Optimization of 360-based photogrammetric pipeline for narrow and dark indoor surveys

Abstract

Photogrammetry has been commonly used for surveying architectural structures over the past two decades; however, this technique faces limitations, particularly in complex, narrow, or dark spaces due to optical constraints. Recently, advanced technologies such as 360° cameras have started to influence established photogrammetric protocols. Capturing an entire scene in a single spherical shot offers significant value for the rapid and efficient surveying of indoor spaces, although new methodologies are needed for data management, storage, and computational optimization. Building on previous work conducted within the SESAMES Project, a validated pyramidal scheme highlights the limitations of a single-camera linear sequence and emphasizes the importance of orienting “omnidirectional” sensors. This paper introduces SQUILLIDAE (Spherical acQUisition Instrument for Low-Light and InDoor nArrow spacEs), a cost-efficient and versatile multi-camera system designed to enhance photogrammetric techniques derived from spherical imagery. It details our approach from initial investigations to final design choices and includes a comparative analysis of results, as well as several perspectives for future technical improvements.