I am looking for a reliable, robust 2D “Filipodia Detector”.
I have not tried, but i guess a simple Random Forest approach would not be reliable enough for our data.
(see attached example)
Autocontext might be an option, but a NN-based approach would likely work best.
(ideally with a pre-trained classifier).
(Zoom in 1 shows the filipodia, Zoom in 2 the label in white)
Do you have any recommendations ?
Thanks a lot & Kind regards