Modified Sauvola binarization for degraded document images

Kaur, Amandeep; Rani, Usha; Josan, Gurpreet Singh

doi:10.1016/j.engappai.2020.103672

Cited by 15 publications

(7 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Its main purpose is to select the appropriate threshold to divide the image into two parts: background and default. In Figures 12 and 13, we show a flaw segmentation results using Sauvola method applied to PM, SSAD, LVAD and the proposed anisotropic diffusion method (Kaur et al, 2020; Sauvola and Pietikäinen, 2000).…”

Section: Resultsmentioning

confidence: 99%

Defect identification in magnetic tile images using an improved nonlinear diffusion method

Gharsallah

Braiek

2021

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

Visual inspection of surface defects is a crucial step in the magnetic tile manufacturing process. Magnetic tile images suffer from a non-uniform illumination, texture and noise that disperse irregularly in flawless image areas. As a result, common edge detection and threshold segmentation techniques fail to identify these kinds of defects. In this work, we present a robust algorithm for defect identification in magnetic tile images. The proposed method is based on a new anisotropic diffusion filtering model. Unlike traditional anisotropic diffusion models that take into account only gradient magnitude information, the proposed model combines together gradient magnitude and a new local difference image feature. The aim is to remove bright shapes and undesirable artifacts in the faultless region in magnetic tile images. In addition, the method activates a smoothing process in the flawless region to homogenize the background and simultaneously a sharpening in the defect boundaries to highlight anomalies. Experimental results on a number of magnetic tiles samples containing different types of defects have demonstrated the efficiency of the proposed diffusion method.

show abstract

Section: Resultsmentioning

confidence: 99%

Defect identification in magnetic tile images using an improved nonlinear diffusion method

Gharsallah

Braiek

2021

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

show abstract

“…It is well known that heuristic binarization methods with hyper-parameters could rarely achieve their upper-bound performance unless the method hyperparameters are individually tuned for each input document image [12], and this is also the main pain point of Sauvola approach.…”

Section: Related Sauvola Approachesmentioning

confidence: 99%

“…Many efforts have been made to mitigate this pain point. For example, [14] introduces a multi-grid Sauvola variant that analyzes multiple scales in the recursive way; [13] proposes a hyper-parameter free multi-scale binarization solution called Sauvola MS [2] by combining Sauvola results of a fixed set of window sizes, each with its own empirical k and r values; [8] improves the classic Sauvola by using contrast information obtained from pre-processing to refine Sauvola 's binarization; [12] estimates the required window size w in Sauvola by using the stroke width transform matrix. Table 1 compares these approaches with the proposed SauvolaNet , and it is clear that only SauvolaNet is end-to-end trainable.…”

Section: Related Sauvola Approachesmentioning

confidence: 99%

“…Postproc? Sauvola [28] user specified single [14] auto user specified auto multiple [13] fixed multiple [8] fixed single [12] auto fixed fixed multiple SauvolaNet auto learned learned multiple…”

Section: Related Sauvola Approachesmentioning

confidence: 99%

“…As mentioned in many works [13,12], one disadvantage when using Sauvola algorithm is the tuning of hyper-parameters. Among all three hyperparameters, namely, the window size w, the degradation level k, and the input deviation r, w is the most important.…”

Section: Pixelwise Window Attentionmentioning

confidence: 99%

See 2 more Smart Citations

SauvolaNet: Learning Adaptive Sauvola Network for Degraded Document Binarization

Deng¹,

Wu²,

Zhou³

2021

Preprint

View full text Add to dashboard Cite

Inspired by the classic Sauvola local image thresholding approach, we systematically study it from the deep neural network (DNN) perspective and propose a new solution called SauvolaNet for degraded document binarization (DDB). It is composed of three explainable modules, namely, Multi-Window Sauvola (MWS), Pixelwise Window Attention (PWA), and Adaptive Sauolva Threshold (AST). The MWS module honestly reflects the classic Sauvola but with trainable parameters and multi-window settings. The PWA module estimates the preferred window sizes for each pixel location. The AST module further consolidates the outputs from MWS and PWA and predicts the final adaptive threshold for each pixel location. As a result, SauvolaNet becomes end-to-end trainable and significantly reduces the number of required network parameters to 40K -it is only 1% of MobileNetV2. In the meantime, it achieves the State-of-The-Art (SoTA) performance for the DDB task -SauvolaNet is at least comparable to, if not better than, SoTA binarization solutions in our extensive studies on the 13 public document binarization datasets. Our source code is available at https://github.com/Leedeng/SauvolaNet.

show abstract

SauvolaNet: Learning Adaptive Sauvola Network for Degraded Document Binarization

Deng¹,

Zhou³

2021

Document Analysis and Recognition – ICDAR 2021

View full text Add to dashboard Cite

Modified Sauvola binarization for degraded document images

Cited by 15 publications

References 29 publications

Defect identification in magnetic tile images using an improved nonlinear diffusion method

Defect identification in magnetic tile images using an improved nonlinear diffusion method

SauvolaNet: Learning Adaptive Sauvola Network for Degraded Document Binarization

SauvolaNet: Learning Adaptive Sauvola Network for Degraded Document Binarization

Contact Info

Product

Resources

About