Comparison of manual and semi-automatic underwater imagery analyses for monitoring of benthic hard-bottom organisms at offshore renewable energy installations

Abstract

The construction of new offshore wind farms is one of the strategies to fulfill growing demands for “green” renewable energy. Underwater imagery is an important tool in the environmental monitoring of offshore renewable energy installations, especially in rocky benthic environment where traditional techniques are not applicable. Underwater video from the high energy Norwegian Sea coast was used for this study. Traditional manual point-based benthic cover estimations from selected frames were tested against a semi-automatic approach which involved making mosaic images from underwater videos. The study demonstrates that results of manual and semi-automatic benthic cover estimations are similar, although the manual analysis has a much larger spread in the variability of the data with many outliers due to the limited amount of points used in the analysis. Although the number of benthic features that could be extracted by computer using color is fewer than those that can be detected with the human eye, the described semi-automatic method is less biased and less costly in terms of qualified staff. Implementation of the semi-automatic method does not require any programming skills and has the ability to quickly and simply process larger amount of underwater imagery which would be of decisive advantage to the industry.