B. Bogart wrote:
Hey IOhannes,
did you find the multi-blob tracker you mentioned on list?
well here it is. i haven't yet added it to the CVS, as i think the functionality should be broken into 2 objects: [pix_multiblob] (that detects multiple blobs in one image) and a [blobtracker] object that does the tracking (this is: ensure consistency of indices over time) of feature-vectors.
the [pix_multiblob] is finished and i think i will submit it; the [blobtracker] (which will most likely be an abstraction based on iemmatrix) has bothered me the whole day (and i don't know why ;-))
mfg.adr IOhannes
Hey Great news!
Thanks Johannes, I'm really looking forward to showing this in my Gem workshops. Please keep me posted about blobtracker!
B.
IOhannes m zmoelnig wrote:
B. Bogart wrote:
Hey IOhannes,
did you find the multi-blob tracker you mentioned on list?
well here it is. i haven't yet added it to the CVS, as i think the functionality should be broken into 2 objects: [pix_multiblob] (that detects multiple blobs in one image) and a [blobtracker] object that does the tracking (this is: ensure consistency of indices over time) of feature-vectors.
the [pix_multiblob] is finished and i think i will submit it; the [blobtracker] (which will most likely be an abstraction based on iemmatrix) has bothered me the whole day (and i don't know why ;-))
mfg.adr IOhannes
B. Bogart wrote:
Hey Great news!
Thanks Johannes, I'm really looking forward to showing this in my Gem workshops. Please keep me posted about blobtracker!
ok, i have submitted an abstraction [pix_blobtracker] to Gem/abstractions
this abstraction depends on Gem for the blob-detection ([pix_multiblob]) and on iemmatrix (which is on the pd-cvs cvs.puredata.info:/cvsroot/pure-data/externals/iem/iemmatrix ) for the actual tracking. while the weighted matching algorithm is far from being perfected it seems like it works quite well.
ah, and since there is no help-patch for [pix_multiblob] yet: you can set the minimum size of a blob with the "blobSize" message (ranges from 0 to 1 with a default of 0.1)
mfg.asd.r IOhannes