Digital image processing (2nd ed.)
Digital image processing (2nd ed.)
On Image Analysis by the Methods of Moments
IEEE Transactions on Pattern Analysis and Machine Intelligence
Three-Dimensional Shape Analysis Using Moments and Fourier Descriptors
IEEE Transactions on Pattern Analysis and Machine Intelligence
Pattern Spectrum and Multiscale Shape Representation
IEEE Transactions on Pattern Analysis and Machine Intelligence
Self-organization and associative memory: 3rd edition
Self-organization and associative memory: 3rd edition
Introduction to statistical pattern recognition (2nd ed.)
Introduction to statistical pattern recognition (2nd ed.)
The revised Fundamental Theorem of Moment Invariants
IEEE Transactions on Pattern Analysis and Machine Intelligence
IEEE Transactions on Pattern Analysis and Machine Intelligence
Image Analysis and Mathematical Morphology
Image Analysis and Mathematical Morphology
Morphological residues and a general framework for image filtering and segmentation
EURASIP Journal on Applied Signal Processing
Global Image Feature Extraction Using Slope Pattern Spectra
ICIAR '08 Proceedings of the 5th international conference on Image Analysis and Recognition
Adaptive feature selection for classification of microscope images
WILF'05 Proceedings of the 6th international conference on Fuzzy Logic and Applications
Granulometries and Opening Trees
Fundamenta Informaticae
| Hi-index | 0.01 |
Plankton form the base of the food chain in the ocean and arefundamental to marine ecosystem dynamics. The rapid mapping ofplankton abundance together with taxonomic and size compositionis very important for ocean environmental research, but difficultor impossible to accomplish using traditional techniques. In thispaper, we present a new pattern recognition system to classifylarge numbers of plankton images detected in real time by theVideo Plankton Recorder (VPR), a towed underwater videomicroscope system. The difficulty of such classification iscompounded because: 1) underwater images are typically verynoisy, 2) many plankton objects are in partial occlusion, 3) theobjects are deformable and 4) images are projection variant,i.e., the images are video records of three-dimensional objectsin arbitrary positions and orientations. Our approach combinestraditional invariant moment features and Fourier boundarydescriptors with gray-scale morphological granulometries to forma feature vector capturing both shape and texture information ofplankton images. With an improved learning vector quantizationnetwork classifier, we achieve 95% classification accuracy onsix plankton taxa taken from nearly 2,000 images. This result iscomparable with what a trained biologist can achieve by usingconventional manual techniques, making possible for the firsttime a fully automated, at sea-approach to real-time mapping ofplankton populations.