Character Recognition Software Research Articles

On the security of text-based 3D CAPTCHAs

CAPTCHAs have become a standard security mechanism that are used to deter automated abuse of online services intended for humans. However, many existing CAPTCHA schemes to date have been successfully broken. As such, a number of CAPTCHA developers have explored alternative methods of designing CAPTCHAs. 3D CAPTCHAs is a design alternative that has been proposed to overcome the limitations of traditional CAPTCHAs. These CAPTCHAs are designed to capitalize on the human visual system's natural ability to perceive 3D objects from an image. The underlying security assumption is that it is difficult for a computer program to identify the 3D content. This paper investigates the robustness of text-based 3D CAPTCHAs. In particular, we examine three existing text-based 3D CAPTCHA schemes that are currently deployed on a number of websites. While the direct use of Optical Character Recognition (OCR) software is unable to correctly solve these text-based 3D CAPTCHA challenges, we highlight certain patterns in the 3D CAPTCHAs can be exploited to identify important information within the CAPTCHA. By extracting this information, this paper demonstrates that automated attacks can be used to solve these 3D CAPTCHAs with a high degree of success.

Portable Camera-Based Assistive Text and Product Label Reading From Hand-Held Objects for Blind Persons

We propose a camera-based assistive text reading framework to help blind persons read text labels and product packaging from hand-held objects in their daily lives. To isolate the object from cluttered backgrounds or other surrounding objects in the camera view, we first propose an efficient and effective motion-based method to define a region of interest (ROI) in the video by asking the user to shake the object. This method extracts moving object region by a mixture-of-Gaussians-based background subtraction method. In the extracted ROI, text localization and recognition are conducted to acquire text information. To automatically localize the text regions from the object ROI, we propose a novel text localization algorithm by learning gradient features of stroke orientations and distributions of edge pixels in an Adaboost model. Text characters in the localized text regions are then binarized and recognized by off-the-shelf optical character recognition software. The recognized text codes are output to blind users in speech. Performance of the proposed text localization algorithm is quantitatively evaluated on ICDAR-2003 and ICDAR-2011 Robust Reading Datasets. Experimental results demonstrate that our algorithm achieves the state of the arts. The proof-of-concept prototype is also evaluated on a dataset collected using ten blind persons to evaluate the effectiveness of the system's hardware. We explore user interface issues and assess robustness of the algorithm in extracting and reading text from different objects with complex backgrounds.

English

Document image analysis is the process or techniques used for images of documents to obtain a computer-readable description from pixel data. A document image analysis product is the Optical Character Recognition (OCR) software that recognizes text in a scanned document image. OCR makes it possible for the user to edit or search the document's contents. In this paper we proposed a novel method for identification of Latin text from Devanagari script image document. There are many documents in Devanagari where a single document page may contain English text as well with Devanagari. In bilingual documents two scripts are generally mixed together within a single text line. There are existing methods for recognition of both script but methods lack the ability to recognize multiple scripts mixed within a single text line.

Characteristics of flattening filter free beams at low monitor unit settings

Newer linear accelerators (linacs) have been equipped to deliver flattening filter free (FFF) beams. When FFF beams are used for step-and-shoot intensity-modulated radiotherapy (IMRT), the stability of delivery of small numbers of monitor units (MU) is important. The authors developed automatic measurement techniques to evaluate the stability of the dose profile, dose linearity, and consistency. Here, the authors report the performance of the Artiste™ accelerator (Siemens, Erlangen, Germany) in delivering low-MU FFF beams. A 6 MV flattened beam (6X) with 300 MU/min dose rate and FFF beams of 7 (7XU) and 11 MV (11XU), each with a 500 MU/min dose rate, were measured at 1, 2, 3, 5, 8, 10, and 20 MU settings. For the 2000 MU/min dose rate, the 7 (7XUH) and 11 MV (11XUH) beams were set at 10, 15, 20, 25, and 30 MU because of the limits of the minimum MU settings. Beams with 20 × 20 and 10 × 10 cm(2) field sizes were alternately measured ten times in intensity modulated (IM) mode, with which Siemens linacs regulate beam delivery for step-and-shoot IMRT. The in- and crossplane beam profiles were measured using a Profiler™ Model 1170 (Sun Nuclear Corporation, Melbourne, FL) in multiframe mode. The frames of 20 × 20 cm(2) beams were identified at the off-axis profile. The 6X beam profile was normalized at the central axis. The 7 and 11 MV FFF beam profiles were rescaled to set the dose at the central axis at 145% and 170%, respectively. Point doses were also measured using a Farmer-type ionization chamber and water-equivalent solid phantom to evaluate the linearity and consistency of low-MU beam delivery. The values displayed on the electrometer were recognized with a USB-type camera and read with open-source optical character recognition software. The symmetry measurements of the 6X, 7XU, and 11XU beam profiles were better than 2% for beams ≥ 2 MU and improved with increasing MU. The variations in flatness of FFF beams ≥ 2 MU were ± 5%. The standard deviation of the symmetry and flatness also decreased with increasing MU. The linearity of the 6X beam was ± 1% and ± 2% for the beams of ≥ 5 and ≥ 3 MU, respectively. The 7XU and 11XU beams of ≥ 2 MU showed linearity with ± 2% except the 7XU beam of 8 MU (+2.9%). The profiles of the FFF beams with 2000 and 500 MU/min dose rate were similar. The characteristics of low-MU beams delivered in IM mode were evaluated using an automatic measurement system developed in this study. The authors demonstrated that the profiles of FFF beams of the Artiste™ linac were highly stable, even at low MU. The linearity of dose output was also stable for beams ≥ 2 MU.

