Method and apparatus for converting an overlay area into a 3D image

摘要

Apparatuses and methods for converting an overlay area of a two-dimensional (2D) monoscopic image into a three-dimensional (3D) image are provided. A depth map representing a perspective based on an input monoscopic image is generated to provide depth values of the input monoscopic image. At least one overlay area and at least one background image included in the input monoscopic image are then detected, and a depth value of one or more or both of the overlay area and the background image are corrected to a desired value. A stereoscopic image is generated based on the corrected depth value, such that the resulting stereoscopic image prevents inconsistent viewing depths between an overlay area and the background image in order to prevent viewer fatigue.

主权项

1. A method of converting an overlay area into a three-dimensional (3D) image, the method comprising:
generating a depth map of an input monoscopic image, wherein the depth map provides depth values of the input monoscopic image; detecting at least one overlay area included in the input monoscopic image, wherein the at least one overlay area comprises at least one of a text area and an on screen display area; selectively correcting a depth value of at least one of the detected at least one overlay area and a background image on the depth map, according to determination whether a type of at least one of the detected at least one overlay area is related to the text area or the on screen display area; and generating at least one stereoscopic image based on the corrected depth value, wherein the selectively correcting is performed by uniformly allocating a predetermined depth value to the detected at least one overlay area, when the correcting is performed on the detected at least one overlay area, wherein the selectively correcting comprises:
when the detected at least one overlay area is the text area, correcting only the depth value of the detected at least one overlay area; andwhen the detected at least one overlay area is the one screen display area, correcting both of the depth values of the detected at least one overlay area and the background image, or only the depth value of the background image, and wherein the detecting comprises detecting the at least one overlay area by analyzing at least one color component of the input monoscopic image to extract the at least one overlay area.