Classification and standardization of field images associated with a field in a form

摘要

A system and method for processing form images including strokes. A stroke identification module identifies the position of each stroke in each of form image. A geometry engine identifies a group of overlapping strokes from an overlay of the plurality of form images. The geometry engine generates a field bounding box encompassing the group of strokes and generates a field image from each form image based on the field bounding box. A label detector analyzes an area around the field image in the form image and generates a label image. A field image classifier determines stroke features for field images associated with a field. The field image classifier classifies the field images into one or more groups based on the stroke features. A stroke standardizer determines a representative image for field images associated with a field.

主权项

1. A computer-implemented method for classifying a plurality of field images, the method comprising:
receiving the plurality of field images associated with a field; determining, with one or more processors, stroke features for each of the plurality of field images; analyzing, with the one or more processors, the stroke features to determine one or more patterns; classifying the plurality of field images into one or more groups based on the one or more patterns; retrieving field images in a first group from the one or more groups; determining a representative image from the retrieved field images based on the stroke features; replacing the field images of the first group with the representative image; and generating a table with the plurality of field images including the representative image for the field images that are classified as being part of the first group.