Mobile augmented reality applications for library services

PurposeThe purpose of this paper is to introduce mobile augmented reality applications for library uses and next generation library services.Design/methodology/approachExamples are drawn from museum and archives informatics, computer science applied research, and computer vision research as well as original research and development work from the Undergraduate Library at the University of Illinois.FindingsMobile augmented reality uses include augmenting physical book stacks browsing, library navigation, optical character recognition, facial recognition, and building identification mobile software for compelling library experiences.Originality/valueThe paper suggests uses of mobile augmented reality applications in library settings and models a demonstration prototype interface.

The control application and simulation – particle swarm optimization exploration of control application for user intention toward mobile Mandarin learning system

This study introduces the use of fuzzy neural networks with intent to clearly visualize the advantages of the approach. Particle swarm optimization as the underlying optimization tool is discussed. Different models of fuzzy neural networks can be envisioned. It is worth stressing that the underlying concept is general enough and any other formalism of information granules could be considered as well. Over the past decades, the Chinese market has developed rapidly, especially on intelligent robot interaction and even tension leg platform. Consequently, Chinese learning has increased. E-learning has also become another learning mechanism and is already quite popular. In the study, optical character recognition software is developed in the android phone system. A questionnaire is designed as a base technology acceptance model, and, according to the dimensions, to make assumptions. Using a structural equation model analysis framework to analyze the recovery over the questionnaire, the conclusion of this study and the possible direction of future research are finally arrived at.

Developing integrated workflows for the digitisation of herbarium specimens using a modular and scalable approach

Digitisation programmes in many institutes frequently involve disparate and irregular funding, diverse selection criteria and scope, with different members of staff managing and operating the processes. These factors have influenced the decision at the Royal Botanic Garden Edinburgh to develop an integrated workflow for the digitisation of herbarium specimens which is modular and scalable to enable a single overall workflow to be used for all digitisation projects. This integrated workflow is comprised of three principal elements: a specimen workflow, a data workflow and an image workflow.The specimen workflow is strongly linked to curatorial processes which will impact on the prioritisation, selection and preparation of the specimens. The importance of including a conservation element within the digitisation workflow is highlighted. The data workflow includes the concept of three main categories of collection data: label data, curatorial data and supplementary data. It is shown that each category of data has its own properties which influence the timing of data capture within the workflow. Development of software has been carried out for the rapid capture of curatorial data, and optical character recognition (OCR) software is being used to increase the efficiency of capturing label data and supplementary data. The large number and size of the images has necessitated the inclusion of automated systems within the image workflow.

Toward a computer vision-based wayfinding aid for blind persons to access unfamiliar indoor environments

Independent travel is a well known challenge for blind and visually impaired persons. In this paper, we propose a proof-of-concept computer vision-based wayfinding aid for blind people to independently access unfamiliar indoor environments. In order to find different rooms (e.g. an office, a lab, or a bathroom) and other building amenities (e.g. an exit or an elevator), we incorporate object detection with text recognition. First we develop a robust and efficient algorithm to detect doors, elevators, and cabinets based on their general geometric shape, by combining edges and corners. The algorithm is general enough to handle large intra-class variations of objects with different appearances among different indoor environments, as well as small inter-class differences between different objects such as doors and door-like cabinets. Next, in order to distinguish intra-class objects (e.g. an office door from a bathroom door), we extract and recognize text information associated with the detected objects. For text recognition, we first extract text regions from signs with multiple colors and possibly complex backgrounds, and then apply character localization and topological analysis to filter out background interference. The extracted text is recognized using off-the-shelf optical character recognition (OCR) software products. The object type, orientation, location, and text information are presented to the blind traveler as speech.

