• What is deconvolution?
• PSF, Noise, Edges, Speed
• Ops Deconvolution, current + future
• Future goals
• Preliminary Results

• A procedure used to reverse convolution
• Convolution (Blurring)
• Deconvolution (deblurring)
• improves measurements

Describes response of imaging system to a point like object

• Need a point spread function to deconvolve
• Measured (image of subresolution object)
• Theoretical Calculation
• COSMOS
• Written in c++/c
• Open!
• PSF Generator
• Widefield PSF (Icy)

Images also contain noise

• Convolution attenuates high frequencies
• Deconvolution restores high frenquencies
• Deconvolution can amplify noise

• Solve the equation in the frequency domain

• Use an Iterative approach
• Richardson Lucy

• Dey, et al. 2006
• IOCBIOS
• Python implementation
• Open!
• Deconvolution Lab
• Fiji Plugin
• Closed source

• Boundary Conditions
• Non-circulant deconvolution BIG Lab

• Algorithm acceleration
• Take a bigger step at each iteration
• Vector Acceleration (Biggs)
• Hardware acceleration
• Fast math libraries
• Multi-threading
• GPU

Currently implemented in Ops (beta)

• Richardson Lucy
• Richardson Lucy Total Variation
• Vector Acceleration
• Non-circulant edge handling

• Optimize memory use and speed
• Diffraction based PSF
• Integrate Icy Algorithms
• GPU
• Phantoms for testing

IJ Macro to crop (to center) and reverse PSF

run("Make Substack...", "  slices=28-155");
run("Reverse");
						

# @OpService ops
# @UIService ui
# @Dataset image
# @Dataset psf

from net.imglib2.meta import ImgPlus

deconvolved=ops.deconvolve().richardsonLucyTV(None, image, psf, [0,0,0], None, \
None, None, None, None, None, 100, 0.0001, True, True);

ui.show("deconvolved", ImgPlus(deconvolved));	

						

Left to right: Original, ImageJ-Ops-RLTV

Left to right: Original, ImageJ-Ops-RLTV