DeepErase: Weakly Supervised Ink Artifact Removal in Document Text Images

TL;DR

DeepErase is a neural network-based preprocessor that effectively removes ink artifacts from scanned document images, significantly improving OCR accuracy for both printed and handwritten texts, especially on out-of-distribution datasets.

Contribution

The paper introduces DeepErase, a weakly supervised neural approach for ink artifact removal in document images, with a novel data assembly method for training without pixel-level annotations.

Findings

Achieves high segmentation accuracy of ink artifacts.

Boosts OCR recognition accuracy substantially.

Performs well on out-of-distribution datasets like IRS tax forms.

Abstract

Paper-intensive industries like insurance, law, and government have long leveraged optical character recognition (OCR) to automatically transcribe hordes of scanned documents into text strings for downstream processing. Even in 2019, there are still many scanned documents and mail that come into businesses in non-digital format. Text to be extracted from real world documents is often nestled inside rich formatting, such as tabular structures or forms with fill-in-the-blank boxes or underlines whose ink often touches or even strikes through the ink of the text itself. Further, the text region could have random ink smudges or spurious strokes. Such ink artifacts can severely interfere with the performance of recognition algorithms or other downstream processing tasks. In this work, we propose DeepErase, a neural-based preprocessor to erase ink artifacts from text images. We devise a method to programmatically assemble real text images and real artifacts into realistic-looking "dirty" text images, and use them to train an artifact segmentation network in a weakly supervised manner, since pixel-level annotations are automatically obtained during the assembly process. In addition to high segmentation accuracy, we show that our cleansed images achieve a significant boost in recognition accuracy by popular OCR software such as Tesseract 4.0. Finally, we test DeepErase on out-of-distribution datasets (NIST SDB) of scanned IRS tax return forms and achieve double-digit improvements in accuracy. All experiments are performed on both printed and handwritten text. Code for all experiments is available at https://github.com/yikeqicn/DeepErase