Data concepts and their relevance for data capture in large scale digitisation of biological collections

Logistically, the data associated with biological collections can be divided into three main categories for digitisation: i) Label Data: the data appearing on the specimen on a label or annotation; ii) Curatorial Data: the data appearing on containers, boxes, cabinets and folders which hold the collections; iii) Supplementary Data: the data held separately from the collections in indices, archives and literature. Each of these categories of data have fundamentally different properties within the digitisation framework which have implications for the data capture process. These properties were assessed in relation to alternative data entry workflows and methodologies to create a more efficient and accurate system of data capture. We see a clear benefit in the prioritisation of curatorial data in the data capture process. These data are often only available at the cabinets, they are in a format suitable for allowing rapid data entry, and they result in an accurate cataloguing of the collections. Finally, the capture of a high resolution digital image enables additional data entry to be separated into multiple sweeps, and optical character recognition (OCR) software can be used to facilitate sorting images for fuller data entry, and giving potential for more automated data entry in the future.

Probabilistic management of OCR data using an RDBMS

The digitization of scanned forms and documents is changing the data sources that enterprises manage. To integrate these new data sources with enterprise data, the current state-of-the-art approach is to convert the images to ASCII text using optical character recognition (OCR) software and then to store the resulting ASCII text in a relational database. The OCR problem is challenging, and so the output of OCR often contains errors. In turn, queries on the output of OCR may fail to retrieve relevant answers. State-of-the-art OCR programs, e.g., the OCR powering Google Books, use a probabilistic model that captures many alternatives during the OCR process. Only when the results of OCR are stored in the database, do these approaches discard the uncertainty. In this work, we propose to retain the probabilistic models produced by OCR process in a relational database management system. A key technical challenge is that the probabilistic data produced by OCR software is very large (a single book blows up to 2GB from 400kB as ASCII). As a result, a baseline solution that integrates these models with an RDBMS is over 1000x slower versus standard text processing for single table select-project queries. However, many applications may have quality-performance needs that are in between these two extremes of ASCII and the complete model output by the OCR software. Thus, we propose a novel approximation scheme called Staccato that allows a user to trade recall for query performance. Additionally, we provide a formal analysis of our scheme's properties, and describe how we integrate our scheme with standard-RDBMS text indexing.

Understanding design research: A bibliometric analysis of Design Studies (1996–2010)

Purpose The purpose of this paper is to investigate the core themes of design research by analysing citations of papers in the journal Design Studies. It also aims to find out the evolution and future trends of design research. Methodology Employing a bibliometrics and network analysis, the paper analyses citations and co-citations from Design Studies. Instead of using the standard analysis method of aggregating author co-citations, this study conducts the analysis at the individual publication level. Research limitations/recommendations Due to the limitation of the Scopus database, this study only managed to extract articles with full citations in Design Studies from 1996 to 2010. Further studies could also include articles from 1979 to 1995 by either using some character recognition software or manually extracting them. Findings The study identifies the core themes centered on design process and design cognition. In addition, it also reveals that the research method protocol analysis has become more popular in recent years among researchers. Originality/value The main contribution of this paper is the use of a network analysis technique to analyse 12 107 citations in the 459 articles published in Design Studies between 1996 and 2010. Unlike previous review papers which relied heavily on the qualitative observations and reflections of the authors, this paper is the first comprehensive quantitative analysis in this field.

Developing Best Practices for Machine Translation Using Google Translate and OCR Terminal

This article describes a process of developing a set of best practices for machine translation at an off-site storage facility. Optical Character Recognition (OCR) software is tested, along with various machine translation programs, to develop a free system to translate non-English materials for staff and users. There is also a brief literature review regarding some reasons why libraries should be concerned with translating materials into other languages.

Enhancing Surveillance for Hepatitis C through Public Health Informatics

Disease surveillance for hepatitis C in the United States is limited by the occult nature of many of these infections, the large volume of cases, and limited public health resources. Through a series of discrete processes, the Massachusetts Department of Public Health modified its surveillance system in an attempt to improve timeliness and completeness of reporting and case follow-up of hepatitis C. These processes included clinician-based reporting, electronic laboratory reporting, deployment of a Web-based disease surveillance system, automated triage of pertinent data, and automated character recognition software for case-report processing. These changes have resulted in an increase in the timeliness of reporting.

