Skip to content

4. Deconvolution

Pradeep Rajasekhar edited this page Sep 2, 2022 · 2 revisions

Deconvolution

To perform deconvolution (Richardson Lucy), you will need the PSFs. We have provided simulated PSFs for the Zeiss lattice lightsheet generated by Zen software here. In the napari-lattice inerface, tick Deconvolution and then click the box Click to Select PSFs for deconvolution. You can choose a multichannel psf image in the first box or choose psfs for each channel in every box. The order of the PSF images/channels must follow the order of the image channels. To accelerate deconvolution, under Choose processing device, select cuda_gpu, which will use pycudadecon library for deconvolution and requires you to have an NVIDIA GPU. The cpu options uses Richardson-Lucy algorithm from scikit-image library. You can also define the number of iterations for running RL deconvolution.

You can couple deconvolution with Cropping and Workflows as well.

Deconvolution

Clone this wiki locally