, An artificial intelligence-based system for predicting the optimal planting time for onion seedlings, characterized in that the above indoor onion seedling growth video data is temperature or humidity.
In paragraph 1, An artificial intelligence-based system for predicting the optimal planting time for onion seedlings, characterized in that the above outdoor open-field environment data is one of temperature, humidity, light intensity, pH, or EC.
A step in which a data collection unit collects indoor onion seedling growth image data from an image acquisition device, collects outdoor open-field environment data from various sensors installed outdoors, and collects weather data from a weather agency server; The data preprocessing unit analyzes the growth status of seedlings from the indoor onion seedling growth image data using a CNN (Convolutional Neural Network), normalizes the outdoor open-field environment data and the weather data, and supplements missing values; and An artificial intelligence-based method for predicting the optimal planting time for onion seedlings, characterized by comprising the step of: an optimal planting time prediction unit predicting future weather data from the growth state of the seedling, the outdoor open-field environment data normalized and missing values supplemented, and the weather data using an LSTM (Long Short-Term Memory) or RNN (Recurrent Neural Network) model; and predicting the optimal planting time from the growth state of the seedling, the outdoor open-field environment data normalized and missing values supplemented, and the predicted weather data using XGBoost (eXtreme Gradient Boosting) or GBM (Gradient Boosting Machine).
In Paragraph 7, A step in which the above data collection unit further collects past pest occurrence data from the pest database; The above data preprocessing unit generates a learning dataset by mapping the pest occurrence time and environmental conditions from the past pest occurrence data; and An artificial intelligence-based method for predicting the optimal planting time for onion seedlings, characterized by further including the step of predicting pests from a training dataset obtained by mapping the timing of pest occurrence and environmental conditions, the outdoor open-field environment data normalized to compensate for missing values and the weather data, through Logistic Regression and Support Vector Machine (SVM).

Description

System and Method for Predicting Optimum Displacement Time After Raising Seedlings of Onion Based on Artificial Intelligence The present invention relates to an artificial intelligence-based system for predicting the optimal planting time for onion seedlings and a method for predicting the same. In particular, it relates to an artificial intelligence-based system for predicting the optimal planting time for onion seedlings and a method for predicting the optimal planting time from various data based on artificial intelligence after onion seedling cultivation in a farm. Generally, onions are crops cultivated by sowing seeds and transplanting seedlings that have developed. In this case, the seedlings refer to plants that have grown to a height of 25–30 cm, a stem thickness of 6–8 mm, and 3–4 leaves. They are transported in bundles, and after ensuring sufficient distance, they are planted with ample spacing. In other words, furrows are dug in the field to be cultivated for onions, seedlings are placed in the furrows in rows and at intervals, and then new furrows are dug at a location spaced apart from the previously formed furrows to cover the roots of the seedlings. The process of placing seedlings in new furrows is then repeated to transplant the seedlings. Since this seedling transplanting process relies on manual labor, a large number of workers are required; however, due to the labor shortage in rural areas, even this is difficult, leading to a surge in farmers giving up on onion cultivation. Accordingly, a technology for automatically sowing seeds was previously disclosed in Utility Model Publication No. 20-1989-0015774 (Title: Seeder), which consists of a furrowing device adjusted by a spring, a seed container equipped with a seed moving device and a seed moving aid, and a seed cover equipped with a spring and seed cover wings. FIG. 1 is a block diagram of an artificial intelligence-based system for predicting the optimal planting time of onion seedlings according to one embodiment of the present invention. FIG. 2 is a schematic diagram showing the process of predicting the optimal planting time for onion seedlings based on artificial intelligence according to one embodiment of the present invention. FIG. 3 is a schematic diagram showing the process of selecting the appropriate timing for irrigation, fertilization, and pesticide spraying according to an open-field environment in accordance with one embodiment of the present invention. FIG. 4 is a schematic diagram of an integrated concept for predicting the optimal planting time for onion seedlings and predicting pests and diseases based on artificial intelligence according to one embodiment of the present invention. FIG. 5 is a flowchart of an artificial intelligence-based method for predicting the optimal planting time for onion seedlings according to one embodiment of the present invention. Figure 6 is a flowchart with a pest and disease prediction method added to Figure 5. The advantages and features of the present invention and the methods for achieving them will become clear by referring to the embodiments described below in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. The embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims. Hereinafter, an artificial intelligence-based system for predicting the optimal planting time for onion seedlings and a prediction method according to an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram of an artificial intelligence-based system for predicting the optimal planting time of onion seedlings according to one embodiment of the present invention. An artificial intelligence-based onion seedling optimal planting time prediction system (1000) may include a data collection unit (100), a data preprocessing unit (200), an optimal planting time prediction unit (300), and a pest and disease prediction unit (400). The data collection unit (100) can collect indoor onion seedling growth image data from the image acquisition device (10), collect outdoor open-field environment data from various sensors (20) installed outdoors, collect weather data from the weather agency server (30), and collect past pest and disease occurrence data from the pest and disease database (40). Here, the image acquisition device (10) may be a time-lapse camera capable of capturing a plant at intervals selected from 0.3 to 60 seconds, but is not limited thereto. The data preprocessing unit (200) receives a time-lapse video of an indoor onion seedling from the data collection unit (100) and analyzes the growth status of the seedling. Here, the growth status of the