Teepe's profile - activity

overview network karma followed questions activity

2013-07-24 03:36:23 -0600	commented answer	Track rain cells The blobtrack_sample code already use Kalman filter (parameter btpp_name==Kalman) Thank for the reference about the color issue. I look in the src of blobtrackingauto.cpp and I found these lines concerning the deletion of blob : `if(r.width > 4 && r.height > 4 && r.x < w && r.y < h && r.x >=0 && r.y >=0 && r.x+r.width < w && r.y+r.height < h && area > 0) { aver = cvSum(cvGetSubRect(pFG,&mat,r)).val[0] / area; /* if mask in blob area exists then its blob OK/ if(aver > 0.1255)Good = 1; }` So I think i could modify the if (...) to avoid my blob to jump to far but I would rather prefer an other solution (I don't feel very comfortable ... (more)
2013-07-23 06:24:38 -0600	received badge	● Editor (source)
2013-07-23 04:50:00 -0600	asked a question	Track rain cells Hello, I try to use opencv to track rain cell displacement. The cells move slowly, have lot of holes and can disappear. I found blobtrack_sample.cpp and the result is not bad (see fig) but I would like to improve it a little bit. I identified two problem, which i presume can be solved if i change some parameters. First problem : sometimes the ellipse jumps far from the rain cell (especially when rain cell move out from the frame) and try to catch an other rain cell. Second problem : the background obviously do not change but blob detection is still not very good. How can I do to detect a blob as soon as the pixel is not white ? I presume that I have to modify some parameters in blobtrack.hpp and blobtrackingauto.cpp... Thank a lot for any help.