AI and
Computer Vision

We have been conducting research on

generating 3D models of greenhouses,

farmlands, and orchard environments using

multiple 2D images acquired through

drones and mobile robots.

​

Recently, our focus has been on real-time

3D modeling of greenhouses and orchard

environments using multiple cameras.

We have been conducting research on

measuring and mapping the characteristics

of fruit trees using orchard 3D models.

​

Recently, we have been using deep learning

instance segmentation algorithms to accurately

measure and map the number, size, and

positions of fruits for each orchard tree.

We have been conducting research on

target recognition for flower thinning, fruit thinning,

and pruning tasks in orchar management robots.

​

Recently, our focus has been on performing

accurate pruning position recognition using

deep learning-based 3D image analysis.

We have been conducting research on crop

and farmland characterization using drone

and PTZ (Pan-Tilt-Zoom) CCTV cameras.

 

Recently, we have been focusing on utilizing

deep-learning based computer vision

technologies to measure farmland water levels

and topography, as well as to estimate

biomass and yield for crops.

We have been conducting research

on real-time weed recognition in field

environments.

 

Recently, our focus has been on real-time

recognition of weeds and their growth points

in both plastic-film mulch row and bare soil

row for the development

of a laser weeding robot.

We have been conducting research for

monitoring and analyzing the characteristics

of broiler chickens in a large indoor poultry

farm designed for animal welfare.

 

Recently, we have been analyzing the number

of chickens, their activity levels, feeding and

drinking behavior, as well as measuring their

body weights using PTZ CCTV cameras.