Overview
As an application of AI / IoT in institutional cultivation, we are conducting joint research with Air Water Inc. on yield prediction, vegetative state prediction, and expert systems.
Purpose
We are conducting research with the goal of realizing a highly practical facility gardening AI system (yield prediction, tree condition estimation, expert).
Contents
We are working on the following three themes.
(1) High-precision short-term yield prediction using AI
(2) Extraction of vegetative state features by tomato tree motion image analysis
(3) Expert system for farm management support
I will explain the outline of each theme.
(1) High-precision short-term yield prediction using AI
Short-term yield forecasting is a very important issue for farm management in terms of management profit. In this theme, we aim to realize more accurate short-term yield prediction by comprehensively judging external factors (temperature, amount of solar radiation, etc.), internal factors (tomato tree condition), current fruit set condition, etc. ..
(2) Extraction of vegetative state features by tomato tree motion image analysis
We are conducting research on phenotyping, which extracts the features of tomato trees from images and videos and estimates their state. Specifically, we are working on the estimation of leaf area index (LAI) and the estimation of tomato tree condition such as balance and tree vigor.
(3) Expert system for farm management support
Farm management in facility cultivation requires enormous knowledge, experience, and intuition, which is an obstacle to new entrants. In order to reduce this obstacle, we are working on research on an expert system that systematizes the knowledge of experienced growers (farm managers) and supports actual farm management.