Image-based character recognition

摘要

Various embodiments enable a computing device to capture multiple images (or video) of text and provide at least a portion of the same to a recognizer to separately recognize text from each image. Each of the recognized outputs will typically include one or more text strings for each image. Substrings common to each of the one or more text strings are computed and compared to each text string within each image to determine an alignment consensus for each substring within the text. A template string is generated that includes each common substring in a position corresponding to a determined alignment for a respective substring. A character frequency vote is then applied to unresolved portions and the final text string is determined by filling the unresolved spaces with the character having the highest occurrence rate for a respective space.

主权项

1. A computer-implemented method performed by a portable computing device, comprising:
obtaining a first image of text at a first angle of view relative to the text using a camera of the portable computing device; obtaining at least one second image of the text at a second angle of view relative to the text; processing the first image and the at least one second image using a character recognition algorithm; identifying a respective string of characters and a respective bounding box for the respective string of characters associated with the text of the first image and the at least one second image; determining, using a longest common substring algorithm, a substring that is common to the first image and the at least one second image; determining, for the substring, a respective first position of the substring within the respective string of characters of the first image and the at least one second image; generating a template string that includes the substring, wherein a second position of the substring within the template string is based, at least in part, on the respective first position of the substring within the respective string of characters of the first image and the at least one second image; determining one or more missing characters in the template string by analyzing occurrences of characters in the respective string of characters of the first image and the at least one second image; and generating a merged string for display on the portable computing device, the merged string including the template string and the one or more missing characters.