You could try labeling with an amine specific probe but at pH 6.5  This way you should only label the N-terminus (since the pKa of the epsilon amine group in lysine is about 10 and the reagent will target NH2.  You could use an Alexa488 succinimidyl ester for example since that should be good for the microscopy.

If you can have the peptide synthesized, you could have the fluorescent probe put wherever you want it.

Does the peptide contain a Cys residue? That could also be labeled specifically.

You could also work without modifying your peptide, to be sure that it binds to the membrane in its physiologic state and that the labeling is not leading to nonphysiologic effects or misinterpretations. Detection could be achieved using an antibody specific to your peptide, with either this primary antibody being labeled (then it could also be an scfv) or in combination with standard commercially available labeled anti-fc antibodies.

hi....

Functionalize your peptide to a terminal acetylene or azide group and then conjugate this peptide with a fluophore like BODPY using Click Chemistry.

Cadavarine coupled with a fluorescent dye would be used to amidate a carboxy group on your peptide (C-terminus or Asp/Glu side chain), but when you try to couple it all your free amino groups can also potentially react unless you protect them first.  Cys or N-terminus suggestions above are good if you can't synthesize the peptide with a functional group. Be aware that large dyes can cause non-specific binding, so you'd want to have a labelled control peptide as well.

You can try to order synthesis of your peptide with N- or C-terminal modification using FLAG sequence consensus DYKDDDDK and to detect binding of your peptide with cells using anti-FLAG antibody conjugate. We did it with 46 amino acid synthetic peptide and it worked very well on human THP-1 cells as well as mouse NIH-3T3 cells, as detected by M2 anti-FLAG monoclonal antibody.