A Mechanism for Copyrighted Video Copy Detection and Identification

Rapid development in the field of multimedia technology has made easier to store and access large amount of multimedia data. There are lots of easy editing and publishing mechanism available which makes duplication of multimedia data feature very easy which can cause the violation of digital rights. So, copy rights security becomes a critical problem for the mutimedia data over the cloud. This has led to the requiring mutimedia copy protection and detection in a large range of the multimedia applications to find illegally-made duplicated copyrighted multimedia copies over the Internet. Watermarking technique was used for same purpose. Using this technique some distortion like Fourier transformation is added in multimedia copy and to play or view the multimedia object in original form this distortion need to be removed by some softwares or players, which makes dependent on those players. However Content based copy detection (CBCD) technique is independent of any software based dependency and identifies copyrighted materials from its content itself. In this technique signatures are created from content itself using SIFT. Content owners specify multimedia objects that they want to protect. Then, system creates signatures from that file and inserts them in the distributed index. The Crawl component once a day downloads recent objects from online hosting sites. The signatures for a query object are created. After downloading of all objects and the signatures are created, the signatures are matched with all indexed signature. If signature matches then this will trap the site which duplicates copyrighted copy without rights. This is effective mechanism, but to fasten this process we have used Audio video cryptosteganography which adds secrete information in multimedia object itself which can neither be removed nor can be identified visually as watermarks are visible. This secrete information is considered as copyrighted information before forming the digital signature. If this secrete information matches with secrete information from database then all contents in the multimedia will be cross verified using CBCD and in this way copyrighted material will be identified.