Video Copy Detection providing Localized Matches.

摘要: With the availability of large scale online video platforms like YouTube, copyright infringement becomes a severe problem, such that demand for robust copy detection systems is growing. Such system must find multiple occurrence protected material within clips are created, modified, remixed and uploaded by user. A particular challenge to exact position in – potentially huge reference database. For this purpose, paper presents Content Based Copy Detection both detects copies query videos against database gives an alignment between them. finding aligning matching shot, fast search candidates conducted, as second step found using dynamic programming minimization well-known edit distance from text retrieval. The introduced approach was evaluated on public available MUSCLE-VCD-2007 [LTJB07] data corpus showed competitive results compared ACM CVPR 2007 evaluation. GI-Topic: KI-BV (artificial intelligence image understanding)