{"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/sparse-depth-sensing-for-resource-constrained","title":"Sparse Depth Sensing for Resource-Constrained Robots","arxiv_id":"1703.01398","date":"2017-03-04","proceeding":null,"authors":["Fangchang Ma","Luca Carlone","Ulas Ayaz","Sertac Karaman"],"abstract":"We consider the case in which a robot has to navigate in an unknown\nenvironment but does not have enough on-board power or payload to carry a\ntraditional depth sensor (e.g., a 3D lidar) and thus can only acquire a few\n(point-wise) depth measurements. We address the following question: is it\npossible to reconstruct the geometry of an unknown environment using sparse and\nincomplete depth measurements? Reconstruction from incomplete data is not\npossible in general, but when the robot operates in man-made environments, the\ndepth exhibits some regularity (e.g., many planar surfaces with only a few\nedges); we leverage this regularity to infer depth from a small number of\nmeasurements. Our first contribution is a formulation of the depth\nreconstruction problem that bridges robot perception with the compressive\nsensing literature in signal processing. The second contribution includes a set\nof formal results that ascertain the exactness and stability of the depth\nreconstruction in 2D and 3D problems, and completely characterize the geometry\nof the profiles that we can reconstruct. Our third contribution is a set of\npractical algorithms for depth reconstruction: our formulation directly\ntranslates into algorithms for depth estimation based on convex programming. In\nreal-world problems, these convex programs are very large and general-purpose\nsolvers are relatively slow. For this reason, we discuss ad-hoc solvers that\nenable fast depth reconstruction in real problems. The last contribution is an\nextensive experimental evaluation in 2D and 3D problems, including Monte Carlo\nruns on simulated instances and testing on multiple real datasets. Empirical\nresults confirm that the proposed approach ensures accurate depth\nreconstruction, outperforms interpolation-based strategies, and performs well\neven when the assumption of structured environment is violated.","url_abs":"http://arxiv.org/abs/1703.01398v3","url_pdf":"http://arxiv.org/pdf/1703.01398v3.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":"sparse-depth-sensing-for-resource-constrained","repo_url":"https://github.com/sparse-depth-sensing/sparse-depth-sensing","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"compressive-sensing","task_name":"Compressive Sensing"},{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"navigate","task_name":"Navigate"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}