MRT letter: Optimal composite depth function for 3D shape recovery of microscopic objects

Generally, in shape from focus techniques, a single focus measure is used in estimating the three-dimensional structure of microscopic objects. However, the performance of a single focus measure is limited to estimate accurately the depth map of diverse type of objects. To cope with this problem, we propose genetic programming based novel approach by developing an optimal composite depth (OCD) function for accurate depth estimation. This OCD function optimally combines the initial depth and focus information extracted from individual focus measures. An improved performance of this function is reported for synthetic and real world microscopic objects.