2017 Volume 16 Issue 2 Pages 95-104
Using a smart phone while walking has been becoming a social issue. Acceleration-based applications to prevent texting while walking are now being developed, but their accuracy rate is low and false detection occurs because only a single element is used based on the required number of steps. The problem in texting while walking is that people are gazing at the smart phone to operate it while walking. Therefore, the ultimate objective of this study was set to develop a highly accurate application in consideration of false detection issues, focusing on the face authentication technology and eye detection with the in-camera. We developed a trial application, conducted eye gaze detection verification experiments using the face authentication technology, and achieved a high accuracy rate of about 96%. In addition, eye detection initially showed a low detection rate in the study, but we were able to minimize false positives by using a cascade classifier. Based on this result, we believe that this technology can be used to detect the gaze of smart phone users.
