Within the Department of Computer Science at the University of York, UK, we have been conducting research into the problem of detecting tracks (narrowband frequencies) in spectrogram images. This research has resulted in several publications, some of which we believe may be of interest to your community as they give insight into the detection of marine mammal vocalisations in passive sonar data.
The research was initiated by conducting a broad survey of existing algorithms that have been applied to the problem of spectrogram track detection and this resulted in the paper "A Survey of Spectrogram Track Detection Algorithms", Applied Acoustics, (http://dx.doi.org/10.1016/j.apacoust.2009.08.007) also available from http://www.cs.york.ac.uk/arch/publications/byyear/2009/Lampert10Survey.pdf. Our most recent work has been concerned with the development of several novel low-level track detection operators and applying computer vision and pattern recognition algorithms to the problem. The results of this research have been published in the paper "A Detailed Investigation into Low-Level Feature Detection in Spectrogram Images", Pattern Recognition, (http://dx.doi.org/10.1016/j.patcog.2011.02.014) also available from http://www.cs.york.ac.uk/arch/publications/byyear/2011/file.2011-03-10.82433 94815. We also have a paper under review titled "Active Contour Detection of Narrowband Frequency Tracks in Spectrogram Images", IEEE Transactions in Pattern Analysis and Machine Intelligence, which extends our previously presented work (http://dx.doi.org/10.1016/j.patrec.2009.09.021) and describes a high-level track detection algorithm. Implementations of the algorithms that have been used within these papers are available from http://stdetect.googlecode.com. This implementation is geared towards recreating the results presented in the publications; however, for evaluation we have included functions to call the detection methods directly. It is our hope that this research may benefit the conservation and monitoring of marine mammals. We are keen to hear from anyone interested in this work (and please pass this information to anyone who you think may be interested). Yours sincerely, Tom Lampert _____________________________________ Dr. Thomas Lampert Research Associate Advanced Computer Architecture Group Department of Computer Science University of York. Phone: +44 (0)1904 325641 Office: CSE/219 Website: http://www-users.cs.york.ac.uk/tomal/ Email Disclaimer: http://www.york.ac.uk/docs/disclaimer/email.htm
_______________________________________________ MARMAM mailing list [email protected] https://lists.uvic.ca/mailman/listinfo/marmam
