JKay

Hello guys, I'm a student from OSPP and glad to find this issue . As a part of my project, I have already implemented the Chan-Vese Segmentation which is a...

@Dale-Black We can chat on Slack :) My id: Yifan Chen

Different from my previous work: [https://github.com/JKay0327/Chan-Vese](https://github.com/JKay0327/Chan-Vese), I try to modified the vectorized codes by using `for` cycle. What's more, I find that there appears a lot of repeat calculations when...

For 3D image, I take image `mri` in `TestImages` as an example, here is the demo: ```julia julia> using Images, TestImages julia> mri = testimage("mri") julia> mosaicview(mri, ncol=9) ``` ...

This is a summary for works we've done in this PR: * Implement function `chan_vese`. We use `for` loop to solve the pixel wide problem, so that we got a...

A list to do before merge: - [ ] Add tests for `init_level_set` and 3D inputs. - [ ] RGB images should be available. But the algorithm for colorant images...

>Any reason not to do https://github.com/JuliaImages/ImageSegmentation.jl/pull/84/files#r716011493? If there is, I'm fine with it, but worth checking whether it's an oversight. It is shown that this will cause a nearly `100ms`'s...

In my test, I've already change the code into: ```julia # using kernelfactor initial_level_set(sz) = broadcast(*, kernelfactors(ntuple(i -> sin.((pi / 5) * (0:sz[i]-1)), length(sz)))...) ... julia> @btime chan_vese(img_gray, μ=0.25, λ₁=1.0,...