menu top right top corner

menu top

ViS E-Mail

menu bottom



Introduction Background Instruments Approaches Researchers

In-situ marine particle imaging

Our priority application for the marine snow instrument is quantification of vertically migrating algal mats in the oligotrophic North Pacific that transport nitrate from below the thermocline to the surface. This transport flux substantially impacts the cycling of nitrogen and overall primary production rates. Images archived from recent field work wil be used. A proposed application (proposal pending) involves determining the size and abundance of large organic particles within the pervasive benthic nepheloid layer that represent a rich, near-bottom food source and sites of biogenic transformations of organic matter. Detrital particles and zooplankton are expected to dominate the particles in this unique marine habitat. Our priority focus for the ZOOVIS instrument will be to extract euphausiid (and other scatterers) densities, lengths and orientations from images collected in the Knight Inlet fjord system in central British Columbia. Our VPR analysis will focus on the extraction and identification of the copepod Calanus finmarchicus and its primary invertebrate predators from five cruises in the Gulf of Maine that were part of the US GLOBE NW Atlantic Program.

A: Particle image from with a structured lighting, high resolution color video camera system (Pilskaln et al, 1998). B: Image collected with the zooplankton visualization and imaging system (ZOOVIS) designed to collect high resolution images of meso- and macrozooplankton from relatively large volumes of water (102-103 mls) (Benfield et al. 2002). Image contains at least 13 euphausiids. C: Algal mat image from No. Pacific obtained with a towed VPR with an analogue black and white video camera and synchronized strobe (Darkangelo et al. 1996; Villareal et al. 1999). White bar is 1 cm. D: Zooplankton images collected with a VPR. Digitized video fields are scanned by the software for targets (referred to as regions of interest, ROI s) which meet previously defined criteria of brightness, focus and size. ROIs are classified visually into taxonomic groups by examination of each image in a thumbnail browsing program (Benfield et al. 1996). Top row: Calanus finmarchicus copepods; middle row: complete and partial images of euphausiids likely Meganyctiphanes norvegica; and bottom row: Limacina retroversa pteropods. These represent 3 of maybe 20 or 30 possible classes encountered in a study.

left bottom corner right bottom corner
Computer Vision Laboratory University of Massachusetts