Narrow Results

Page segmentation is one of the important and basic research subjects of document analysis. There are two major kinds of page segmentation methods, i.e. hierarchical and no-hierarchical ones. Most traditional techniques such as top–down and bottom–up approaches belong to the hierarchical method. Though these two approaches have been used till now, they are not effective for processing documents with high geometric complexity and the process of splitting document needs iterative operations which is time consuming. A non-hierarchical method called the modified fractal signature (MFS) was presented in recent years. It can overcome the above weaknesses, however the MFS needs to calculate modified fractal signature which makes the theory very complex. In this thesis, we present a new page segmentation approach: Median Order Statistic Filter (MedOSF) — Maximum Order Statistic Filter (MaxOSF) approach which is more direct and much simpler. We use the MedOSF to remove the salt–pepper noise of the document and use the MaxOSF to do the page segmentation. In practice, they not only can adaptively process the documents with high geometrical complexity, but also save a lot of computing time.

Text is commonly printed on a complex background. Segmenting text is an important part in document analysis. In the past some methods have been shown for the segmentation of texts with images. However, previous studies have not sufficiently addressed complex compound documents. This investigation presents an algorithm for the segmentation of text in various document images. The proposed segmentation algorithm applies a new multilayer segmentation method to separate the text from various compound document images, independent from the text and background overlapping or not. This method solves various problems associated with the complexity of background images. Experimental results obtained using various document images scanned from book covers, advertisements, brochures and magazines, reveal that the proposed algorithm can successfully segment Chinese and English text strings from various backgrounds, regardless of whether the texts are over a simple, slowly varying or rapidly varying background texture.

In this article, a new minimum spanning tree-based method for shape description and matching is proposed. Its properties are checked through the problem of graphical symbols recognition. Recognition invariance in front shift and multi-oriented noisy objects was studied in the context of small and low resolution binary images. The approach seems to have many desirable properties, even if the construction of graphs induces an expensive algorithmic cost. In order to reduce time computing, an alternative solution based on image compression concepts is provided. The recognition is realized in a compact space, namely the Discrete Cosine space. The use of block discrete cosine transform is discussed and justified. The experimental results led on the GREC2003 database show that the proposed method is characterized by a good discrimination power, a real robustness to noise, with an acceptable time computing. The position with a reference approach like Zernike moments is also investigated to measure the relevance of the proposed technique.

Document image segmentation is an important research area of document image analysis which classifies the contents of a document image into a set of text and non-text classes. Previous existing methods are often designed to classify text and halftone therefore they perform poorly in classifying graphics, tables and circuit, etc. In this paper, we present a robust multi-level classification method using multi-layer perceptron (MLP) and support vector machine (SVM) to segment the texts from non-texts and thereafter classify them as tables, graphics and halftones. This method outperforms previously existing methods by overcoming various issues associated with the complexity of document images. Experimental results prove the effectiveness of our proposed method. By virtue of our multi-level classification approach, the text components, halftone components, graphic components and table components are accurately classified respectively which would highly improve OCR accuracy to reduce garbage symbols as well as increase compression ratio thereafter simultaneously.

Techniques to identify printed and handwritten text in scanned documents differ significantly. In this paper, we address the question of how to discriminate between each type of writing on registration forms. Registration-form documents consist of various type zones, such as printed text, handwriting, table, image, noise, etc., so segmenting the various zones is a challenge. We adopt herein an approach called “multiscale-region projection” to identify printed text and handwriting. An important aspect of our approach is the use of multiscale techniques to segment document images. A new set of projection features extracted from each zone is also proposed. The classification rules are mining and are used to discern printed text and table lines from handwritten text. The proposed system was tested on 11118 samples in two registration-form-image databases. Some possible measures of efficiency are computed, and in each case the proposed approach performs better than traditional methods.

