Producing, capturing and using visual identification tags for moving objects

摘要

A visual Reference tag is formed from an array of cells, wherein each cell is visually distinct from all other cells in a given neighborhood and each of said cells contains a single visual cue, for example a unique color, selected from a finite number of visual cues.

主权项

1. A method for decoding one or more visual reference tags from a digitized image, each reference tag comprising an array of cells, said method comprising steps of: finding a central core of each tag using multi-scale color histograms, performing local white balancing for each tag, recovering the extent and the content of each tag, determining each tag's color sequence along an encoding path, and decoding integer numbers represented by the color sequence for each tag.