A System to Find the Change of One’s Vision Implicitly

Myopia has become one of the most serious health problems in the world as it grows rapidly in the young people. Efficient vision monitoring methods are required to determine myopia in its early stages so as to take it for further treatment. Current technologies for vision examination are usually either expensive or time costly. This paper proposes a nearsightedness monitoring system, which exploits the widely used smartphones to detect the deterioration of nearsightedness by monitoring and analyzing the distance between the eyes and the smartphone screen. The detection process is implicit since the system is implemented on the daily used phones and people do not have to be interrupted frequently. The proposed system consists of two key components: activity recognition component and the eye detecting component. The activity recognition component is used to determine whether a person is watching the phone. Once a person is watching the phone, the eye detecting component can be triggered to click a photograph of the person using the front camera of the phone and localize the two eyes on the picture taken. The distance between eyes and the screen is estimated by the ratio of the two-eye distance in the picture and the width of the picture. This paper uses the 3-axis acceleration sensor and the front camera which are equipped on most of the smartphones for activity recognition and eye detection respectively. A prototype for has been developed in order to evaluate the effectiveness of the system under various environmental conditions. From more than half year monitoring period which consisted of about 20 volunteers, we were able to accurately detect the degradation of nearsightedness in two volunteers.