|
|
|
A Quantitative Comarison to Water Column Correction Techniques in Marine Environments Using High Spatial Resolution Data
Nima Pahlevan
MS.c. student
KNTU,
Iran
Email: n_pahlavan@sina.kntu.ac.ir
ABSTRACT
Mapping of substrate cover types such as Coral Reefs, Algae, Seagrass, etc. in coastal regions has been performed in large scale areas ýusing satellite-based observation since mid 1980's. Although the high spatial resolution data such as IKONOS images has overcome ýsome spatial limitations, low signal-to-noise ratio due to destructive atmospheric effects and intervening water column has lead to a ýproblematic image processing task. In this study, after conducting registration by means of large scale maps and removal of ýatmospheric impacts applying pseudo invariant features (PIFs) on a pan-sharpened IKONOS image, two distinct methods namely, ýBierwirth law (first method) and Depth Invariant Indices (DII) (second method), have been utilized for elimination of water column ýeffects. Efficiency of the two methods has been evaluated by comparison of the generated thematic maps with the field observations. ýPrior to classification, the image was divided to two zones according to DOP segmentation and the other processing stages were ýidentical during the study. Results of this research have shown that the use of Bierwirth law performed well against DII ýcorresponding to classification outcomes. In other words, Kappa index of first method was 0.73 rather than the 0.68 of second one. ýHowever, these values increased by masking out deep areas (> 4m) to 0.79 and 0.74, respectively. We concluded that both methods ýwere almost poor in deep areas particularly the second one and either in-situ measurements of attenuation coefficients of water ýcolumn or more accurate physical models could improve the level of accuracy.ý
|
|
|