System and method for assisting in computer interpretation of surfaces carrying symbols or characters

摘要

The present disclosure relates to image processing and analysis and in particular automatic segmentation of identifiable items in an image, for example the segmentation and identification of characters or symbols in an image. Upon user indication, multiple images of a subject are captured and variations between the images are created using lighting, spectral content, angles and other factors. The images are processed together so that characters and symbols may be recognized from the surface of the image subject.

主权项

1. A method comprising the acts of:
receiving a user request through a user interface of a general purpose computer; capturing, in response to the user request, a first image of a first surface under a first set of lighting conditions, the first surface having three dimensions, wherein aspects of the first and second dimensions define at least two boundaries of the first surface and aspects of the third dimension encode raised character information; storing the first image in a memory resource; capturing, also in response to the user request, a second image of the first surface under a second set of lighting conditions; storing the second image in the memory resource; using a processor resource of the general purpose computer to combine the first image with the second image to form a first combined image by employing differences between the first and second images to enhance the decipherability of the encoded raised character information; and receiving output of a character recognition operation, the output comprising a set of machine readable characters corresponding to the raised character information encoded in the third dimension of the first surface, wherein the differences between the first and second images include differences in the raised character information that result from differences between the first set of lighting conditions and the second set of lighting conditions.