Image analysis techniques to accompany a new in situ ichthyoplankton imaging system

G. Tsechpenakis, C. Guigand, R. K. Cowen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Citations (Scopus)

Abstract

We have built a high resolution towed digital imaging system (ISIIS) capable of imaging water volumes sufficient to accurately quantify even rare plankton (e.g. larval fish) in situ. This imaging system produces very high resolution imagery at very high data rates necessitating automated image analysis. As we are interested in the identification and quantification of a large number of organisms, some of which are relatively similar to each other, we are developing an automated system for detection and recognition of organisms of interest using computer vision tools. Our method aims at (i) the detection of multiple regions (organisms) of interest automatically, while filtering out noise and out-of-focus organisms, and (ii) the classification of the detected organisms into pre-defined categories using shape and texture information. For the organisms detection, we use a probabilistic scheme based on image statistics to locate the regions of interest and - based on size and shape constraints - we filter out the noise, i.e., regions that are detected but do not correspond to organisms. For the classification of the detected organisms, we use the Scale Invariant Feature Transform (SIFT) for matching between the detected regions and the organism images in our database ("dictionary").

Original languageEnglish
Title of host publicationOCEANS 2007 - Europe
StatePublished - Nov 28 2007
EventOCEANS 2007 - Europe - Aberdeen, Scotland, United Kingdom
Duration: Jun 18 2007Jun 21 2007

Other

OtherOCEANS 2007 - Europe
CountryUnited Kingdom
CityAberdeen, Scotland
Period6/18/076/21/07

Fingerprint

ichthyoplankton
Image Analysis
Imaging System
image analysis
Imaging systems
Image analysis
Plankton
Glossaries
Shape Constraint
Fish
Computer vision
Digital Imaging
High Resolution Imaging
Textures
Scale Invariant Feature Transform
Statistics
Mathematical transformations
Region of Interest
Imaging techniques
Computer Vision

Keywords

  • Automated analysis for ISIIS
  • Computer vision
  • Detection of organisms in plankton images
  • Organism recognition

ASJC Scopus subject areas

  • Computer Science Applications
  • Oceanography
  • Ocean Engineering
  • Modeling and Simulation

Cite this

Tsechpenakis, G., Guigand, C., & Cowen, R. K. (2007). Image analysis techniques to accompany a new in situ ichthyoplankton imaging system. In OCEANS 2007 - Europe [4302271]

Image analysis techniques to accompany a new in situ ichthyoplankton imaging system. / Tsechpenakis, G.; Guigand, C.; Cowen, R. K.

OCEANS 2007 - Europe. 2007. 4302271.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tsechpenakis, G, Guigand, C & Cowen, RK 2007, Image analysis techniques to accompany a new in situ ichthyoplankton imaging system. in OCEANS 2007 - Europe., 4302271, OCEANS 2007 - Europe, Aberdeen, Scotland, United Kingdom, 6/18/07.
Tsechpenakis G, Guigand C, Cowen RK. Image analysis techniques to accompany a new in situ ichthyoplankton imaging system. In OCEANS 2007 - Europe. 2007. 4302271
Tsechpenakis, G. ; Guigand, C. ; Cowen, R. K. / Image analysis techniques to accompany a new in situ ichthyoplankton imaging system. OCEANS 2007 - Europe. 2007.
@inproceedings{e3c59d63d89d42b4972ac68364c99a59,
title = "Image analysis techniques to accompany a new in situ ichthyoplankton imaging system",
abstract = "We have built a high resolution towed digital imaging system (ISIIS) capable of imaging water volumes sufficient to accurately quantify even rare plankton (e.g. larval fish) in situ. This imaging system produces very high resolution imagery at very high data rates necessitating automated image analysis. As we are interested in the identification and quantification of a large number of organisms, some of which are relatively similar to each other, we are developing an automated system for detection and recognition of organisms of interest using computer vision tools. Our method aims at (i) the detection of multiple regions (organisms) of interest automatically, while filtering out noise and out-of-focus organisms, and (ii) the classification of the detected organisms into pre-defined categories using shape and texture information. For the organisms detection, we use a probabilistic scheme based on image statistics to locate the regions of interest and - based on size and shape constraints - we filter out the noise, i.e., regions that are detected but do not correspond to organisms. For the classification of the detected organisms, we use the Scale Invariant Feature Transform (SIFT) for matching between the detected regions and the organism images in our database ({"}dictionary{"}).",
keywords = "Automated analysis for ISIIS, Computer vision, Detection of organisms in plankton images, Organism recognition",
author = "G. Tsechpenakis and C. Guigand and Cowen, {R. K.}",
year = "2007",
month = "11",
day = "28",
language = "English",
isbn = "1424406358",
booktitle = "OCEANS 2007 - Europe",

}

TY - GEN

T1 - Image analysis techniques to accompany a new in situ ichthyoplankton imaging system

AU - Tsechpenakis, G.

AU - Guigand, C.

AU - Cowen, R. K.

PY - 2007/11/28

Y1 - 2007/11/28

N2 - We have built a high resolution towed digital imaging system (ISIIS) capable of imaging water volumes sufficient to accurately quantify even rare plankton (e.g. larval fish) in situ. This imaging system produces very high resolution imagery at very high data rates necessitating automated image analysis. As we are interested in the identification and quantification of a large number of organisms, some of which are relatively similar to each other, we are developing an automated system for detection and recognition of organisms of interest using computer vision tools. Our method aims at (i) the detection of multiple regions (organisms) of interest automatically, while filtering out noise and out-of-focus organisms, and (ii) the classification of the detected organisms into pre-defined categories using shape and texture information. For the organisms detection, we use a probabilistic scheme based on image statistics to locate the regions of interest and - based on size and shape constraints - we filter out the noise, i.e., regions that are detected but do not correspond to organisms. For the classification of the detected organisms, we use the Scale Invariant Feature Transform (SIFT) for matching between the detected regions and the organism images in our database ("dictionary").

AB - We have built a high resolution towed digital imaging system (ISIIS) capable of imaging water volumes sufficient to accurately quantify even rare plankton (e.g. larval fish) in situ. This imaging system produces very high resolution imagery at very high data rates necessitating automated image analysis. As we are interested in the identification and quantification of a large number of organisms, some of which are relatively similar to each other, we are developing an automated system for detection and recognition of organisms of interest using computer vision tools. Our method aims at (i) the detection of multiple regions (organisms) of interest automatically, while filtering out noise and out-of-focus organisms, and (ii) the classification of the detected organisms into pre-defined categories using shape and texture information. For the organisms detection, we use a probabilistic scheme based on image statistics to locate the regions of interest and - based on size and shape constraints - we filter out the noise, i.e., regions that are detected but do not correspond to organisms. For the classification of the detected organisms, we use the Scale Invariant Feature Transform (SIFT) for matching between the detected regions and the organism images in our database ("dictionary").

KW - Automated analysis for ISIIS

KW - Computer vision

KW - Detection of organisms in plankton images

KW - Organism recognition

UR - http://www.scopus.com/inward/record.url?scp=36348943441&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=36348943441&partnerID=8YFLogxK

M3 - Conference contribution

SN - 1424406358

SN - 9781424406357

BT - OCEANS 2007 - Europe

ER -