Shape and Texture Based Classification of Fish Species

Rasmus Larsen; Hildur Ólafsdóttir; Bjarne Kjær Ersbøll

doi:10.1007/978-3-642-02230-2_76

Shape and Texture Based Classification of Fish Species

Rasmus Larsen, Hildur Ólafsdóttir, Bjarne Kjær Ersbøll

Research output: Contribution to journal › Conference article › Research › peer-review

Abstract

In this paper we conduct a case study of ¯sh species classi- fication based on shape and texture. We consider three fish species: cod, haddock, and whiting. We derive shape and texture features from an appearance model of a set of training data. The fish in the training images were manual outlined, and a few features including the eye and backbone contour were also annotated. From these annotations an optimal MDL curve correspondence and a subsequent image registration were derived. We have analyzed a series of shape and texture and combined shape and texture modes of variation for their ability to discriminate between the fish types, as well as conducted a preliminary classfication. In a linear discrimant analysis based on the two best combined modes of variation we obtain a resubstitution rate of 76 %

Original language	English
Book series	Lecture Notes in Computer Science
Volume	5575
Pages (from-to)	745-749
ISSN	0302-9743
DOIs	https://doi.org/10.1007/978-3-642-02230-2_76
Publication status	Published - 2009
Event	16th Scandinavian Conference on Image Analysis (SCIA) - Oslo, Norway Duration: 15 Jun 2009 → 18 Jun 2009 http://www.scia2009.org/

Conference

Conference	16th Scandinavian Conference on Image Analysis (SCIA)
Country/Territory	Norway
City	Oslo
Period	15/06/2009 → 18/06/2009
Internet address	http://www.scia2009.org/

Access to Document

10.1007/978-3-642-02230-2_76

OpenUrl availability

Full text

Cite this

@inproceedings{f7101846fbc645ca843b00dd97b9f404,

title = "Shape and Texture Based Classification of Fish Species",

abstract = "In this paper we conduct a case study of ¯sh species classi- fication based on shape and texture. We consider three fish species: cod, haddock, and whiting. We derive shape and texture features from an appearance model of a set of training data. The fish in the training images were manual outlined, and a few features including the eye and backbone contour were also annotated. From these annotations an optimal MDL curve correspondence and a subsequent image registration were derived. We have analyzed a series of shape and texture and combined shape and texture modes of variation for their ability to discriminate between the fish types, as well as conducted a preliminary classfication. In a linear discrimant analysis based on the two best combined modes of variation we obtain a resubstitution rate of 76 %",

author = "Rasmus Larsen and Hildur {\'O}lafsd{\'o}ttir and Ersb{\o}ll, {Bjarne Kj{\ae}r}",

year = "2009",

doi = "10.1007/978-3-642-02230-2_76",

language = "English",

volume = "5575",

pages = "745--749",

journal = "Lecture Notes in Computer Science",

issn = "0302-9743",

publisher = "Springer",

note = "16th Scandinavian Conference on Image Analysis (SCIA) ; Conference date: 15-06-2009 Through 18-06-2009",

url = "http://www.scia2009.org/",

}

TY - GEN

T1 - Shape and Texture Based Classification of Fish Species

AU - Larsen, Rasmus

AU - Ólafsdóttir, Hildur

AU - Ersbøll, Bjarne Kjær

PY - 2009

Y1 - 2009

N2 - In this paper we conduct a case study of ¯sh species classi- fication based on shape and texture. We consider three fish species: cod, haddock, and whiting. We derive shape and texture features from an appearance model of a set of training data. The fish in the training images were manual outlined, and a few features including the eye and backbone contour were also annotated. From these annotations an optimal MDL curve correspondence and a subsequent image registration were derived. We have analyzed a series of shape and texture and combined shape and texture modes of variation for their ability to discriminate between the fish types, as well as conducted a preliminary classfication. In a linear discrimant analysis based on the two best combined modes of variation we obtain a resubstitution rate of 76 %

AB - In this paper we conduct a case study of ¯sh species classi- fication based on shape and texture. We consider three fish species: cod, haddock, and whiting. We derive shape and texture features from an appearance model of a set of training data. The fish in the training images were manual outlined, and a few features including the eye and backbone contour were also annotated. From these annotations an optimal MDL curve correspondence and a subsequent image registration were derived. We have analyzed a series of shape and texture and combined shape and texture modes of variation for their ability to discriminate between the fish types, as well as conducted a preliminary classfication. In a linear discrimant analysis based on the two best combined modes of variation we obtain a resubstitution rate of 76 %

U2 - 10.1007/978-3-642-02230-2_76

DO - 10.1007/978-3-642-02230-2_76

M3 - Conference article

SN - 0302-9743

VL - 5575

SP - 745

EP - 749

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

T2 - 16th Scandinavian Conference on Image Analysis (SCIA)

Y2 - 15 June 2009 through 18 June 2009

ER -

Shape and Texture Based Classification of Fish Species

Abstract

Conference

Access to Document

OpenUrl availability

Fingerprint

Cite this