发明名称 Systems and methods for computation-efficient image processing system architecture
摘要 Embodiments relate to systems and methods for a computation-efficient image processing system architecture. Image data can be transmitted from a computer, online service, and/or other image source to an output device having a set of image processing modules in two or more image paths, including an edge detection module and a video decoding module. The edge detection module can produce edge tag output, and the video decoding module, operating in parallel, can generate decoded video output. The edge tag output and decoded video output can be transmitted to a set of downstream image processing modules, including modules for color trapping, edge smoothing, and other operations. Because earlier processing stages share information with downstream modules which require the same or related data, redundant processing can be reduced or eliminated. Complex image operations can therefore be carried out, and high-quality output can be generated, without sacrificing responsiveness.
申请公布号 US9307119(B2) 申请公布日期 2016.04.05
申请号 US201213587331 申请日期 2012.08.16
申请人 XEROX CORPORATION 发明人 Li Xing;Metcalfe David Jon;Yao Meng;Stevens Michael Dale;Phillippi Paul W.
分类号 H04N1/56;G06K1/12 主分类号 H04N1/56
代理机构 MH2 Technology Law Group LLP 代理人 MH2 Technology Law Group LLP
主权项 1. A method of processing image information, comprising: accessing an image source to receive image data; performing a video decoding operation on the image data to generate a decoded video output; performing, on the decoded video output, at least— a black trapping operation to locate black areas of the decoded video output and generate a black trapping output,a thin line growth operation to generate a thin line growth output,a black enhancement operation to generate a black enhancement output, anda color estimation operation to generate a color estimation output; combining at least the black trapping output, the thin line growth output, and the black enhancement output to generate an adjusted video output; performing, in parallel with the video decoding operation, an edge detection operation on the image data to generate an edge tag output; performing an edge smoothing operation based on the edge tag output and the color estimation output to generate an edge smoothed output; and merging the adjusted video output and the edge smoothed output to generate a finished image output.
地址 Norwalk CT US