Postal Address Block Location Using a Convolutional Locator Network
Generate an AI Snapshot to get a quick, structured summary of this paper.
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
The use of a convolutional neural network to perform address block location on machine-printed mail pieces and a simple set of rules was used to generate ABL candidates from the network output.
Abstract
This paper demonstrates that using a CLN to find abstract features of an object, rather than locating the entire object, provides a reasonable amount of insensitivity to the shape and scale of the object. In particular, the completely identified address blocks in the final test set had aspect ratios which ranged from 1.3 to 6.1 and their absolute X and Y dimensions both varied over a 3:1 range. They contained anywhere from 2 to 6 lines of text