Fast autofocusing in digital holography using scaled holograms

Digitally-recorded holograms are numerically reconstructed to reveal the 3D profile of the object. When the distance of the object to the recording plane is not known a priori, one has to perform reconstruction at discrete steps and decide the best focus distance according to some criteria. There are autofocusing algorithms that utilize a metric to assess the quality of the images, such as the sharpness of the object, to decide the best focus distance. However, the computational cost of this search rapidly increases as the resolution of the system becomes higher since the object image has to be numerically reconstructed from the hologram first. Here, we investigate the behavior of the sharpness curves of the object images obtained from the scaled holograms. We find that it is possible to estimate the autofocus distance accurately with this scaling, and the resultant autofocus speed improvement is on the order of square of the scaling ratio. Experimental results are presented and they are in excellent agreement with the simulations.