Bird Species Identification and Population Estimation by Computerized Sound Analysis

This project developed hardware and software to automate the monitoring of birds. Results demonstrate that automated monitoring can provide comparable or superior species detection to current point count survey methods for species that vocalize, and can acquire more comprehensive and reliable data than current methods, particularly for rare and infrequent species. Reliable, indisputable biological survey data in the form of recordings can also avoid legal challenges and disputes that can delay projects. This project developed and refined hardware that can be deployed by any field biologist or competent technician and acquire field data for weeks or months at a time for later retrieval and processing. Processing long duration recordings by manual listening to find focal species vocalizations would present a daunting task, and require at least as much time as the duration of the recordings. The signal processing software developed by this project can automatically analyze this data burden to rapidly scan and identify target species. Unlike intermittent personnel-based surveys, the automated bioacoustic monitoring system developed by this project can operate continuously and thereby sample more intensively (and economically) than that possible with human observers and thus enable more confident species evaluation, and allow a more thorough assessment of species movements, abundance, and presence or absence. Continuous monitoring can also provide more consistent data from survey to survey to better reveal long-term population trends of species.

• Record URL:
  https://dot.ca.gov/-/media/dot-media/programs/research-innovation-system-information/documents/final-reports/ca04-0661-finalreport-a11y.pdf
• Record URL:
  https://merritt.cdlib.org/api/presign-file/ark%253A%252F13030%252Fm5838dkh/2/producer%252FCA04-0661.pdf
• 
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/oclc/883430574
• Corporate Authors:

  California Department of Transportation

  Division of Research and Innovation
  1227 O Street, MS-83
  Sacramento, CA  United States  94273-0001

  Humboldt State University

  Arcata, California  United States 
• Authors:
  • Szewczak, Joseph
  • Michael, Morrison
• Publication Date: 2010-12-6

Language

• English

Media Info

• Media Type: Digital/other
• Features: Appendices; Figures; References; Tables;
• Pagination: 120p

Subject/Index Terms

• TRT Terms: Acoustic signal processing; Birds; Sound
• Geographic Terms: California
• Subject Areas: Environment; Highways; I15: Environment;

Filing Info

• Accession Number: 01506425
• Record Type: Publication
• Source Agency: UC Berkeley Transportation Library
• Report/Paper Numbers: CA04-0661
• Contract Numbers: 65A0184
• Files: CALTRANS, TRIS, STATEDOT
• Created Date: Feb 3 2014 9:20AM