发明名称 SYSTEM AND METHOD FOR PROVIDING CONTENT-AWARE PERSISTENT ADVERTISEMENTS
摘要 A method and system of providing content-aware persistent advertisements comprises the steps of ingesting and collecting VDNA (Video DNA) fingerprints of media contents playing on various network-enabled terminals including Internet browsers, mobile devices, tablets, smart televisions and so on; sending the ingested fingerprints along with other information such as metadata, user's location, etc, to the content identification server via Internet or mobile networks, and selecting accurate advertisements according to the media content and relevant information, then finally pushing the advertisements back to user's terminal. With VDNA fingerprint identification, media contents are identified by content instead of metadata or other surrounding information, so that the result of identification can be considered genuine, and the chosen advertisements based on the media content can be persistent across multiple terminals playing the same media content.
申请公布号 US2015237410(A1) 申请公布日期 2015.08.20
申请号 US201414181683 申请日期 2014.02.16
申请人 Yu Lei;Wang Yangbin 发明人 Yu Lei;Wang Yangbin
分类号 H04N21/81;H04N21/8358;G06Q30/02 主分类号 H04N21/81
代理机构 代理人
主权项 1. A method for providing content-aware persistent advertisements including persistency of geo-location, time, environment and user preferences, based on identified contents by dynamic VDNA (Video DNA) fingerprint extraction, said method comprising: a) extracting and collecting VDNA fingerprints of media contents playing on various network-enabled terminals including Internet browsers, mobile devices, tablets, smart TVs (television), b) sending said extracted VDNA fingerprints along with other information including metadata and user's location to a content identification server via Internet or mobile networks, and c) selecting accurate advertisements according to said media contents identified by said VDNA fingerprints, and pushing said content-aware persistent advertisements back to said terminals,wherein said dynamic VDNA fingerprint extraction comprises: a) sampling a media frame as an image, b) dynamically dividing an input image into certain variable amount of equally sized squares, c) computing average value of RGB (Red, Green and Blue) values from each pixel in each said square, d) said VDNA fingerprint of said image being a two dimensional vector of the values from all divided squares, and e) said VDNA fingerprint is used to synchronize with said media frame including master frame and sample frame.
地址 Hangzhou CN