Dear all, I am quite a newbie to this area. I am not sure about what is the following algorithm: Suppose I am having an moving object, with a bunch of cameras, each equipped with signal processing capability.The signal processing unit will track/predict the motion of that object, something like a motion estimation. Is this kind of "motion estimation" algorithm already exist and well developed? Is this a good research topic? Anybody give me some ideas? Thanks a lot, -Walala
what is this algorithm for object tracking?
Started by ●March 20, 2004
Reply by ●March 20, 20042004-03-20
On Sat, 20 Mar 2004 00:11:14 -0500, "walala" <mizhael@yahoo.com> wrote:>Dear all, > >I am quite a newbie to this area. I am not sure about what is the following >algorithm: > >Suppose I am having an moving object, with a bunch of cameras, each equipped >with signal processing capability.The signal processing unit will >track/predict the motion of that object, something like a motion estimation. > >Is this kind of "motion estimation" algorithm already exist and well >developed? Is this a good research topic? Anybody give me some ideas?This already exists, our company sold a product that did exactly what you describe. That particular one needed only one reference camera. I've also seen a solution from another company that we may do business with that uses several reference cameras and tracks objects in three dimensions. I don't know much about the details, but it's certainly out there and used in commercial products right now. -- Be seeing you.
Reply by ●March 20, 20042004-03-20
"walala" <mizhael@yahoo.com> wrote in message news:c3gjpb$74a$1@mozo.cc.purdue.edu...> Dear all, > > I am quite a newbie to this area. I am not sure about what is thefollowing> algorithm: > > Suppose I am having an moving object, with a bunch of cameras, eachequipped> with signal processing capability.The signal processing unit will > track/predict the motion of that object, something like a motionestimation.> > Is this kind of "motion estimation" algorithm already exist and well > developed? Is this a good research topic? Anybody give me some ideas?Kalman tracking is one technique that is often used. There is also the 'Condensation' algorithm. Leon
Reply by ●March 20, 20042004-03-20
On Sat, 20 Mar 2004 00:11:14 -0500, "walala" <mizhael@yahoo.com> wrote:>Suppose I am having an moving object, with a bunch of cameras, each equipped >with signal processing capability.The signal processing unit will >track/predict the motion of that object, something like a motion estimation.A rather old example of tracking using Fuzzy logic (1992) can be found in: http://www-2.cs.cmu.edu/afs/cs/project/ai-repository/ai/areas/fuzzy/systems/tracker/0.html "This directory contains a simulation of target tracking by a fuzzy expert system. The fuzzy system is compared in real time with the results obtained from a Kalman-filter controller from which it was automatically generated using Dr. Bart Kosko's DCL technique." HTH Geo
Reply by ●March 20, 20042004-03-20
"Thore Karlsen" <sid@6581.com> wrote in message news:9pln505bsb1pqe1kbuhh16jpo71mt7qjek@4ax.com...> On Sat, 20 Mar 2004 00:11:14 -0500, "walala" <mizhael@yahoo.com> wrote: > > >Dear all, > > > >I am quite a newbie to this area. I am not sure about what is thefollowing> >algorithm: > > > >Suppose I am having an moving object, with a bunch of cameras, eachequipped> >with signal processing capability.The signal processing unit will > >track/predict the motion of that object, something like a motionestimation.> > > >Is this kind of "motion estimation" algorithm already exist and well > >developed? Is this a good research topic? Anybody give me some ideas? > > This already exists, our company sold a product that did exactly what > you describe. That particular one needed only one reference camera. I've > also seen a solution from another company that we may do business with > that uses several reference cameras and tracks objects in three > dimensions. > > I don't know much about the details, but it's certainly out there and > used in commercial products right now. > > -- > Be seeing you.Hi Thore, Thanks a lot for your information. That's interesting. You mentioned two companies. Would you mind letting me know the companies name so I can browse their products on Internet? Thanks a lot, -Walala
Reply by ●March 22, 20042004-03-22
walala wrote:> Dear all, > > I am quite a newbie to this area. I am not sure about what is the > following algorithm: > > Suppose I am having an moving object, with a bunch of cameras, each > equipped with signal processing capability.The signal processing unit > will track/predict the motion of that object, something like a motion > estimation. > > Is this kind of "motion estimation" algorithm already exist and well > developed? Is this a good research topic? Anybody give me some ideas? >This kind of thing has existed in Radar four years. You might want to look at Kalman filters and its variants (e.g. Extended KF, Unscented KF) as well as Particle Filters, alpha-beta-gamma trackers.... Carlos
Reply by ●March 22, 20042004-03-22
Carlos wrote:> walala wrote: >> Dear all, >> >> I am quite a newbie to this area. I am not sure about what is the >> following algorithm: >> >> Suppose I am having an moving object, with a bunch of cameras, each >> equipped with signal processing capability.The signal processing unit >> will track/predict the motion of that object, something like a motion >> estimation. >> >> Is this kind of "motion estimation" algorithm already exist and well >> developed? Is this a good research topic? Anybody give me some ideas? >> > > This kind of thing has existed in Radar four years. You might wantOoops!! I meant "for years". My brain's turning to mush :-) Carlos
Reply by ●March 22, 20042004-03-22
Carlos wrote:> ... Unscented KF ...Is that a typo? If not, what scents are commonly available? Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●March 22, 20042004-03-22
Jerry Avins wrote:> Carlos wrote: > >> ... Unscented KF ... > > Is that a typo? If not, what scents are commonly available? >Lol. The Unscented Kalman Filter is kosher, but I haven't tried it out yet. See e.g.: http://cslu.cse.ogi.edu/nsel/ukf/test.html Ta, Carlos
Reply by ●March 22, 20042004-03-22
Carlos wrote:> Jerry Avins wrote: > >>Carlos wrote: >> >> >>> ... Unscented KF ... >> >>Is that a typo? If not, what scents are commonly available? >> > > > Lol. The Unscented Kalman Filter is kosher, but I haven't tried it out yet. > See e.g.: > http://cslu.cse.ogi.edu/nsel/ukf/test.html > > Ta, > CarlosAbsolutely amazing! Have you any idea what the name refers to? Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