Textual information embedded in multimedia can provide a vital tool for indexing and retrieval. Text extraction process has many inherent problems due to the variation in font sizes, color, backgrounds and resolution. Text detection and localization are the most challenging phases of text extraction process whereas text extraction results are highly dependent upon these phases. This paper focuses on the text localization because of its very fundamental importance. Two effective feature vectors are introduced for the classification of the text and nontext objects. First feature vector is represented by the Radon transform of text candidate objects. Second feature vector is derived from the detailed geometrical analysis of text contents. Union of two feature vectors is used for the classification of text and nontext objects using support vector machine (SVM). Text detection and localization results are evaluated on two publicly available datasets namely ICDAR 2013 and IPC-Artificial text. Moreover, results are compared with state-of-the-art techniques and the Comparison demonstrates the superiority of the presented research.

The purpose of this paper is to retrieve and study the highly cited papers as well as the correlation between the citation frequency and the download frequency of the 20 traditional Chinese medicine journals in China, in order to provide the guidance for improving the influence and academic quality of these journals.

Bibliometric analyses were conducted on 1103 papers of 20 traditional Chinese medicine journals from 2011 to 2020 by retrieving for the China Academic Journal Network Publishing Database (CAJD) in China National Knowledge Infrastructure (CNKI). SPSS 17.0 software was used to analyze the correlation between the citation frequency and the download frequency via conducting regression fitting and establishing the mathematical models.

The results showed that the total citations of the 1103 papers were 93051 times and the average citations were 84.36 times per paper. The total downloads of the 1103 papers were 2058442 times, and the average downloads were 1866.22 times per paper. China Journal of Chinese Materia Medica ranked first according to the number of papers, total citations and total downloads. The citations of Journal of Chinese Medicinal Materials ranked first based on the number of citations per paper. One of Li’s paper had been cited the most (983 times). There were 629 (57.03%) papers whose first author was from universities. The scopes of the first authors were distributed in 29 regions and 2 special administrative regions (Macao, Hong Kong) in China. The authors from Beijing published 283 (25.66%) papers, ranking number one. The number of papers supported by funds was 882 (79.96%). The research results of correlation showed that the citation frequency and the download frequency of the highly cited papers had a highly positive correlation from both journal and paper level for whether the sample data of journals was normally distributed or nonnormally distributed. The correlation coefficients of the 20 journals at journal level and that at paper level were 0.9765 and 0.6677, respectively. The correlation was better at journal level than that at paper level, while the optimal regression fitting was all cubic polynomial. Among the 1103 papers, there were 684 (62.01%) research papers and 419 (37.99%) review papers. The main citation period of the top 15 papers was from the 2nd year to the 6th year after publication, accounting for 78.39%.

Papers on clinical therapeutics research, papers on the pharmacological effects and its mechanism of traditional Chinese medicine, and papers on traditional Chinese medicine and natural medicine were the main source of the highly cited papers of the traditional Chinese medicine journals. Editors of the journals should focus on the above-mentioned research areas to select manuscripts for exploiting the excellent sources extensively, while paying attention to review papers, focusing on national major or key projects, paying attention to network spreading, stabilizing authors with quality services, in order to improve the influence and the academic quality of journals.

The compressor is the most core component of refrigeration products, and its performance directly affects the overall performance of the product. The exhaust silencing chamber and coil are key components of a refrigeration compressor, and their working stability and lifespan determine the reliability and economy of the refrigeration compressor. This paper obtains reference data for the fluid domain of the exhaust silencing chamber through simulation analysis and image processing of the existing exhaust silencing chamber structure and inner diameter size of the coil. Then, the optimal model of the prioritized exhaust chamber is obtained through pattern recognition method. Based on the image processing method, the internal flow field of the exhaust chamber are obtained, providing a strong reference for the optimal structural design of the exhaust silencing chamber and coil of the refrigeration compression body.

