In a number of application scenarios, proper video signals may exhibit simultaneous correlation characteristics over the space and time dimensions which jointly describe periodic features or behaviors. Examples of such scenarios may be found in video monitoring of physical systems, sport and athlete coaching with automatic video supervision, biomedical applications to newborn video monitoring for the detection of epileptic seizures or apnea episodes, surveillance systems and others. A general Maximum Likelihood (ML) approach to the detection of common periodic features possibly present in a set of video signals and the estimation of their characteristics, such as the fundamental frequency and the local amplitude, is proposed. Application examples in various scenarios are presented and the performance of the proposed ML solutions is shown to be effective.
Extraction of Periodic Features from Video Signals / Alinovi, D.; Raheli, R.. - (2017), pp. 95-100. (Intervento presentato al convegno MMEDIA 2017, The Ninth International Conferences on Advances in Multimedia tenutosi a Venice, Italy nel April 2017).
Extraction of Periodic Features from Video Signals
D. Alinovi;R. Raheli
2017-01-01
Abstract
In a number of application scenarios, proper video signals may exhibit simultaneous correlation characteristics over the space and time dimensions which jointly describe periodic features or behaviors. Examples of such scenarios may be found in video monitoring of physical systems, sport and athlete coaching with automatic video supervision, biomedical applications to newborn video monitoring for the detection of epileptic seizures or apnea episodes, surveillance systems and others. A general Maximum Likelihood (ML) approach to the detection of common periodic features possibly present in a set of video signals and the estimation of their characteristics, such as the fundamental frequency and the local amplitude, is proposed. Application examples in various scenarios are presented and the performance of the proposed ML solutions is shown to be effective.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.