A stochastic model for the detection of coherent motion

A computational model is presented for the detection of coherent motion based on template matching and hidden Markov models. The premise of this approach is that the growth in detection sensitivity is greater for coherent motion of structured forms than for random coherent motion. In this preliminary study, a recent experiment was simulated with the model and the results are shown to be in agreement with the above premise. This model can be extended to be part of a more complex and elaborate computational visual system.

• Type: Journal
• Year: 2004
• Where: BC
• Authors: Jason Lee, Willy Wong
• Keywords: BC 2004, Coherent Motion, Computational Model, Markov Models