Institute of Information Theory and Automation

Publication details

Convolutional Neural Networks for Direct Text Deblurring

Conference Paper (international conference)

Hradiš M., Kotera Jan, Zemčík P., Šroubek Filip


serial: Proceedings of BMVC 2015

action: The British Machine Vision Conference (BMVC) 2015 /26./, (Swansea, GB, 07.09.2015-10.09.2015)

project(s): GA13-29225S, GA ČR, 938213/2013, GA UK, 7H14004, GA MŠk

keywords: image deblurring, text deblurring, convolutional neural networks, image restoration

preview: Download

abstract (eng):

In this work we address the problem of blind deconvolution and denoising. We focus on restoration of text documents and we show that this type of highly structured data can be successfully restored by a convolutional neural network. The networks are trained to reconstruct high-quality images directly from blurry inputs without assuming any specific blur and noise models. We demonstrate the performance of the convolutional networks on a large set of text documents and on a combination of realistic de-focus and camera shake blur kernels. On this artificial data, the convolutional networks significantly outperform existing blind deconvolution methods, including those optimized for text, in terms of image quality and OCR accuracy. In fact, the networks outperform even state-of-the-art non-blind methods for anything but the lowest noise levels. The approach is validated on real photos taken by various devices.

RIV: JD

Responsible for information: admin
Last modification: 21.12.2012
Institute of Information Theory and Automation