Signature-guided character recognition

摘要

Where the recognition of small characters (e.g., text, numbers or symbols) expressed in substantially large images is desired, the recognition process may be facilitated by identifying a signature or a pattern of marked identifiers (e.g., bar codes) within the image, and determining where such characters are typically located in relation to the signature or pattern of identifiers. Because the recognition of characters within images typically occupies a substantial amount of a computer's processing capacity, focusing a recognition technique on portions where such characters are frequently located within an image that includes the signature or pattern, and not on the entire image, the time required in order to process an image in order to recognize such characters may be markedly reduced.

主权项

1. A character recognition system comprising:
a digital camera; and a computing device in communication with the digital camera, wherein the computing device is configured to at least:
capture at least one image of at least two faces of an object using the digital camera;identify a plurality of identifiers within the at least one image, wherein the plurality of identifiers comprises a first identifier provided on a first face of the object and a second identifier provided on a second face of the object;determine a position of the first identifier on the first face of the object based at least in part on the at least one image;determine a position of the second identifier on the second face of the object based at least in part on the at least one image;determine an orientation of the first identifier on the first face of the object based at least in part on the at least one image;determine an orientation of the second identifier on the second face of the object based at least in part on the at least one image;define a signature based at least in part on the position of the first identifier on the first face, the orientation of the first identifier on the first face, the position of the second identifier on the second face and the orientation of the second identifier on the second face within the at least one image;determine whether information regarding a position of recognizable characters associated with the defined signature is included in a data store; andupon determining that information regarding the position of recognizable characters on at least one of the at least two faces of the object associated with the defined signature is included in the data store,
extract a portion of the at least one image corresponding to the position of recognizable characters on at least one of the at least two faces of the object associated with the defined signature; andevaluate the portion of the at least one image according to a character recognition analysis.