Estimating Norway lobster abundance from deep-water videos: an automatic approach
暂无分享,去创建一个
Video technology has been playing an increasing role in marine science, both for habitat mapping and estimating commercial species abundance. However, when quantification is needed, it is usually based on manual counting, a subjective and time-consuming task. The present work proposes a methodology to automatically quantify the abundance of Norway lobsters, Nephrops norvegicus, by counting lobsters or their burrows from video sequences, as a reliable complement to the currently used operator-based approach. The methodology is validated using a set of test video sequences captured at the Portuguese continental slope, using a monochrome camera mounted on a trawl gear, being characterised by non-uniform illumination, artefacts at image border, noise and marine snow. The analysis includes, after a pre-processing stage, the segmentation of regions of interest and the corresponding classification into one of the three targeted classes: Norway lobsters, burrows and others (including trawl impact marks). The developed software prototype, named IT-IPIMAR N. norvegicus (I2N2), is able to provide an objective, detailed and comprehensive analysis to complement manual evaluation, for lobster and burrow density estimation.
[1] Thomas Serre,et al. Modeling feature sharing between object detection and top-down attention , 2005 .
[2] R. J. A. Atkinson,et al. Population biology of the Norway lobster, Nephrops norvegicus (L.) in the Firth of Clyde, Scotland - I: Growth and density , 1997 .
[3] Jason S. Link,et al. Ecological Considerations in Fisheries Management: When Does it Matter? , 2002 .
[4] Paulo Lobato Correia,et al. Fishery-independent estimation of benthic species density — a novel approach applied to Norway lobster Nephrops norvegicus , 2008 .