Greenhouse Automation using Machine Learning

In Japan, the productivity of the agricultural sector is a key concern for society. One of the solutions to maintain this productivity is the increased automation of greenhouses.

We are interested in using robots for the maintenance of tomato cultures. Is it possible to build robots to automatically pick fruits when they are ready to harvest, that can identify plants that are sick or not developing properly, and that can intelligently perform maintenance tasks in those plants?

Segmentation of Tomato Plants

This research concerns the use of machine learning techniques to process images taken inside a greenhouse, and also the development of robots that can successfully operate in that environment.

Other information

This research is done in partnership with the Faculty of Life and Environmental Sciences, and the Faculty of Engineering, both at the University of Tsukuba, and funded by the OPERA project.