发明名称 Method and apparatus for adjusting 3D depth of object and method for detecting 3D depth of object
摘要 A method for adjusting three-dimensional (3D) depth including performing block matching on left-eye/right-eye frames of a stereo video signal to generate a motion vector of a specific object and to obtain 3D depth of the specific object; and adjusting 3D depth of an additive object according to that of the specific object, wherein the additive object is one of a subtitle object, an on screen display (OSD) object or a program guide object.
申请公布号 US8970670(B2) 申请公布日期 2015.03.03
申请号 US201113300840 申请日期 2011.11.21
申请人 MStar Semiconductor, Inc. 发明人 Wang Chien Shun
分类号 H04N13/00 主分类号 H04N13/00
代理机构 Edell, Shapiro & Finnan, LLC 代理人 Edell, Shapiro & Finnan, LLC
主权项 1. A method for dynamically adjusting three-dimensional (3D) depth of an additive object of a stereo video signal, comprising: using a calculating circuit, detecting 3D depth of a specific object in a plurality of frames of the stereo video signal; and using an adjusting circuit, adjusting the 3D depth of the additive object according to the 3D depth of the specific object, wherein the plurality of frames comprises a left-eye frame and a right-eye frame, and the step of detecting the 3D depth of the specific object in the plurality of frames of the stereo video signal comprises: performing block matching on the left-eye frame and the right-eye frame to generate a motion vector corresponding to the specific object, wherein the block matching is performed only in a horizontal direction; and obtaining the 3D depth of the specific object according to the motion vector, wherein the step of performing block matching on the left-eye frame and the right-eye frame to generate the motion vector corresponding to the specific object comprises: performing block matching on the left-eye frame and the right-eye frame to respectively generate a plurality of motion vectors corresponding to a plurality of image blocks; counting the number of the plurality of motion vectors corresponding to each motion vector value to obtain a specific motion vector; and defining an image block corresponding to the specific motion vector as an image region corresponding to the specific object, with the motion vector of the specific object being the specific motion vector.
地址 Hsinchu Hsien TW