Methods and apparatus for robust decoding of glyph address carpets

Abstract

A captured image includes a portion of a set of two-dimensional address codes. The portion of address codes can be decoded to determine a discrete pointer uniquely defining the portion. The captured image is first processed to determine the orientation of the portion, and then decoded based on the orientation to determine the discrete pointer. To determine the orientation of the portion, the portion is first analyzed to determine values at discrete locations within the portion. The values at each location form a matrix of binary data. The values of the matrix are then correlated to determine an orientation of the captured portion of two dimensional address codes. After determining the orientation of the portion, the values can be further analyzed to determine a discrete pointer that identifies the location of the portion within the address space defined by the two-dimensional address codes.

Dependencies are inferred from the claim text.