Convolutional neural network-based regression for depth prediction in digital holography
This work addresses a specific bottleneck in digital holography reconstruction, offering a faster alternative to traditional depth estimation techniques.
The study tackled the problem of needing prior depth knowledge for object reconstruction in digital holography by proposing a CNN-based regression method that directly predicts depth positions from holograms with millimeter precision, eliminating the time-consuming depth search required in previous methods.
Digital holography enables us to reconstruct objects in three-dimensional space from holograms captured by an imaging device. For the reconstruction, we need to know the depth position of the recoded object in advance. In this study, we propose depth prediction using convolutional neural network (CNN)-based regression. In the previous researches, the depth of an object was estimated through reconstructed images at different depth positions from a hologram using a certain metric that indicates the most focused depth position; however, such a depth search is time-consuming. The CNN of the proposed method can directly predict the depth position with millimeter precision from holograms.