公开(公告)号：US11501471B2

公开(公告)日：2022-11-15

申请号：US17168868

申请日：2021-02-05

Applicant: CASIO COMPUTER CO., LTD.

Inventor： Shogo Imao

IPC: G06T11/00 ,  G06T7/11 ,  G06N3/04 ,  G06N3/08 ,  G06F11/00

Abstract: A method for generating virtual and real composite image data includes: acquiring captured image data capturing an image of a real space as seen from a user's point of view; inputting the captured image data into a trained model, the training model outputting segmentation data segmenting the captured image data into a first region in which a target object is displayed, a second region in which at least a part of the user's body is displayed, and a third region that is other than the first and second regions; and compositing data of the first region and data of the second region with a virtual space image data based on the segmentation data.