Brian Northan

Hi @gselzer it looks to me like both smooth and sharpen are almost exactly the same. Also it would be nice to be able to take advantage of the work...

Hi @gselzer You are right we should make the ```Op``` a ```SliceWise2DConvolve``` and then let each Op do it's own normalization. In the case of the ```findEdges``` it looks like...

Hi @gselzer A few more thoughts after looking more closely at the IJ1 implementations. 1. [Smooth](https://github.com/imagej/imagej1/blob/58fec6733d7797d87bb111a3aa2ba2ce8ed9503b/ij/process/ShortProcessor.java#L660) looks to me like it is just a mean. Or am I missing a...

Hi @gselzer 1. Have you tested to see if ```DefaultMeanFilter``` and ```DefaultSmooth``` produce the same result?? I think the line in question is used for rounding purposes, and the [setReal](...

Hi @dscho, @ctrueden So the first steps for this effort are to create a JTransform version of FFTMethods and then to create interfaces for fft ops, and then ops that...

Hi Curtis fft and ifft should capture all the variations for the image transforms. Each variation hopefully can be determined by the inputs. There have been a couple of recent...

@ctrueden, Below I've documented the steps used in @StephanPreibisch's convolution function (https://github.com/imglib/imglib2-algorithm-gpl/blob/master/src/main/java/net/imglib2/algorithm/fft2/FFTConvolution.java see run() ) The FFT ops (https://github.com/imagej/imagej-ops/pull/76) were designed with this workflow in mind... 1. Calculate the desired...

@ctrueden, @dscho, @dietzc, @StephanPreibisch I updated the "milestones" above to reflect recent work and what needs to be done next. I can take a look at SymmetricSeparableConvolvolution next. I also...

Hi Christian The convolver ops generally have two implementations, a higher level one using `Img` and a lower level one that uses `RandomAccessibleInterval`. The high level op uses the `ImgFactory`...

Hi Christian > We have several input options here: If the incoming RandomAccessibleInterval actually is an Img >(instanceof test ... even extendible for other types?!) we can use the same...