University of Crete (UOC)
Σχεδίαση, ανάπτυξη και αξιολόγηση ενός απεικονιστικού συστήματος πολλαπλών ανιχνευτών, για τη μελέτη της συμπεριφοράς ιχθύων
Abstract
dc:descriptionNowadays, the study of fish behavior requires more sophisticated computer vision systems to combine multiple replicates, long experimental periods and in detail description of behavioral patterns in discrete steps. In this work, we present the design and application of a multi-camera system, together with the required processing and analysis software, that offers quantitative and qualitative analysis of fish behavioral patterns. It enables the simultaneous monitoring of 9 cameras, under long time recordings, with high time accuracy and minimal frame losses. This novel contribution to aquacultural engineering also includes image filtering, background calculation, object detection and automated recognition of specific behavioral traits. Additionally, the system is capable of extracting specific information from alarms generated by user-defined parameters. The described system in this thesis successfully completed ten successive experiments, monitoring fish activity in tanks, accumulating data from almost one full year of recorded experimental period. All trials performed within those experiments provided new information on fish specific behavioral patterns in aquaculture. Technical and economic performance of the described system highlights its potential to be applied in any behavioral study under laboratory-scale conditions.
Degree
thesis:*- Grantor dc:publisher
- University of Crete (UOC)
- Year dc:date
- 2019
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Papadakis, Vasileios
- Παπαδάκης, Βασίλειος
Subjects
dc:subject × 8Rights
dc:rights- Statement dc:rights
-
- BY
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/45112