Text Binarization Research Articles

METHODS OF DETERMINING THE THRESHOLD VALUE OF IMAGE AS A WAY TO IMPROVE AUGMENTED REALITY

Background. Augmented reality technology is at the initial stage of its development and requires further development of new and improvement of existing methods and algorithms. The most common and reliable approach to the implementation of augmented reality is the marker approach. The definition of the marker is divided into several stages, one of the most important of which is to determine the threshold of the image. Properly chosen algorithm for determining this threshold can significantly increase the efficiency of the algorithm for finding the marker. Objective. The aim of the study is a comparative analysis of threshold algorithms based on the type of information used and evaluation of their effectiveness based on a set of objective indicators of segmentation quality. Comparison is made for images and documents. For an objective comparison of efficiency, we used a combination of 4 quality criteria for form segmentation. Methods. The following performance criteria were used to determine the performance of the threshold determination methods: classification error, boundary determination error, relative foreground error, and non-uniformity region. Two approaches were considered to obtain an average performance score according to the previous criteria. The first approach was the arithmetic averaging of normalized scores obtained by all criteria. Using a specific threshold algorithm for the image, the average value of all criteria was determined, which was an indicator of its segmentation quality. In turn, the sum of these quality indicators for all images was determined as the performance indicator of the algorithm. The second approach used rank averaging, so that for each test image there was a ranking of algorithms of the threshold value from 1 to 16 according to each criterion separately. Then the ranks (not the actual scores) were averaged both by images and by all criteria. Results. A general unified comparative analysis of the performance of 16 algorithms for determining the threshold value of the image was conducted, which makes it possible to improve the functioning of technologies that depend on determining the threshold of images, including the mechanism of augmented reality marker recognition. Conclusions. It has been observed that the clustering-based method, namely the Kittler method, and the entropy-based method, namely the Kapoor method, are the best algorithms for determining thresholds for specialized images. Similarly, the Kittler clustering method and the Sauwol and White local variable methods are the best working algorithms for binarization of text documents.

Text Recognition of Cardboard Pharmaceutical Packages by Utilizing Machine Vision

In this paper, text recognition of variably curved cardboard pharmaceutical packages is studied from the photometric stereo imaging point-of-view with focus on developing a method for binarizing the expiration date and batch code texts. Adaptive filtering, more specifically Wiener filter, is used together with haze removal algorithm with fusion of LoG-edge detected sub-images resulting an Otsu thresholded binary image of expiration date and batch code texts for future analysis. Some results are presented, and they appear to be promising for text binarization. Successful binarization is crucial in text character segmentation and further automatic reading. Furthermore, some new ideas will be presented that will be used in our future research work.

Identification and Recognition of Text and Face Using Image Processing for Visually Impaired

Nowadays, visual information is the most consequential information,so blind people or visually impaired people has this disadvantage as compulsory information about circumventing is not available to them. The objective of this project is to guide visually impaired people with an astute contrivance utilizing an Android Phone. This system predicated on Android technology is designed for them in order to solve the infeasible situation that affects the blind people. Our system presents an automated system which will provide them to read text patterns that are printed on documents. On the text patterns, localization and binarization are done utilizing OCR. The recognize text is converted to an audio and is provided to the user. Consequently, Deals with Character Recognition. The difficulty faced to identify people during conversation is a very big disadvantage for these people in many professional and edifying situations. To increase the accuracy of face recognition smartphones are utilized in these settings, we have an archetype and face recognition are tested for end users. This system uses Smartphone technology to provide audio output to the blind people.

Fourth order indirect diffusion coupled with shock filter and source for text binarization

In this paper, we present a binarization method for degraded document images with contrast variation or low contrast, which couples a selective shock filter and a time-dependent source to a fourth order indirect diffusion process. The role of the selective shock filter is to enhance the edges as far as possible, while the time-dependent source plays an increasingly predominant role in the process of binarization. The fourth order indirect diffusion not only keeps the merit of noise immunity of the direct diffusion counterpart, but also preferably preserves the low-frequency components to avoid blur and dislocation of weak edges. We develop a novel numerical scheme for solving our model effectively and efficiently, by incorporating finite differencing with the parallel splitting up method in the literature. The proposed algorithm is tested on seven publicly available datasets (Document Image Binarization COmpetition (DIBCO) 2009 to 2014 and 2016). The experimental results indicate that our model is very effective for binarization of document images with contrast variation or low contrast, and has mostly the best effect for document images from the DIBCO datasets, compared with four partial differential equation (PDE) models and four benchmark and recent non-PDE methods.

