Ultrafast sensors and depth cameras are key enablers for imaging through complex geometries, through scattering, and beyond the line of sight. However, despite accelerating advances in imaging electronics and imaging applications, the optics of such cameras have been inherited from conventional low-speed photography cameras. This has limited ultrafast cameras and their applications to the design constraints of conventional optics. Here, we exploit time as an extra dimension in the optical design and demonstrate that by folding large spaces in time using time-resolved cavities, one can enable new camera capabilities without losing the targeted information. We demonstrate lens tube compression by an order of magnitude, together with ultrafast multi-zoom imaging and ultrafast multispectral imaging by time-folding the optical path at different regions of the imaging optics. Considering the vast variety of designs that could emerge by time-folding conventional imaging optics, we expect this technique to have a broad impact on time-resolved imaging and depth-sensing optics.

超快传感器和深度相机是实现复杂几何图形，散射和超出视线范围成像的关键因素。然而，尽管在成像电子学和成像应用方面加速发展，但是这种照相机的光学器件已经从传统的低速摄影照相机继承而来。这将超快相机及其应用限制在常规光学器件的设计约束中。在这里，我们将时间作为光学设计的一个额外维度，并证明了通过使用时间分辨型腔及时折叠大空间，可以启用新的相机功能而不会丢失目标信息。我们展示了镜筒压缩一个数量级，通过在成像光学器件的不同区域对光路进行时间折叠，将其与超快多缩放成像和超快多光谱成像一起使用。考虑到时间折叠式常规成像光学器件可能会出现各种各样的设计，我们希望该技术对时间分辨成像和深度感测光学器件产生广泛的影响。