Dear Shumoos

there is no problem when you select the threshold perfectly.

Regards

Dear Shumoos Taha Hammadi,

If you want to get any serious reply or help with your matter I suggest you attach at least a .jpg or .bmp out of your recent analysis/analyses to this thread.

It makes no sense to discuss anything about not knowing what your problem really is (specimen: ok: we know about EYE (animal-human?) / Cornea epithelium.... but nothing about methods to get sections(?): paraffin-, cryo-, semithin plastic /resin sections, stain(s) used, method of image acquisition etc, etc.).

Sorry to bother you with my request,  but hopefully you'll answer some of my questions.

Thanks Dr. Muzhir Al-Ani for your answer, actually I tried to work on threshold but I didn't get good result.

Thanks Dr.Wolfgang H. Muss. I'm working on EYE human and I should detect the nerve and cells, but at this stage just on nerve segmentation, and I'II attach to you some picture after enhance it.

I hope you can help me.

Do you use certain pre-processing operations? Your images are very intensity inhomogeneous, so you should use some operations in order to improve they, before the segmentation.

Dear Shumoos Taha Hammadi,

thanks for attaching some images. Nevertheless I am guessing still:

i) fluorescence or histochemical stain / FLUOR / Normal-transmitted light  (e.g.darkfield, reflectance) microscopy?

ii) staining method

iii) specimens: paraffin, cryo, resin sections, whole specs native (these all would be possibilities).

iv) I would second (indeed not knowing about the origin of images, hardware, software for image processing, if any) Luminita Moraru's opinion as above....

I am sorry about not being able to help again.

I think Fiji/ImageJ should work well for this. There's also the Cell Profiler from the Broad Inst. Most of them have pre processing and cell counting pipelines or modules built in.