Accuracy and stability improvement of tomography video signatures

Video signature techniques based on tomography images address the problem of video identification. This method relies on temporal segmentation and sampling strategies to build and determine the unique elements that form the signature. In this paper an extension for these methods is presented; first an implementation for high contrast sub-sampling and up-sampling filters is used to increase tomography edge stability, then a robust temporal video segmentation system is used to replace the original method applied to determine shot changes more accurately, and finally a new feature extraction method, derived from the previously proposed sampling pattern, was implemented and tested, resulting in a highly distinctive set of signature elements. The video set for the test was recorded from live broadcast TV, using as queries 20 different commercials, with the objective of analyzing the audit proficiency of the signatures for new applications. The performance obtain over the version presented in is significant with 99.58% of Recall and 99.35% of Precision prediction.