Computer Vision, Image Processing, Object Tracking, Estimating Object, Bayes Network, Particle FIlter
Article Human opinion dynamics: An inspiration to solve complex opti...
It looks a bit obscure. Perhaps you could summarise the technique and tell people here what it is about in terms that the estimation theory community will understand. It is not clear from the abstract of the paper that this is a tracking technique.
To improve the prediction of particle filter by using Human opinion dynamic algorithm in order to track the object effectively. The Human opinion dynamic algorithm or CODO (Continuous Opinion Dynamic Optimizer) as proposed by the authors is based on social influence. the social influence has been formulated by considering two factors example distance between two individuals and the social ranking of the individuals
Hello. How does human opinion relate to video image tracking? In general your prediction is only as good as your model. if you want to improve the prediction performance you need to improve the latter.
In the particle filter, there is a problem called the sample impoverishment, to solve that problem can use HOD algorithm. [Sampling Process]
Yeah ! that is human opinion, by using the same factor on the particles in the particle filter, we can solve the impoverishment problem. I got promising results for certain data set in image processing.
Re sampling of the particles are being done by using that CODO and rest same particle filter is being used. Error rate has reduced a little but computational cost is little high when compared to the regular particle filter.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing
- Image Processing