About this paper
This paper describes DepthPano, the depth-image format used in the Metrological Research Foundation's spatial data architecture for indoor terrestrial laser scanning in residential and small-commercial settings. The format is presented as an engineering description: what is implemented, how it works, what trade-offs were accepted, and what file-size and query-performance results are achieved in operational use.
The paper is not a theoretical contribution. Mapping 3D point cloud data onto 2D image grids is a well-established technique in computer vision; the MPEG V-PCC codec, the differential PNG approach of Arikawa et al. (2017), and the equirectangular comparison work of Dumić et al. (2022) are among the published works that have explored variations of the same architectural idea. The Foundation's DepthPano implementation is one application of this architectural family, with engineering choices made for the specific operational context of NDIS-funded residential accessibility assessment in Australia.
The contribution of this paper is the documented implementation: the encoding details, the integration with the rest of the Foundation's bundle architecture, the operational results, and the direct comparison with the published alternatives. Readers looking for a novel compression algorithm or a new spatial data structure should consult the cited prior work; readers looking for a working format applied at scale in a specific domain may find the present description useful.
The format is published in the public domain under open access terms. The Foundation has no patent claims on the format itself. Adopters are welcome to use, modify, or extend the format without restriction; the paper is offered as documentation rather than as a basis for licensing.
