These systems (reflectors) are not exactly omnidirectional (at least not vertically) because it is impossible to film everything that is above the mirror, and for this reason cameras must be placed high up inside the location.

By knowing the mirror shape, it was possible to correct the distortion and reconstruct the original scene. The first 360° devices were made from a traditional video camera filming a scene reflected on a mirror (usually of paraboloid shape). The proposed algorithm was tested on the CVIP360 dataset, and empirical results demonstrate that the estimation error is negligible for distancing applications. The proposed distance estimation algorithm is based on geometry facts regarding the acquisition process of the omnidirectional device, and is uncalibrated in practice: the only required parameter is the camera height. The CVIP360 dataset includes 16 videos acquired outdoors and indoors, annotated by adding information about the pedestrians in the scene (bounding boxes) and the distances to the camera of some points in the 3D world by using markers at fixed and known intervals. This paper presents two new contributions to the research community: the CVIP360 dataset, an annotated dataset of 360° videos for distancing applications, and a new method to estimate the distances of objects in a scene from a single 360° image. Omnidirectional (or 360°) cameras are acquisition devices that, in the next few years, could have a big impact on video surveillance applications, research, and industry, as they can record a spherical view of a whole environment from every perspective.

Garmin virb edit system requirements