Shape and Texture Based Classification of Fish Species.

Conference Paper · January 2009with14 Reads
Source: DBLP
Conference: Image Analysis, 16th Scandinavian Conference, SCIA 2009, Oslo, Norway, June 15-18, 2009. Proceedings
    • "C. Spampinato et al. [10] classifies 360 images of ten different species and achieves an average accuracy of about 92%. R. Larsen et al. [11] classify three fish species and achieve a recognition rate of 76%. In contrast, this paper investigates novel techniques to perform effective live fish recognition in an unrestricted natural environment. "
    [Show abstract] [Hide abstract] ABSTRACT: Live fish recognition in the open sea is a challenging multi-class classification task. We propose a novel method to recognize fish in an unrestricted natural environment recorded by underwater cameras. This method extracts 66 types of features, which are a combination of color, shape and texture properties from different parts of the fish and reduce the feature dimensions with forward sequential feature selection (FSFS) procedure. The selected features of the FSFS are used by an SVM. We present a Balance-Guaranteed Optimized Tree (BGOT) to control the error accumulation in hierarchical classification and, therefore, achieve better performance. A BGOT of 10 fish species is automatically constructed using the inter-class similarities and a heuristic method. The proposed BGOT-based hierarchical classification method achieves about 4% better accuracy compared to state-of-the-art techniques on a live fish image dataset.
    Full-text · Conference Paper · Nov 2012 · Ecological Informatics
  • [Show abstract] [Hide abstract] ABSTRACT: A research tool is developed to support marine biologists’ research by providing means of analyzing long term and continous fish monitoring video content. Such analysis can be used for instance to discover ecological phenomena such as changes in fish abundance and species composition over time and area, as well as patterns in these changes. Two characteristics sets our system apart from traditional ecological data collecting and processing methods, e.g., with divers. First, the continuous video recording results in an enormous amounts of monitoring data, currently around a year of video recording (containing 4 million fish) have been processed. Second, different from traditional manual recording and analyzing the ecological data, the whole recording, analyzing and result presenting pipeline within our system is automated. On the one hand, it saves effort of manually examining every video footage which is unfeasible. On the other hand, no automatic video analysis methods is perfect, so the user interface provides marine biologists with multiple options to verify the data. Marine biologist can look at the underlying videos, looking a different certainty level and running different automatic video analysis software to verify the tools finding. This research tool enables marine biologists for the first time to analyise long-term and continous underwater videos recording.
    Full-text · Article · Jan 2013
  • [Show abstract] [Hide abstract] ABSTRACT: Live fish recognition in the open sea is a challenging multi-class classification task. We propose a hierarchical classification approach to recognize live fish from underwater videos. However, the hierarchical method accumulates misclassified samples into deeper layers and these accumulated errors reduce the average accuracy. We propose a set of heuristics to help construct more accurate hierarchical trees and, therefore, control the error accumulation. We create an automatically generated tree based on these heuristics and compare it to a baseline tree on a live fish image dataset. The proposed hierarchical classification method achieves about 4% better accuracy compared to state-of-the-art techniques.
    Full-text · Article · Jan 2016 · Ecological Informatics