Advancing Seabed Imaging: Integrated Mosaicking Technologies for Enhanced Marine Exploration

Underwater video capture platforms are used to photograph the seabottom for benthic surveys. Most mapping technologies rely upon external positioning to create maps from this video data, which vastly increases the cost of the platform. This work introduces two pioneering software solutions with minimal external positioning requirements for seabed imagery processing for marine science. The first package, "mosaic-library," is a Python-based tool tailored for creating detailed mosaics from videos with complex 3D motion. It streamlines seabed exploration by offering video processing tools, image registration methods, and mosaicking map generation, essential for bottom feature classification and biodiversity monitoring. The second package, "SledgeStitcher", automates the manual process of analyzing seabed videos obtained from benthic sledge platforms for ecological studies. By converting video and navigation data into geographically tagged bottom map images, it eases the operator's workload in identifying and counting marine organisms like burrows, sea pens, and flatfish, providing accurate comparisons across different studies.