We give a brief introduction to sequential spatial processes. We discuss their definition, formulate a Markov property, and indicate why such processes are natural tools in tackling high level vision problems. We focus on the problem of tracking a variable number of moving objects through a video stream, and discuss the relationship with the popular Hough transform. A list of pointers to the literature concludes the paper.
|Title of host publication||Stereology and Image Analysis - ECS10 (Proceedings 10th European Congress, Milan, Italy, June 22-26, 2009)|
|Place of Publication||Bologna|
|Publisher||Esculapio Pub. Co.|
|Publication status||Published - 2009|
|Name||The MIRIAM Project Series|
Lieshout, van, M. N. M. (2009). Sequential spatial processes for image analysis. In V. Capasso (Ed.), Stereology and Image Analysis - ECS10 (Proceedings 10th European Congress, Milan, Italy, June 22-26, 2009) (The MIRIAM Project Series; Vol. 4). Esculapio Pub. Co..