Generation of Stereo Images from the Heterogeneous Cameras

Abstract

Received: 15 January 2021 Accepted: 25 March 2021 As mobile devices with multiple cameras have been widespread, the cameras can be used for generating stereo images for depth sensing. However, it is difficult to use the raw images from the cameras because the cameras have different characteristics such as resolutions, distortions, and field of view in many cases. In this paper, we suggest methods to generate the stereo images from the heterogeneous cameras within 0.329 seconds and verify the methods. Although we use the images from a smartphone to generate the stereo images, the suggested methods can be used for any other devices with heterogeneous cameras. To the best of our knowledge, the suggested methods are the world-first one that generates the stereo images from the images by real mobile devices.