{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/collaborative-dense-slam","title":"Collaborative Dense SLAM","arxiv_id":"1811.07632","date":"2018-11-19","proceeding":null,"authors":["Louis Gallagher","John B. McDonald"],"abstract":"In this paper, we present a new system for live collaborative dense surface\nreconstruction. Cooperative robotics, multi participant augmented reality and\nhuman-robot interaction are all examples of situations where collaborative\nmapping can be leveraged for greater agent autonomy. Our system builds on\nElasticFusion to allow a number of cameras starting with unknown initial\nrelative positions to maintain local maps utilising the original algorithm.\nCarrying out visual place recognition across these local maps the system can\nidentify when two maps overlap in space, providing an inter-map constraint from\nwhich the system can derive the relative poses of the two maps. Using these\nresulting pose constraints, our system performs map merging, allowing multiple\ncameras to fuse their measurements into a single shared reconstruction. The\nadvantage of this approach is that it avoids replication of structures\nsubsequent to loop closures, where multiple cameras traverse the same regions\nof the environment. Furthermore, it allows cameras to directly exploit and\nupdate regions of the environment previously mapped by other cameras within the\nsystem. We provide both quantitative and qualitative analyses using the\nsynthetic ICL-NUIM dataset and the real-world Freiburg dataset including the\nimpact of multi-camera mapping on surface reconstruction accuracy, camera pose\nestimation accuracy and overall processing time. We also include qualitative\nresults in the form of sample reconstructions of room sized environments with\nup to 3 cameras undergoing intersecting and loopy trajectories.","url_abs":"http://arxiv.org/abs/1811.07632v2","url_pdf":"http://arxiv.org/pdf/1811.07632v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"collaborative-dense-slam","repo_url":"https://github.com/robotvisionmu/densemonoslam","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"camera-pose-estimation","task_name":"Camera Pose Estimation"},{"task_slug":"pose-estimation","task_name":"Pose Estimation"},{"task_slug":"surface-reconstruction","task_name":"Surface Reconstruction"},{"task_slug":"visual-place-recognition","task_name":"Visual Place Recognition"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}