Image Analysis Techniques for the Study of Fish School Structure from Acoustic Survey Data

A new image analysis based processing system is described for acoustic fisheries data. In a series of a simple steps the system can identify schools and display their positions on a screen. Modifications to the identification process can be entered interactively by the operator. The system provides a wide range of parameters associated with each school, e.g., height, width, depth, shape, density, that can be entered into a database. Applications of this type of system for improving the accuracy of acoustic stock assessment, studying fish distribution patterns, and elucidating fish schooling behaviour are discussed.