Researches

What is needed to conserve tropical forests and ensure appropriate and efficient forest management?

I believe one of the answers is to accurately assess the condition and value of the forest.

In particular, understanding carbon stocks and biodiversity is expected to lead to new incentives for forest conservation in the future, as in international movements such as REDD+.

I am working from both a research and business perspective to develop new technologies to enable appropriate use and decision making for such forests.

Fig. Automatic tree species identification from popular UAV (phantom 4 pro) using deep learning

With this in mind, I have developed “DF Scanner“, a UAV-based forest analysis software, so that many people can use the deep learning-based tree species identification technology I have been developing. This software not only can identify and learn tree species using deep learning, but also can detect tree tops, separate tree crowns, estimate DBH and timber volume. In the future, I plan to make it possible to evaluate carbon stocks and biodiversity.

I hope that this research and development will help improve the efficiency of forest management and conservation.

Masanori Onishi