Abstract:
Classification in underwater imagesis a challenging task as images are often captured inextreme environmental conditions with poor illumination, hazy background, etc. Oce...Show MoreMetadata
Abstract:
Classification in underwater imagesis a challenging task as images are often captured inextreme environmental conditions with poor illumination, hazy background, etc. Ocean scientists who are involved in such analysis, prefer automatic classification as manual classification is costly and time consuming. Techniques based on intensity information alone may not result in accurate segmentation of underwater objects. Statistical features representing the texture information of the object and background is needed. A set of 14 texture features was computed for underwater images and features like autocorrelation, sum average, sum variance and sum entropy were able to accurately classify object of interest from background. A fuzzy neural network was designed and texture features were trained and tested for classification. The proposed adaptive fuzzy neural network obtained a maximum classification accuracy of 97%.
Published in: 2017 13th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD)
Date of Conference: 29-31 July 2017
Date Added to IEEE Xplore: 25 June 2018
ISBN Information: