Centre of Applied Marine Biotechnology logo or CAMBio for short
 
  Funding Contact
Home>Research>Aquaculture and Fisheries>Current Projects>Project 2
 

Title

Machine vision for the identification of shellfish larvae.

Objectives

Morphological identification of bivalve larvae to the species level is difficult because of their size, similar shape, similar colour and because of the absence of protruding parts to aid in the identification. Consequently the process is both time-consuming and requires the availability of personnel with considerable experience in the field. The application of computational pattern recognition methods to the automated identification and size analysis of scallop larvae is being investigated in this project.

Machine Vision

Preliminary studies based on the shape of scallop and non-scallop larvae and using binary invariant moments i.e. features that are invariant to shift (position within the image), to scale (induced either by growth or differential image magnification) and to rotation indicated some potential for automatic identification. This project will focus on enlargement of the data set, the implementation and evaluation of several classification techniques, and the development of a robust segmentation technique based on Markov random field models, which combine spatial contiguity with colour/grey level similarity to overcome difficulties where some parts of the object are less optically dense than the background.

icon Duration

2005-2008

icon Partners/Collaborators

Dept. of Computer Science, Royal Holloway, Univ. of London

icon Funding Agency

Higher Education Authority Technological Sector Research Programme - Strand 1

Contact for Further Details

Jonathan Campbell – Lecturer
Email: jonathan.campbell@lyit.ie

John Slater – Principal Investigator
Email: john.slater@lyit.ie

Christopher Hudy - Postgraduate researcher
Email: christopher.hudy@lyit.ie

+ Publications/Presentations

2006 Hudy, C., Campbell, J. & Slater, J.
Contour based methods for segmentation of low contrast images.
In Derek Molloy, Ovidiu Ghita, and Robert Sadleir, (Editors).
Proceedings IMVIP 2006, Irish Machine Vision and Image Processing Conference 2006, pages 138–145, Dublin City University, Dublin, Ireland, August 2006.

2005 Ramos, V., Campbell, J., Slater, J., Gillespie, J., Bendezu, I. and Murtagh, F.
Swarming around shellfish larvae.
In Proceedings 2nd World Congress on Lateral Computing, pages 97-104, Bangalore, India, December 2005. Springer-Verlag, LNCS Series.

2005 Hudy, C., Campbell, J. & Slater, J.
Segmentation methods for the identification of shellfish larvae.
Presented at IT&T Annual Conference, Carrigaline, Co. Cork 25th – 26th October 2005.

2005 Campbell, J., Slater, J., Gillespie, J., Bendezu, I., & Murtagh, F.
Pattern recognition methods for identification of shellfish larvae.
Proceedings of the Irish Machine Vision and Image Processing Conference (IMVIP 2005), QUB, Belfast, Northern Ireland, U.K., 30th – 31st August 2005.

2005 Hudy, C., Campbell, J. & Slater, J.
Segmentation Methods for Optical Microscopy Images: A Practical Survey.
Proceedings of the IMVIP. Queens University Belfast, Northern Ireland, U.K., 30th - 31st August 2005.

2004 Campbell, J.C., Slater, J.W., Gillespie, J., Bendezu, I.F. & Murtagh, F.
The identification of bivalve larvae using computational pattern recognition methods.
Presented at TecNet Marine Research Workshop, G.M I.T. 24-25th Mar 2004.

2003 Campbell, J.C., Slater, J.W., Gillespie, J., Bendezu, I.F. & Murtagh, F.
Pattern recognition methods for the identification of bivalve larvae.
Poster at BioNet Conference, Galway, Ireland. Dec. 2003.

2003 Slater, J.W.
Forecasting the scallop spatfall using morphological identification of scallop larvae.
Presented at 14th Int. Pectinid Conference, Florida, USA. 23-29th April 2003.
 
icon
  
© CAMBio. All Rights Reserved.
  
icon