Experiments with the Full Fish4Knowledge Dataset

Research output: Chapter in Book/Report/Conference proceedingChapter


This chapter presents some of the experiments possible when using a very large ecological observation dataset, such as the Fish4Knowledge dataset. The dataset was acquired over 1000 days, observing 12 h a day using 9 undersea cameras at 3 locations. 23 different species were recognized. Each day’s observations vary considerably, but analysis of the large dataset allows trends to be observed. Key results are (1) that there is only little variation in fish observation through the daylight hours, (2) that typhoons only temporarily disrupt the abundance measures, and (3) different habitats show different ratios of species.
Original languageEnglish
Title of host publicationFish4Knowledge: Collecting and Analyzing Massive Coral Reef Fish Video Data
EditorsB. Robert Fisher, Yun-Heh Chen-Burger, Daniela Giordano, Lynda Hardman, Fang-Pang Lin
Place of PublicationCham
PublisherSpringer International Publishing
Number of pages21
ISBN (Electronic)978-3-319-30208-9
ISBN (Print)978-3-319-30206-5
Publication statusPublished - 2016

Publication series

NameIntelligent Systems Reference Library
PublisherSpringer International Publishing
ISSN (Print)1868-4394


Dive into the research topics of 'Experiments with the Full Fish4Knowledge Dataset'. Together they form a unique fingerprint.

Cite this