OMOTE: Real-time face tracking and projection mapping

Forget the faces of historic monuments, the new frontier of projection mapping is the faces of humans.

Created by Nobumichi Asai and friends, technical details behind the process are scant at the moment. From this Tumblr post, it’s clear that step one is a 3D laser scan of the model’s face.


And then… you know, eventually you get this:


Hat tip to David Hobizal.

Tags: , , , ,

About the author

Justin Cone

Together with Carlos El Asmar, Justin co-founded Motionographer, F5 and The Motion Awards. He currently lives in Austin, Texas with is wife, son and fluffball of a dog. Before taking on Motionographer full-time, Justin worked in various capacities at Psyop, NBC-Universal, Apple, Adobe and SCAD.

One Comment

Lucas Luo

I saw this last week, today I watch this again and it still amazed me and a bit creepy. Although 3D projection is not a brand new idea for now, back to 2010, there were already some amazing 3D projections on the building which you can found on Youtube. ( However, they just project on an autographic view of the project in the past. But this time is totally different, they projection the mapping on a face which is a three dimensional form. By watching though the video, we can found that even the person rotate her face or turning around, the projection still keep the same quality which means there is no bad spot or wrong perspective. The technique behind this is real-time 3d scanning and 3d projecting and the little light dots on her face are the reference for 3d scanner to locate her face. If you look carful enough, you will find that the girl never open her eyes, the eyes were projected on. It’s quite amazing how accurate and fast did the 3d scanner locate her face and calculate the right projection to project.
For the advertising and business standpoint, I think we can use this special technique or just this real-time scan/project process to promote our product. Maybe in the future, we don’t need to make a real working prototype for showing the audience how the product works. We may just need to build a rough white model and project the details and interface on the white model.

Comments are closed.