Great thanks.

I tried the procedure below for applying point cloud matching to morphologies.

I used morphologies from neuronmorpho.org and used neurons from the lobula complex of Drosophila (http://flybrain.mrc-lmb.cam.ac.uk/si/nbl...).

1. Preprocessing all files using: [vaa3d executable] -x blastneuron -f pre_processing -p "#i [input SWC file] # o [output SWC file] #s 2 #r 0"
2. Point Cloud Matching: [vaa3d executable] -x blastneuron -f inverse_projection -p "#t [preprocessed target SWC file] #s [preprocessed subject SWC file] #o [output file] "

This procedure worked well for most pairs. Some pairs however, it did not work well. For  example, when Gad1-F-000062 was target, the point cloud matched version of Cha-F-200000 was worse than the preprocessed version. I have attached the results in a zip file.

Since the functions used above are slightly different from those of the code mentioned in the paper (http://home.penglab.com/proj/blastneuron...), is this procedure correct for point cloud matching of morphologies with blastneuron?

Sincerely
Ajay