Component

The Fourier transform of an image produces a representation in frequency space: i.e. separated according to spatial frequency (effectively scale). The 2D amplitude map of the different spatial frequencies is symmetrical, and is commonly displayed with low spatial frequencies (large features) in the centre, highest spatial frequencies (small features) at the edges. Fourier filtering involves suppressing or enhancing features in the Fourier domain before carrying out an inverse Fourier transform to obtain a filtered real-space image. ImageJ's _Process > FFT > Bandpass Filter_ implements two common Fourier-filtering functions: 1. filtering for specific sizes of feature in an image by selecting minimum and maximum feature sizes (selecting a radial band of frequencies in Fourier space); 2. filtering out repetitive horizontal or vertical stripes by cutting out a zero-frequency stripe in the orthogonal direction in frequency space. The example image above shows the effect of filtering for 2 feature size ranges: 0-8 pixels, and 8-256 pixels; where the former appears "flattened" or washed-out, and the latter very blurred. The small images displayed to the lower-right of each filtered image correspond to the mask applied to the Fourier transform. Such filtering can be useful prior to global thresholding, for noise suppression, etc.

Fluorescence spectroscopy by image correlation is a technique that allows analysing and characterizing the different molecular dynamics from a sequence of fluorescence images. Many image correlation techniques have been developed for different applications but in particular to study the mechanisms of cell adhesion during migration. These techniques can be used with most imaging modalities: e.g. fluorescence widefield, confocal microscopy, TIRFM. They allow to obtain information such as the density in molecules, diffusion coefficients, the presence of several populations, or the direction and speed of a movement corresponding to active transport when spatial and temporal correlations are taken into account (STICS: Spatio-Temporal Image Correlation Spectroscopy).

This plugin is based on ICS_tools plugin by Fitz Elliott, available here.

Some bugs have been removed, ROI does not need to be squared, fitting is weighted in order to give more weight to the smaller lags (temporal or spatial)

Exemple of use on sample data [fluorescent beads](http://biii.info/node/2577 "Beads") - Select an ROI, start by ICS to get the right PSF size - Then run TICS and select diffusion, or diffusion plus flow model. Remove the first line (autocorrelation) which corresponds to the noise autocorrelation before fitting.

HDF5 is a data format for storing extremely large and complex data collections. This Fiji/ImageJ HDF5 plugin saves and loads 2D - 5D datasets with flexible options.

In Fiji, the plugin is downloadable via update site "HDF5".

Measurement of kymograph generally deals with slanted streaks in kymograph to measure velocity of movements. This is a pretty much manual procedure that needs hands-on work. Kymoquant analyzes kymograph by treating patterns appearing in kymograph as texture: for a specified area in kymograph, it detects the most likely orientation of streaks and outputs this result as velocity. This workflow was created to ease the situation when it is difficult to find major direction of streaks only with eyes.

Rigid registration of time series in 3D. A video tutorial is available (be careful of sounds, the video automatically starts!): [Sample Drift Correction Following 4D Confocal Time-lapse Imaging](http://www.jove.com/video/51086/sample-drift-correction-following-4d-co…)