rigid body with missing points
Stephen Gatesy
We're running into issues with a bone that has four markers implanted (nos. 7, 8, 9, and 10).
When all four markers are digitized, the motion looks correct. It also looks okay when a marker other than the first (7) is missing. So we have good alignment using 7,8, and 10, for example.
But when we lose marker 7, no transforms are calculated. Given that this is our most anterior marker on the knee, it goes out of sight very often and we lose many frames this way. Moreover, the bone has a dramatically bad alignment once all four markers are once again in view. The bone appears to be knocked out of alignment by the missing marker and then can't regain its original relationship to the centroids.
I wonder if there is a check going on that only looks at the first marker for each bone's cluster. Could this be rejecting frames with markers 8,9, and 10 that would otherwise be useful? Can we fix this so that any frame with the 3 marker minimum is calculated?
Of greater concern, can anyone see why a span of missing first markers might throw off subsequent reconstructions within the same sequence? As I understand it, the solutions are for going frame to frame, right? Are markers free to come and go as long as there are 3? Is the first marker in some way special? I thought we were trying to treat them all equally.
Thanks,
steve
Stephen Gatesy
Stephen...@brown.edu
Department of Ecology & Evolutionary Biology
Box G-B209, 171 Meeting Street
Brown University
Providence, RI 02912
401-863-3770 (office)
401-863-9169 (lab)