Text area segmentation from document images by novel adaptive thresholding and template matching using texture cues

This paper presents a new perspective of text area segmentation from document images using a novel adaptive thresholding for image enhancement. Using sliding windows, the texture of the enhanced image is matched with that of a fixed training template image containing the typed letters ‘dB.’ The affine-invariant, low-dimensional difference theoretic texture feature set is used for the texture measurement. The distance matrix is binarized using Otsu threshold, and the ‘0’ pixels indicate the text area. One primary contribution of this paper is the novel adaptive thresholding for document image enhancement prior to the extraction of texture cues. The proposed adaptive thresholding mimics the ability of the human eye to iteratively adjust to varying light intensities through iterative gamma correction followed by contrast stretching so that the text becomes well defined against the background clutter. The text blobs so segmented are binarized using Yanowitz and Bruckstein method of text binarization, and the results are applied for evaluation with respect to the ground-truth annotations. We tested our algorithm on the benchmark DIBCO 2009, 2010, 2011, 2012, 2013 document image datasets in comparison with the state of the art. The high precision–recall and F-score values establish the efficiency of our approach.

Gujarati Text Localization, Extraction and Binarization from Images

Gujarati Text Localization, Extraction and Binarization from Images

Unsupervised refinement of color and stroke features for text binarization

Color and strokes are the salient features of text regions in an image. In this work, we use both these features as cues, and introduce a novel energy function to formulate the text binarization problem. The minimum of this energy function corresponds to the optimal binarization. We minimize the energy function with an iterative graph cut based algorithm. Our model is robust to variations in foreground and background as we learn Gaussian mixture models for color and strokes in each iteration of the graph cut. We show results on word images from the challenging ICDAR 2003/2011, born-digital image and street view text datasets, as well as full scene images containing text from ICDAR 2013 datasets, and compare our performance with state-of-the-art methods. Our approach shows significant improvements in performance under a variety of performance measures commonly used to assess text binarization schemes. In addition , our method adapts to diverse document images, like text in videos, handwritten text images.

Adaptive Histogram Analysis for Scene Text Binarization and Recognition

Scene text binarization and recognition is a challenging task due to different appearance of text in clutter background and uneven illumination in natural scene images. In this paper, we present a new method based on adaptive histogram analysis for each sliding window over a word of a text line detected by the text detection method. The histogram analysis works on the basis that intensity values of text pixels in each sliding window have uniform color. The method segments the words based on region growing which studies spacing between words and characters. Then we propose to use existing OCRs such as ABBYY and Tesseract (Google) to recognize the text line at word and character levels to validate the binarization results. The method is compared with well-known global thresholding technique of binarization to show its effectiveness.

A locally adaptive, diffusion based text binarization technique

This research proposes an adaptive modification to a novel diffusion based text binarization technique. This technique uses linear diffusion with a nonlinear source term to achieve a binarizing effect. This simple isotropic process is compared to the state-of-the-art DIBCO contestants and produces remarkable results given the simplicity of the algorithm. Furthermore, the authors show how using a simple discretization scheme allows for the massively parallel implementation of this process.

Hill Climbing Approach for Text Binarization from Videos

Objectives: To develop presents an efficient algorithm for binarization of text from videos. However, it is a challenging and difficult task in image processing, due to their complicated background and non uniform character size. Method/Analysis: This study focuses a novel approach for binarization of text region from videos based on Hill climbing Algorithm. Symmetric filtering and graph cut algorithms are used to refine the obtained clusters. Finally, an optimal clustering selection algorithm is applied to obtain the text region. Findings: The experimental results show that the proposed text binarization technique is robust in text detection with various character size and complicated background. Data sets are taken from the You Tube Video Text (YVT) harvested from YouTube. Application: This proposed binarization technique is applicable to automated licence plate recognition system, especially to identify Indian vehicles licence plates.

English

Text in images contain important contents for information indexing and retrieval, automatic annotation and structuring of images. Hence text extraction is the crucial stage of analyzing the images. The steps involved in text extraction algorithms are detection, localization, binarization, extraction, enhancement, and recognition of text from the image. Text extraction is a very challenging task due to the variations in text size, font, style, orientation and alignment as well as complex background. Several text extraction techniques based on edge detection, connected component analysis, morphological operators, wavelet transform, texture features, neural network etc. have been developed. This paper provides a review of the various techniques suggested by researchers and their comparative analysis in terms of precision rate, recall rate, detection rate etc.

