Abstract
As the applications using augmented reality have been widespread, the
depth information on mobile devices has been important. For example, it
can be used for the natural interaction with the environment in
augmented reality and the bokeh effects during image processing.
However, it is difficult to get depth information on general mobile
devices without any depth sensors. As one of the methods for the depth
information without any special sensors, stereo vision through recent
multiple cameras can be used to generate the disparity map, which can be
finally used to generate the depth map. However, the cameras have
heterogeneous characteristics each vendor, which makes it challenging to
utilize the cameras. In this research, we propose a method for the
disparity map to use the cameras that most of the recent devices
internally include. With our method, the devices don’t have to include
depth sensors such as ToF (Time Of Flight) and SL (Structured Light) and
could get the depth information at the raw image from the camera within
275.64 msec (3.63 FPS) on commodity devices. As far as we know, the
proposed method is the first research that can generate the disparity
map from captured figures by commercial mobile devices.