Exploring Agriculture Through the Use of AI

Assistant Professor (Agricultural Environmental Engineering and Agricultural Information Engineering)

Konno Shinsuke

In recent years, smart agricultural technologies—such as robotic farming machines and drones—have begun to spread across production sites. At the same time, the use of data in agriculture is rapidly advancing, with sensor-based data collection and AI-driven analysis expected to contribute to labor-saving operations, reduced workload, and the transfer of expert knowledge.

In our laboratory, we collect video data from rice paddies, crop fields, and livestock barns using drones and network cameras. We analyze this data through image recognition powered by artificial intelligence (AI).

For example, we use AI to analyze video footage of calves captured by network cameras in order to identify behavioral patterns and examine their relationship with the rearing environment. We also use drones to capture aerial images of rice paddies, enabling us to assess spatial variation in crop growth and estimate potential yields.

By analyzing data from real agricultural production fields, we aim to contribute to the realization of efficient and precise agricultural practices.

▲Annotating Images for Training an Object Detection AI Model

▲Classifying Calf Postures Using an Object Detection AI Model

▲Analyzing Entire Fields by Creating Orthomosaic Maps from Drone-Captured Aerial Images