Computer-derived image compositing

Abstract

A computer-derived composite image is obtained of the details contained in successive photomicrographs taken by optical sectioning of the collagen networks in the pulmonary interalveolar wall. Three-dimensional information is therefore contained in a two-dimensional portrayal, providing a basis for computer digital analysis otherwise not readily available.

Extended depth-of-field and 3D information extraction in digital holographic microscopy

“…As our second purpose is to perform a 3D reconstruction, we mostly make use of algorithms that retain one slice of the stack (one distance of focus) for each pixel. The chosen methods are: (1) sum-modified Laplacian [4], (2) variance [5], (3) mean longitudinal value of intensity weighted with spatial variance, (4) Tenengrad [5], (5) Pieper and Korpel's differential operator [1], (6) gradient [3] and (7) Eltoukhy's linear combination [7]. We also use (8) the software developed by B. Forster et al [10] to compare the previous algorithms with a wavelets-based method, here complex Daubechies wavelets.…”

“…Better results are obtained when calculating the energy contained in each region of the high-pass filtered images [2]. Many methods, like gradient or Tenengrad's method, use the pixels in the neighbourhood of the current one and define a criterion of sharpness [1,[3][4][5][6]. Another suggestion is the reconstruction of the focalized image using, for each pixel, a linear combination of the corresponding pixels taken from each of the partially focalized images [7].…”

On the extended depth of focus algorithms for bright field microscopy

“…This algorithm is similar to an algorithm presented by Wall et al (1981). In both cases the procedure implies taking the differences between the gray levels of pairs on nearby pixels in each optical slice.…”