A Korean CAPTCHA Study: Defeating OCRs In a New CAPTCHA Context By Using Korean Syllables

Internet is being used for several activities by a great range of users. These activities include communication, e-commerce, education, and entertainment. Users are required to register regarding website in order to enroll web activities. However, registration can be done by automated hacking software. That software make false enrollments which occupy the resources of the website by reducing the performance and efficiency of servers, even stop the entire web service. It is crucial for the websites to have a system which has the capability of differing human users and computer programs in reading images of text. Completely Automated Public Turing Test to Tell Computers and Human Apart (CAPTCHA) is such a defense system against Optical Character Recognition (OCR) software. OCR can be defined as software which work for defeating CAPTCHA images and make countless number of registrations on the websites. This study proposes a new CAPTCHA context that is Korean CAPTCHA by means of the method which is splitting CAPTCHA images into several parts with random rotation values, and drawing random lines on a grid background by using Korean characters only. Lines are in the same color with the CAPTCHA text and they provide a distortion of image with grid background. Experimental results show that Korean CAPTCHA is a more secure and effective CAPTCHA type for Korean users rather than current CAPTCHA types due to the structure of Korean letters and the algorithm we are using: rotation and splitting. In this paper, the algorithm of our method is introduced in detail.

An image-based automatic Arabic translation system

In this paper, we present a system that automatically translates Arabic text embedded in images into English. The system consists of three components: text detection from images, character recognition, and machine translation. We formulate the text detection as a binary classification problem and apply gradient boosting tree (GBT), support vector machine (SVM), and location-based prior knowledge to improve the F1 score of text detection from 78.95% to 87.05%. The detected text images are processed by off-the-shelf optical character recognition (OCR) software. We employ an error correction model to post-process the noisy OCR output, and apply a bigram language model to reduce word segmentation errors. The translation module is tailored with compact data structure for hand-held devices. The experimental results show substantial improvements in both word recognition accuracy and translation quality. For instance, in the experiment of Arabic transparent font, the BLEU score increases from 18.70 to 33.47 with use of the error correction module.

Procedures for measuring women's influence: Data translation and manipulation and related problems

Preparing data from artefact catalogues of previously published German excavation reports, in the project 'Engendering Roman Spaces', required ongoing refinement of data translation and digital manipulation using a variety of software packages. This process included the use of Optical Character Recognition (OCR) software, spreadsheets, database and graphics programs, and the final presentation of the data in ArcGIS as an interpretative tool. With each step, a number of challenges were encountered relating to the quality of the data and original cataloguing processes, and the limitations of the software packages being used. Excavation reports of four Roman military sites - the forts of Vetera I, Ellingen, Oberstimm and Rottweil - are used in this article to highlight the range of problems encountered and solutions arrived to resolve them, a process requiring constant revision and refinement.

Heuristics for identification of bibliographic elements from title pages

This paper presents a methodology for automatic identification of bibliographic data elements from the title pages of books. Also enumerates the various steps like scanning the title pages, running optical character recognition (OCR) software, generating HTML files out of title pages and applying heuristics to identify the bibliographic data elements. Much of the paper deals with the surveys undertaken to analyze the characteristics of various bibliographic descriptive elements like title, author, publisher and other elements. The first survey deals with the sequence of the bibliographic data in the title pages. The second survey deals with the font size, font type and the proximity of each bibliographic element on the title pages. The survey results are then used to develop heuristics, in order to develop a rule‐based expert system which can identify the bibliographic elements on the title pages. The results of the system are presented, along with problems encountered.

A statistical approach to the generation of a database for evaluating OCR software

In this paper we consider a statistical approach to augment a limited database of groundtruth documents for use in evaluation of optical character recognition software. A modified moving-blocks bootstrap procedure is used to construct surrogate documents for this purpose which prove to serve effectively and, in some regards, indistinguishably from groundtruth. The proposed method is validated through a rigorous statistical procedure.

Document image analysis: A primer

Document image analysis refers to algorithms and techniques that are applied to images of documents to obtain a computer-readable description from pixel data. A well-known document image analysis product is the Optical Character Recognition (OCR) software that recognizes characters in a scanned document. OCR makes it possible for the user to edit or search the document’s contents. In this paper we briefly describe various components of a document analysis system. Many of these basic building blocks are found in most document analysis systems, irrespective of the particular domain or language to which they are applied. We hope that this paper will help the reader by providing the background necessary to understand the detailed descriptions of specific techniques presented in other papers in this issue.

Improving the Accessibility of User Accessible Archives

This paper reviews the requirements associated with developing systems designed to convert large printed archives into electronic text format. Consideration of human factors issues include signal detection theory and job design. These two issues will greatly influence system strategies and implementation. Needs assessment and cost analysis of the system will help constrain decisions concerning technology and system structure. Technical issues involve assessing scanner and optical character recognition (OCR) software capabilities and understanding how those relate to system design. Overall system design is expected to require cooperation between smaller organizations that share similar and complimentary needs.