Talk:Condensation algorithm

Robotics Stub‑class Low‑importance

	This article is within the scope of WikiProject Robotics, a collaborative effort to improve the coverage of Robotics on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.RoboticsWikipedia:WikiProject RoboticsTemplate:WikiProject RoboticsRobotics
Stub	This article has been rated as Stub-class on Wikipedia's content assessment scale.
Low	This article has been rated as Low-importance on the project's importance scale.

Wrong description?

The Article says:

One of the most interesting facets of the algorithm is that it does not compute on every pixel of the image. Rather, pixels to process are chosen at random, and only a subset of the pixels end up being processed.

That seems wrong to me. What Sequential Monte Carlo Methods (and the condensation algorithm as a special case of them) do, is that they operate on a subset of possible positions in the state space of the tracked object. However, the state space of the object does not relate directly to the pixels in the image; the state of the object is, as far as my understanding goes, represented by the spline coefficients which describe the momentary shape of the object. To express this in SMC Method / Particle Filter formalism, the camera image is the observation, the spline coefficients represent the state. And the Condensation Algorithm is an application of SMC Methods aka Particle Filters to the domain of optical tracking and image segmentation. (Joise)--82.113.106.22 (talk) 15:20, 30 January 2009 (UTC)[reply]