Analyzing periodic motion classification

Xavier Orriols, Xavier Binefa

    Research output: Contribution to journalArticleResearchpeer-review

    Abstract

    In this paper, we present a new technique for separating different types of periodic motions in a video sequence. We consider different motions those that have different periodic patterns with one or many fundamental frequencies. We select the temporal Fourier Transform for each pixel to be the representation space for a sequence of images. The classification is performed using Non-Negative Matrix Factorization (NNMF) over the power spectra data set. The paper we present can be applied on a wide range of applications for video sequences analysis, such as: background subtraction on non-static backgrounds framework, object segmentation and classification. We point out the fact that no registration technique is applied in the method that we introduce. Nevertheless, this method can be used as a cooperative tool for the existing techniques based on camera motion models (motion segmentation, layer classification, tracking of moving objects, etc). © Springer-Verlag Berlin Heidelberg 2003.
    Original languageEnglish
    Pages (from-to)673-680
    JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
    Volume2652
    Publication statusPublished - 1 Dec 2003

    Fingerprint Dive into the research topics of 'Analyzing periodic motion classification'. Together they form a unique fingerprint.

    Cite this