A Novel, Remote-Controlled BatCam for Censusing Small Colonies of
Bats.  

Thomas H. Kunz, Jimmy C. Chau, Zheng Wu, Lisa Hong, Jonathan
D. Reichard, Margrit Betke, and Thomas D.C. Little, Boston University,
Boston, MA.


Assessing colony and population size and dynamics are among the most
important challenges that face bat ecologists. We describe a
remote-controlled, infrared BatCam capable of recording nightly
emergence of bats and for transmitting digital data over the Internet
to remote stations or to a local laptop computer. The BatCam consists
of a small, weatherized Internet camera connected to a local video
server and gateway for Internet access. The server enables video
recording a predetermined intervals corresponding to expected nightly
emergence. Internet connectivity allows sampling the real-time video
from off-site. This is particularly important for maintaining
confidence in the quality of the collected video when away from the
site. Using the BatCam, we successfully recorded and analyzed video
data from a small bat colony of little brown myotis that emerged
nightly from a bat house located at Moore State Park, Paxton, MA, in
June 2008. We applied automated computer vision algorithms to detect
and track bats in the video. The bats in each video frame were first
detected by an algorithm using background modeling, in which an
observed image frame was compared with an estimate of the background
images without bats. A second algorithm tracked each newly detected
bat through the camera's field of view.  An estimate of colony size
was determined from the number of bats that were tracked throughout
the video sequence recorded during emergence.  This novel,
remote-controlled BatCam and analytical approach promises to be a
valuable addition to the tool kit needed for assessing seasonal and
interyear changes in colony size and may be especially important for
assessing changes in maternity colonies of bats species affected by
White-Nose Syndrome.