A Projector-Camera System for Ironing Support with Wrinkle Enhancement

Abstract

Ironing is one of troublesome houseworks, in which the goal of the task is to remove wrinkles caused during washing. A projector has advantages in physical world instruction over an instruction sheet, a Head Mounted Display, or a smartphone/tablet PC because of direct mapping of instructive information on the target object. In this article, we propose a method to detect wrinkles using machine-learning and a system to present detected wrinkles by enhancing the area of wrinkles through a projector. In total, 47 infrared image features are defined, from which 15 features are finally used, to classify 32 pixels squares (about 4.5 cm squares) of regions of interest into one of four classes including wrinkle, flat, sagging, and tuck. A RandomForest classifier successfully identified 93.0 % of the wrinkle class. The comparison of wrinkle enhancement methods implies that presenting all ROIs on an ironing board at a time is more effective in removing wrinkles than enhancing an area around and ahead of an iron. Also, we found that making a user realize the effect of wrinkle removal is important to reduce wrinkles efficiently and showed prospective solutions for this issue.

Estimating Visibility of Annotations for View Management in Spatial Augmented Reality Based on Machine-Learning Techniques

“…Furthermore, the projected information can be shared with others in a relatively large display space. These characteristics allows applications to be proposed in various domains, which include domestic work [3,4], assembly tasks [5,6,7], education [8,9,10,11], physical training [12], appearance control for visually impaired [13], entertainment [14], etc.…”