Dewarping Document Image By Displacement Flow Estimation with Fully Convolutional Network

TL;DR

This paper introduces a novel FCN-based method for rectifying distorted document images by estimating pixel-wise displacements, significantly improving dewarping accuracy and detail preservation.

Contribution

It presents a new framework combining displacement flow estimation with a Local Smooth Constraint for effective document image rectification.

Findings

Achieved state-of-the-art dewarping performance.

Effectively handles various geometric distortions.

Preserves local details and overall image quality.

Abstract

As camera-based documents are increasingly used, the rectification of distorted document images becomes a need to improve the recognition performance. In this paper, we propose a novel framework for both rectifying distorted document image and removing background finely, by estimating pixel-wise displacements using a fully convolutional network (FCN). The document image is rectified by transformation according to the displacements of pixels. The FCN is trained by regressing displacements of synthesized distorted documents, and to control the smoothness of displacements, we propose a Local Smooth Constraint (LSC) in regularization. Our approach is easy to implement and consumes moderate computing resource. Experiments proved that our approach can dewarp document images effectively under various geometric distortions, and has achieved the state-of-the-art performance in terms of local details and overall effect.