首页 | 本学科首页   官方微博 | 高级检索  
     


An evaluation of the analysis system of video transects used to sample subtidal epibiota
Authors:Konstantin S. Tkachenko
Affiliation:Institute of Marine Biology, Far East Branch of the Russian Academy of Sciences, Vladivostok, 690041, Russia
Abstract:Two popular methods of benthic cover estimation (the point intercept technique with two sets of position points, and digital interactive color segmentation) were compared with an alternative method of digital cover estimation using Bezier curves as a tool for outlining the objects on an images and AutoCAD® software for the final evaluation of abundance. The comparison was done using still video images obtained from two 10-m transects on subtidal rocks off the Rimsky-Korsakov islands in the Sea of Japan (Russia). Ten rectangular sectors (1×0.4 m each) selected randomly within both video transects were analyzed. One-way ANOVA for repeated measurements was used to test the differences between methods. The point intercept technique differed significantly from both methods of digital estimation and had an essential positive bias. The Tukey multiple comparison test revealed the differences among digital estimation methods in the species, which have the complicated color with many contrast spots. The proposed approach using Bezier curves has an advantage over the interactive color segmentation if the objects under selection are lit at different levels or have a contrast coloration or are hidden by canopy organisms. Besides, estimation of cover in the field of AutoCAD® software is more precise and takes less time than that obtained using a scaled grid, available for automatically segmented species. The results showed that digital cover estimation using Bezier curves and AutoCAD® software is a convenient method for analyzing benthic samples at large spatial scales.
Keywords:Benthos   Digital image analysis   Cover estimation   Video transects   Large spatial scale
本文献已被 ScienceDirect 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号