METHOD AND APPARATUS FOR OBTAINING PLENOPTIC CAMERA ARRAY DEPTH IMAGE BASED ON ARTIFICIAL NEURAL NETWORK

摘要

Provided is a method for utilizing plenoptic geometric characteristics of an image by simultaneously using array images photographed by a multi-camera array. To this end, the present specification provides a method for obtaining depth information which includes the following steps: obtaining an array depth image from each of a plurality of cameras included in a multi-camera array; obtaining a plenoptic array image from the array depth image obtained from each of the plurality of cameras; converting the plenoptic array image into the same coordinate system so as to obtain a corrected array image; and obtaining depth information from the obtained corrected array image.;COPYRIGHT KIPO 2019