CUDA-accelerated fast Sauvola’s method on Kepler architecture

Sauvola’s method is one of the top binarization methods for degraded document images. High computational complexity, however, restricts it to non-time-sensitive applications. In this paper, we present a parallel implementation of Sauvola’s method with integral image optimization, called fast Sauvola’s method, on Nvidia Kepler architecture GPUs using CUDA 5.0. Our implementation is evaluated on a GTX 650 graphic card (384 cores) and exhibits an average speedup of about 38 compared to a sequential implementation on a fast CPU, and computational complex of our implementation is constant for any size of local windows.

[1]  Derek Bradley,et al.  Adaptive Thresholding using the Integral Image , 2007, J. Graph. Tools.

[2]  Matti Pietikäinen,et al.  Adaptive document image binarization , 2000, Pattern Recognit..

[3]  Wen-mei W. Hwu,et al.  GPU Computing Gems Jade Edition , 2011 .

[4]  Thomas M. Breuel,et al.  The OCRopus open source OCR system , 2008, Electronic Imaging.

[5]  Franklin C. Crow,et al.  Summed-area tables for texture mapping , 1984, SIGGRAPH.

[6]  Ioannis Pratikakis,et al.  DIBCO 2009: document image binarization contest , 2011, International Journal on Document Analysis and Recognition (IJDAR).

[7]  Zhiyi Yang,et al.  Parallel Image Processing Based on CUDA , 2008, 2008 International Conference on Computer Science and Software Engineering.

[8]  Fatih Murat Porikli,et al.  Fast Construction of Covariance Matrices for Arbitrary Size Image Windows , 2006, 2006 International Conference on Image Processing.

[9]  Kevin Skadron,et al.  Scalable parallel programming , 2008, 2008 IEEE Hot Chips 20 Symposium (HCS).

[10]  B. Kapralos,et al.  I An Introduction to Digital Image Processing , 2022 .

[11]  Paul A. Viola,et al.  Robust Real-Time Face Detection , 2001, International Journal of Computer Vision.

[12]  Ichiro Masaki,et al.  Efficient integral image computation on the GPU , 2010, 2010 IEEE Intelligent Vehicles Symposium.

[13]  Klaus Mueller,et al.  Performance Tuning for CUDA-Accelerated Neighborhood Denoising Filters , 2011 .

[14]  Rahul Sharma,et al.  Parallel Implementation of Souvola’s Binarization Approach on GPU , 2011 .

[15]  Bülent Sankur,et al.  Survey over image thresholding techniques and quantitative performance evaluation , 2004, J. Electronic Imaging.

[16]  C. Messom,et al.  High Precision GPU based Integral Images for Moment Invariant Image Processing Systems , 2008 .

[17]  Thomas M. Breuel,et al.  Efficient implementation of local adaptive thresholding techniques using integral images , 2008, Electronic Imaging.

[18]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[19]  John D. Owens,et al.  GPU Computing , 2008, Proceedings of the IEEE.