English

Image binarization plays a vital role in text segmentation which is used in OCR application. Binarization of text in degraded images is a challenging task due to the variations in colour, size, and font of the text and the results are often affected by complex backgrounds, different lighting conditions, shadows and reflections. A robust solution to this problem can significantly enhance the accuracy of scene text recognition algorithms leading to a variety of applications such as scene understanding, automatic localization, navigation, and image retrieval. In this paper, we propose a novel method to extract and binarize text from images that contains complex background. We use an Independent Component Analysis (ICA) based technique to map out the text region, which is inherently uniform in nature, while removing shadows, specularity and reflections, which are included in the background. This algorithm works better on images with different degradations. We implement our method on DIBCO dataset then we compare our robust algorithm with state-of-art criteria like binarization based on Otsu method and we can prove that our algorithm will give better results.

A novel approach to text binarization via a diffusion-based model

This paper presents a new approach to document image binarization. The method is based on the dynamic process of diffusion, coupled with a nonlinear Fitzhugh–Nagumo type source term that exhibits binarizing properties. These desirable properties lead to a method that is robust to noise and is able to successfully binarize an input document image. We measure the efficacy of our proposed method against industry standards by two methods, a pixel by pixel comparison with the ground truth image and a standard optical character recognition test. Through these measures we illustrate a progressive method that performs at the highest level in the field.

Text Location and Extraction for Business Cards Using Stroke Width Estimation

Text extraction and binarization are the important pre-processing steps for text recognition. The performance of text binarization strongly related to the accuracy of recognition stage. In our proposed method, the first stage based on line detection and shape feature analysis applied to locate the position of a business card and detect the shape from the complex environment. In the second stage, several local regions contained the possible text components are separated based on the projection histogram. In each local region, the pixels grouped into several connected components based on the connected component labeling and projection histogram. Then, classify each connect component into text region and reject the non-text region based on the feature information analysis such as size of connected component and stroke width estimation.

Color space transformations for analysis and enhancement of ancient degraded manuscripts

In this paper we focus on ancient manuscripts, acquired in the RGB modality, which are degraded by the presence of complex background textures that interfere with the text of interest. Removing these artifacts is not trivial, especially with ancient originals, where they are usually very strong. Rather than applying techniques to just cancel out the interferences, we adopt the point of view of separating, extracting and classifying the various patterns superimposed in the document. We show that representing RGB images in different color spaces can be effective for this goal. In fact, even if the RGB color representation is the most frequently used color space in image processing, it does not maximize the information contents of the image. Thus, in the literature, several color spaces have been developed for analysis tasks, such as object segmentation and edge detection. Some color spaces seem to be particularly suitable to the analysis of degraded documents, allowing for the enhancement of the contents, the improvement of the text readability, the extraction of partially hidden features, and a better performance of thresholding techniques for text binarization. We present and discuss several examples of the successful application of both fixed color spaces and self-adaptive color spaces, based on the decorrelation of the original RGB channels. We also show that even simpler arithmetic operations among the channels can be effective for removing bleed-through, refocusing and improving the contrast of the foreground text, and to recover the original RGB appearance of the enhanced document.

Binarization and cleanup of handwritten text from carbon copy medical form images

This paper presents a methodology for separating handwritten foreground pixels, from background pixels, in carbon copied medical forms. Comparisons between prior and proposed techniques are illustrated. This study involves the analysis of the New York State (NYS) Department of Health (DoH) Pre-Hospital Care Report (PCR) [Western Regional Emergency Medical Services, Bureau of Emergency Medical Services, New York State (NYS) Department of Health (DoH), Prehospital Care Report v4.] which is a standard form used in New York by all Basic and Advanced Life Support pre-hospital health care professionals to document patient status in the emergency environment. The forms suffer from extreme carbon mesh noise, varying handwriting pressure sensitivity issues, and smudging which are further complicated by the writing environment. Extraction of handwriting from these medical forms is a vital step in automating emergency medical health surveillance systems.

A Contour-Based Robust Algorithm for Text Detection in Color Images

Text detection in color images has become an active research area in the past few decades. In this paper, we present a novel approach to accurately detect text in color images possibly with a complex background. The proposed algorithm is based on the combination of connected component and texture feature analysis of unknown text region contours. First, we utilize an elaborate color image edge detection algorithm to extract all possible text edge pixels. Connected component analysis is performed on these edge pixels to detect the external contour and possible internal contours of potential text regions. The gradient and geometrical characteristics of each region contour are carefully examined to construct candidate text regions and classify part non-text regions. Then each candidate text region is verified with texture features derived from wavelet domain. Finally, the Expectation maximization algorithm is introduced to binarize each text region to prepare data for recognition. In contrast to previous approach, our algorithm combines both the efficiency of connected component based method and robustness of texture based analysis. Experimental results show that our proposed algorithm is robust in text detection with respect to different character size, orientation, color and language and can provide reliable text binarization result.