Student: Irene Williams

Supervisor: Dr. Joe Leach

Abstract: The industrial revolution has produced a wealth of technology, enabling marine researchers among others to gather multitudes of data. In this quest for knowledge, the techniques used for data collection have become as vastly studied as the natural phenomena themselves. Sampling equipment for almost any phenomenon is available, ranging from the use of space-borne satellites for global events, to high-resolution photography (microscopy) for microscopic detail. Therefore, the increase in available technology has greatly increased the potential areas of study of the marine environment.

The versatility and rapid advancement of this technology has, in many cases, resulted in intensive data collection without adequate use of this data. The collection of oceanic data is an extremely expensive exercise, with the fiscal costs including shipping and staff, plus the cost of the sampling equipment itself. Therefore, it is imperative that the use of existing data is extended, and that more versatile data be collected in the future.

This research concentrates on video technology and investigates the potential use of various systems. Research has included data originally intended for both qualitative and quantitative analysis. Tasks involved extending the use of qualitative data for semi-quantitative and quantitative use, and comparing the results to data from quantitative collection systems. Research into data collection systems will lead to a saving in time and money for the capture of marine research data.

Biography: Irene Williams
Email: irene@sunrise.sli.unimelb.edu.au

More research information:

	More information on single video imagery
	More information on stereo video imagery

	Single Video Data Single video data has been provided by both CSIRO & MAFRI. The CSIRO data is a forward looking oblique source, with depths to 200 metres. MAFRI data is a near-vertical source to 90 metres depth. To the left is an example of an Along Track Video (ATV) stereo pair (from MAFRI data). The pairs come from sequential frames of the video track rather than side-by-side cameras. Research has noted many limitations of this stereo-setup, predominantly involving motion between frames. The image base separation can't be calculated accurately, also changes in height and movement of objects within view hamper stereo measurements.

Stereo Video Data

The Department of Geomatics currently has two stereo video capture platforms operational. The following data was captured at around 3 metres depth and shows a traditional stereo pair. The side overlap is shown, then a forward overlap is added to this for stereo along a video transect.