Motion detection and characterization by an excitable membrane: The "bow wave" model

We present a novel method for detecting and characterizing coherent motion in a set of image frames, using a two-dimensional sheet of locally connected neural elements. Externally applied activity leads to traveling activity waves. Activity waves resulting from points or objects in motion have a characteristic "bow wave" shape, which can be used to establish the existence, direction and speed of the object's motion. This mechanism may be used to model motion detection in biological systems, and may help to explain such psychophysical effects as motion pop-out, robustness of motion signals to noise, and the perception of transparent motion. (C) 1999 Elsevier Science B.V. All rights reserved.