Please join us for the following AOS Seminar on Thursday, May 29 at 4:00PM in Spiess Hall Room 330.
First Speaker: Eric Orenstein, Scripps Institute of Oceanography
Title: Automated Classification of camouflaging cuttlefish
Abstract: An increasing number of visually-based animal-behavior experiments result in vast amounts of image and video data. Manually sifting through this information is difficult, time consuming, and prone to error -- often requiring specially trained personnel. Cuttlefish camouflaging studies, in particular, generate thousands of images that demand careful attention for correct classification. The present work proposes and validates a method to automate the classification process using texture analysis. This approach could be extended to other camouflaging organisms and classification studies.
John Kusters, Director of West Coast Operations, Applied Physical Sciences Corp.
Kevin Cockrell, Engineer, Applied Physical Sciences Corp.
Title: An Environmental Ship Motion Forecasting (ESMF) system
Abstract: Applied Physical Sciences Corp. (APS) is developing a prototype system to demonstrate real-time precise ship motion and wave forecasting for time scales of a few seconds to a few minutes. ESMF is under development as part of the Office of Naval Research (ONR) Seabasing Future Naval Capability (FNC). Seabasing presents a number of challenges for the safe and efficient transfer of materiel between ships at sea, not the least of which is the prediction of ship motion as far as possible into the future. Accurate and timely ship motions forecast depend heavily on the detailed and comprehensive measurement of environmental factors (e.g. waves, wind) in the vicinity. The forecast time is principally limited by the range at which these forcing factors can be measured and subsequently propagated to the ship in time and geo-location. To meet the ESMF requirements the APS system employs a special-purpose, vertically polarized, X-band radar capable of sensing phase resolved ocean backscatter to ranges of 6~7 km. The resultant wave field and forcing functions are then used to drive a real-time Reduced Order Ship Motion (ROM) model and forecasting process.