DIGITAL IMAGE PROCESSING: BINARIZATION

OCR has been an active research area since last few decades. OCR performs the recognition of the text in the scanned document image and converts it into editable form. The OCR process can have several stages like preprocessing, segmentation, recognition and post processing. The preprocessing stage is a crucial stage for the success of OCR, which mainly deals with noise removal. In the present paper, a modified technique for noise removal named as "K-Algorithm" has been proposed, which has two stages as filtering and binarization. The proposed technique shows improvised results in comparison to median filtering technique.

One of the most important and necessary steps
in the process of document analysis and recognition is the
binarization, which allows extracting the foreground from
the background. Several binarization techniques have been
proposed in the literature, but none of themwas reliable for all
image types. This makes the selection of one method to apply
in a given application very difficult. Thus, performance evaluation
of binarization algorithms becomes therefore vital.
In this paper, we are interested in the evaluation of binarization
techniques for the purpose of retrieving words from
the images of degraded Arabic documents. A new evaluation
methodology is proposed. The proposed evaluation
methodology is based on the comparison of the visual features
extracted from the binarized document images with
ground truth features instead of comparing images between
themselves. The most appropriate thresholding method for
each image is the one for which the visual features of the
identified words in the image are “closer” to the features
of the reference words. The proposed technique was used
here to assess the performances of eleven algorithms based
on different approaches on a collection of real and synthetic
images.

Measurement of cardiac contraction velocity using M-mode ultrasonography with the help of digital image processing has been carried out. Analysis of organ movement rate is of great importance in medical imaging as it relates to diagnosis for diseases. This research is aimed at measuring the rate of heart contraction using digital image processing that employs the active contour segmentation method. The sample used is a video of cardiac check using M-mode USG that lasts 30 seconds and has a frame rate of 30 frames/seconds. This image was then extracted as to get 900 USG images each at 480 x 360 pixels. Measurement of the edges of cardiac images was conducted using MATLAB software. Results show that the cardiac in the video shifted farthest at 73/30 seconds, to 10.34 mm against the x-axis and at 28/30 seconds, to 14.00 mm from the y-axis. Meanwhile, the contraction velocity fluctuated.

In this chapter, a binarization technique specifically designed for historical document images is presented. Existing binarization techniques focus either on finding an appropriate global threshold or adapting a local threshold for each area in order to remove smear, strains, uneven illumination etc. Here, a hybrid approach is presented that first applies a global thresholding technique and, then, identifies the image areas that are more likely to still contain noise. Each of these areas is re-processed separately to achieve better quality of binarization. Evaluation results are presented that compare our technique with existing ones and indicate that the proposed approach is effective, combining the advantages of global and local thresholding. Finally, future directions of our research are mentioned.

Our research team enjoyed the privilege of collaborating with Benjamin Sass over a period of several years. We are happy to dedicate this article to him and wish to express our gratitude for what has been both a prodigious and enjoyable experience. The purpose of our joint endeavor has been the introduction of modern techniques from computer science and physics to the realm of Iron Age epigraphy. One of the most important issues addressed during our cooperation was the topic of facsimile creation. Facsimile creation is a necessary preliminary step in the process of deciphering and analyzing ancient inscriptions. Several manual facsimile construction techniques are currently in use: drawing upon collation of the artifact; outlining on transparent paper overlaid on a photograph of the inscription; and computer-aided depiction via software such as Adobe Photoshop, Adobe Illustrator, Gimp or Inkscape (see Summary section below for software web links). Despite their importance for the field of epigraphy, little attention has thus far been devoted to the methodology of facsimile creation (though the recent comprehensive treatment by Parker and Rollston 2016). Recent decades have seen rapid development and consolidation of various computerized image processing algorithms. Among the most basic and popular tasks in this field is the creation of a black-and-white version of a given image, denoted as image binarization (see Fig.1a–b). Often, such a binarized image is used as a first step for further image processing missions, such as Optical Character Recognition (OCR), texts digitization and text analysis tasks. An algorithmic creation of binarizations can therefore be seen as another method of facsimile creation. Furthermore, a relatively new sub-domain of image processing, Historical Imaging and Processing (HIP), specializes in handling antique documents of different types, periods and origins. Accordingly, binarization algorithms stemming from HIP are even more suitable for archaeological purposes.

In this paper, we present a novel approach of document image acquisition with reduced number of measurements, requiring lesser number of sensing elements. Reduction in the number of sensors also directly implies faster acquisition. Our approach is built around the compressed sensing paradigm, in which signals which are sparse can be captured and reconstructed with considerably lesser number of measurements than required by Nyquist sensing. Signal reconstruction algorithms used in compressed sensing are computationally intensive. A brief survey of some of the popular reconstruction algorithms is also given in the paper. In the algorithm proposed in this work, the reconstruction step associated with recovering the binary image from the compressed sensing measurements is pruned to take advantage of the inherent requirement of the binarization of the acquired image. Binarization of captured documents, which is an essential step for post processing is built into the proposed scheme.

This article discusses the quality assessment of binary images. The customary, ground truth based methodology, used in the literature is shown to be problematic due to its subjective nature. Several previously suggested alternatives are surveyed and are also found to be inadequate in certain scenarios. A new approach, quantifying the adherence of a binarization to its document image is proposed and tested using six different measures of accuracy. The measures are evaluated experimentally based on datasets from DIBCO and H-DIBCO competitions, with respect to different kinds of binarization degradations.