Imagers as Biological Sensors

• Main Authors: Ahmadian, Shaun, Allen, Michael, Coe, Sharon, Graham, Eric, Hamilton, Michael, King, Jaime, Rahimi, Mohammad, Taggart, Michael, Yuen, Eric
• Format: Article in Journal/Newspaper
• Language: English
• Published: eScholarship, University of California 2007
• Subjects: Engineering; Avian Studies
• Online Access: http://www.escholarship.org/uc/item/5rb3s6z1

Researchers at CENS are using imagers in two large classes of investigation: (1) Animal Behavior and Survey studies, and (2) Plant Phenology and Belowground Biological Activity studies. A range of imager types are being used, including stationary cameras wired for power and image output, stationary pan-tilt wired cameras, mobile pan-tilt wired cameras, and wireless Cyclops cameras. Animal observing systems at the James Reserve are both being developed and have continued to reliably produce images for a range of research programs involving bird behavior and herpetological surveys. Avian studies goals include the analysis of nest site selection, microclimatic influences on adult breeding and nesting success, and documenting nest predation. Herpetological Studies include the use of pitfall traps to determine diversity and abundance of local fauna. The key science needs have been to increase the frequency and replicates of data captured and to develop image processing software for automated classification of avian behavior and animal pitfall trap capture. Plant and belowground observing systems at the James Reserve include tower-mounted and belowground mini-rhizotron cameras. Above-ground imaging goals include using image analysis for sensing for plant phenological changes and physiological condition, including the timing of leaf flushes and flowering events and analysis for “greenness” of leaves and of landscapes. Below-ground imaging goals involve the analysis of the large number of rhizotron images currently being collected for changes in root and fungal counts and general biological activity associated with concurrently-measured carbon fluxes. The key science needs have been to allow users to not only browse the vast number of captured data and images, but also to control the operation of the systems and to semi-automatically assist in the analysis of images with a toolbox of image analysis software.