Title :
Detection of Trawling Patterns in Seabed Images
Author :
Yu, Xinsheng ; Gong, Dejun ; Tang, Ying ; Wang, Ni
Author_Institution :
Coll. of Marine Geo-Sci., Ocean Univ. of China, Qingdao
Abstract :
A large number of studies have proved that bottom trawling have great impact on benthic habitats. It is desirable to monitor the damaged level of seafloor and the relationships with ecosystem. Recent years, underwater video imagery has been shown a valuable approach to provide detailed information of physical properties of seabed. In this paper, a method to recognizing trawling marks in seabed image was presented, Instead of using techniques of texture process for seabed analysis, geometric feature-based methods have been applied for trawling pattern classification. The proposed method has been tested with real world images and an overall 83% recognition rate has been achieved with leave one out validation method. This offers the potential of helping scientist to process the amount of underwater images for fishery and seabed environment monitoring
Keywords :
feature extraction; oceanographic techniques; pattern classification; seafloor phenomena; benthic habitats; ecosystem; fishery; geometric feature-based methods; physical properties; seabed environment monitoring; seabed images; texture process; trawling pattern classification; trawling patterns detection; underwater video imagery; Ecosystems; Image analysis; Image recognition; Image texture analysis; Monitoring; Pattern analysis; Pattern classification; Pattern recognition; Sea floor; Testing;
Conference_Titel :
OCEANS 2006
Conference_Location :
Boston, MA
Print_ISBN :
1-4244-0114-3
Electronic_ISBN :
1-4244-0115-1
DOI :
10.1109/OCEANS.2006.306901
