News Column

Patent Issued for System and Method of Revising Depth of a 3D Image Pair

January 28, 2014

By a News Reporter-Staff News Editor at China Weekly News -- National Taiwan University ( Taipei , TW) has been issued patent number 8629901, according to news reporting originating out of Alexandria, Virginia , by VerticalNews editors. The patent's inventors are Chen, Liang-Gee ( Taipei , TW); Wu, Chien ( Taipei , TW); Li, Chung-Te ( Taipei , TW); Lai, Yen-Chieh ( Taipei , TW); Cheng, Chao-Chung ( Taipei , TW); Huang, Ling-Hsiu ( Tainan , TW). This patent was filed on May 19, 2011 and was published online on January 14, 2014 . From the background information supplied by the inventors, news correspondents obtained the following quote: "The present invention generally relates to digital image processing, and more particularly to a system and method of revising depth generated based on a three-dimensional (3D) image pair. "When three-dimensional (3D) objects are mapped onto a two-dimensional (2D) image plane by prospective projection, such as an image taken by a still camera or a video camera, a lot of information, particularly 3D depth information, disappears. A 3D imaging system, however, can convey 3D information to a viewer by recording 3D visual information or by re-creating the illusion of depth. Although the 3D imaging technique has been known for over a century, the 3D display becomes more practical and popular owing to availability of high-resolution and low-price displays such as liquid crystal displays (LCDs). "FIG. 1 shows a block diagram of a conventional 3D imaging system 1 that creates depth information by a depth generator 11 according to a 2D image input. The depth information is then processed by the depth-image-based rendering (DIBR) 12 to generate a left (L) image 14A and a right image 14B, which are then displayed and viewed by the viewer. "Moreover, FIG. 2 shows a block diagram of another conventional 3D imaging system 2 that captures a left (L) image 20A and a right image 20B from a target object by two cameras respectively. The depth generator 21 utilities stereo matching technique such as block matching to acquire the left and right depth information from a stereo image pair, L image 20A and R image 20B, respectively. The left and right depth information is then processed by the depth-image-based rendering (DIBR) 22 to generate at least two images with different viewpoint (i.e., at least one left (L') image 24A and at least one right (R') image 24B), which should be viewed by the viewer, according to the matching relation of the L image 20A and R image 201B. "However, for the conventional 3D imaging system 2, there are still some basic constraints in stereo videos, such as occlusion, two-camera setting parameter deviation, which may affect the generated depth information. Therefore, if only considering the matching relation of the stereo image pair, some pixels are probably induced to the unreliable depth information. "For the reason that conventional 3D imaging systems could not effectively display 3D image or video, a need has arisen to propose a novel system and method of revising depth of a 3D image that could recapture or generate 3D depth information, to faithfully and easily recover or approximate a full 3D representation." Supplementing the background information on this patent, VerticalNews reporters also obtained the inventors' summary information for this patent: "In view of the foregoing, it is an object of the embodiment of the present invention to provide a system and method of revising depth of a three-dimensional (3D) image to increase quality of a 3D image or video. "According to one embodiment, a system of revising depth of a three-dimensional (3D) image pair which comprises a depth generator and a depth revisor is disclosed. The depth generator creates at least one initial depth map associated with one image of the 3D image pair, wherein the image comprises a plurality of pixels, and the initial depth map carries an initial depth value of each pixel of one image of the 3D image pair. The depth revisor comprises an inconsistence detection unit and an interpolation unit. The inconsistence detection unit detects the inconsistence among the pixels to estimate a reliable map. The interpolation unit interpolates the initial depth value according to the reliable map and the proximate pixels, so as to generate a revised depth map by revising the initial depth value. "According to another embodiment, a method of revising depth of a three-dimensional (3D) image pair is disclosed. The method comprises the following steps: firstly, at least one initial depth map associated with one image of the 3D image pair is received, wherein the initial depth map carries an initial depth value of each pixel of one image of the 3D image pair. Then, the inconsistence among the pixels of the image is detected to estimate a reliable map the initial depth value is interpolated according to the reliable map and the proximate pixels, so as to generate a revised depth map by revising the initial depth value." For the URL and additional information on this patent, see: Chen, Liang-Gee; Wu, Chien; Li, Chung-Te; Lai, Yen-Chieh; Cheng, Chao-Chung; Huang, Ling-Hsiu. System and Method of Revising Depth of a 3D Image Pair. U.S. Patent Number 8629901, filed May 19, 2011 , and published online on January 14, 2014 . Patent URL: http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=41&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=2026&f=G&l=50&co1=AND&d=PTXT&s1=20140114.PD.&OS=ISD/20140114&RS=ISD/20140114 Keywords for this news article include: Asia , National Taiwan University . Our reports deliver fact-based news of research and discoveries from around the world. Copyright 2014, NewsRx LLC


For more stories covering arts and entertainment, please see HispanicBusiness' Arts & Entertainment Channel



Source: China Weekly News


Story Tools