Most market available interactive projectors use laser curtain technology that limits its usage to flat surface only. It also requires accurate calibration that challenges the user who normally does not have such expertise.
Machine vision technologies by detecting and recognizing human actions such as finger touching and gesture, however, will relax such limitations and expand the range of interaction from flat surface to any 3D surface and even to space.
Despite the recent advances in computer vision and deep learning for object detection and classification, realizing accurate depth perception and generalization of the algorithm to any uncontrolled environment are still challenging in particular with consumer level sensors. This project tries to address those challenges and focuses on developing computer vision and deep learning technologies.
Camera-projector calibration
Scene understanding
Scene interaction
Thank you for your interest in Epson. To subscribe and receive promotional emails, please visit Epson Global to find your local site.