Overview of ImageCLEFcoral 2021: Coral reef image annotation of a 3D environment

This paper presents an overview of the ImageCLEFcoral 2021 task that was organised as part of the Conference and Labs of the Evaluation Forum - CLEF Labs 2021. The task addresses the problem of automatically segmenting and labelling a collection of underwater images that can be used in combination to create 3D models for the monitoring of coral reefs. The training dataset contained 882 images from 6 subsets from 4 locations. 1 subset was complete (containing all the images to build the 3D model) and 5 subsets containing a partial collection. The test data (491 images) contained the images required to complete 4 of the partial image sets from each of the 4 locations (the final partial subset is not used for testing, only training). 8 teams registered to the ImageCLEFcoral task, of which 3 teams submitted 8 runs. Participants' entries showed that although automatic annotation of benthic substrates was possible, developing a generic algorithm to work across multiple geographical locations will be difficult due to the variation of characteristics within and between classification types.