Improving 3D visualisation of image stacks by correction of inhomogeneous sample illumination: application to SIMS imaging

Raw three-dimensional image data recorded by any kind of image detector often suffers from inexact lateral intensity distributions caused by inhomogeneous sample illumination. This can be caused by the optical system or by inhomogeneities of the image detector, e.g, by uneven ageing of an electron multiplier surface. The effect of this image detection deficiency is folding of the image data function with a disturbing background function of the detection hardware which is often gaussian shaped. When the dynamic range of the image data is small in comparison to the variation width of the hardware background function, structural information of the image sets can become hard to identify in three-dimensional volumes. Therefore a new method for automatic calculation of the background function is used to correct the image volumes before visualising them. The algorithm employed is based on recording three pictures (including one vertical and one horizontal shift) prior to recording the three-dimensional image stack and the following solution of the so-called "apparatus elasticity equation" which results in the description of the interfering background. A major advantage of this method is the fact that there is no homogeneous reference sample required and that it is also applicable to uneven and noisy intensity distributions with few features. The efficiency of this method is shown with examples of depth profiles of industrial sample volumes recorded by three-dimensional secondary ion mass spectrometry (SIMS), but is not limited to it; it can be applied to all sorts of microscopic images that suffer from illumination inhomogeneity regardless of its cause.