We developed a check reading system, termed INTERCHEQUE, which recognizes both the legal (LAR) and the courtesy amount (CAR) on bank checks. The version presented here is designed for the recognition of French, omni-bank, omni-scriptor, handwritten bank checks, and meets industrial requirements, such as high processing speed, robustness, and extremely low error rates.

We give an overview of our recognition system and discuss some of the pattern recognition techniques used. We also describe an installation which processes of the order of 70,000 checks per day. Results on a data base of about 170,000 checks show a recognition rate of about 75% for an error rate of the order of 1/10,000 checks.

Current state-of-the-art techniques in 2D chart analysis primarily emphasize the recognition of textual information as a means of comprehending and summarizing chart contents. However, the effective analysis and understanding of information embedded in chart images depends on accurate reverse-engineering of the behavior of depicted variables. In this paper, we propose a methodology, named Abax, as an initial study for recognizing and approximating the mathematical functions that describe the behavior of variables illustrated in chart images, particularly those containing curves. Abax is focused on approximating the values of function parameters using spatial pixel information derived from the identified keypoints of each curve. Qualitative results of the described method are presented as a proof of concept, demonstrating accurate extraction of information from fives types of functions: linear, polynomial, asymptotic, sinusoidal and arbitrary.

Document image processing has become an increasingly important technology in the automation of office documentation tasks. Automatic document scanners such as text readers and OCR (Optical Character Recognition) systems are an essential component of systems capable of those tasks. One of the problems in this field is that the document to be read is not always placed correctly on a flat-bed scanner. This means that the document may be skewed on the scanner bed, resulting in a skewed image. This skew has a detrimental effect on document analysis, document understanding, and character segmentation and recognition. Consequently, detecting the skew of a document image and correcting it are important issues in realizing a practical document reader. This paper presents the use of analyzing the connected components extracted from the binary image of a document page. Such an analysis provides a lot of useful information, and will be used to perform skew correction, segmentation and classification of the document. Moreover, we describe two new algorithms — one for skew detection and one for skew correction. The new skew correction algorithm we propose has been shown to be fast and accurate, with run times averaging under 1.5 CPU seconds and 30 seconds real time to calculate the angle on a 5000/20 DEC workstation. Experiments on over 100 pages show that the method works well on a wide variety of layouts, including sparse textual regions, mixed fonts, multiple columns, and even for documents with a high graphical content.

This article presents an automatic system that takes in grayscale scanned images, which could be mixed text/graphic documents, and performs thresholding and skew detection on the document images. The system consists of two major components; multistage thresholding and skew detection. The proposed skew detection algorithm has no restriction on detectable angle range and does not rely on large blocks of text. It works well on textual document images, graphical images and mixed text and graphic images. The performance of the systems was evaluated using over 60 images that consist of real life documents like envelopes and artificial mixed text/graphic icons. The superior performance of thresholding is clear compared to other techniques from the evaluation. The skew detection algorithm is robust when compared with other methods when very few text lines are present in the document image.

The large and constantly growing amounts of available text documents hold great potential for the exploration of knowledge. However, in the light of the vast quantity and variety of available documents, one fact should not be forgotten: the results of a knowledge discovery in texts are only as good as the underlying document collection. That is why analysts have to ensure that document collections adequately represent the specific area under examination and thereby to minimise the bias and to maximise the generalisable nature of the knowledge brought to light. Surprisingly, knowledge management research has barely paid any attention to the problems of such a document quality assessment and rigorous document selection. This paper addresses that research gap and makes two contributions: In the first step, building on a cross-disciplinary exchange with social research, development of a framework for the quality assessment and collection of documents. This artefact provides concrete guidance for compiling suitable, high-quality document collections and makes a contribution to ensuring “document collection quality” within the context of knowledge discovery in texts. In the second step, the framework is evaluated in a practical demonstration. In this context, the demonstration also exemplifies how different document collections influence the results of knowledge discoveries.

Recognition of handwritten characters is difficult because of variability involved in the writing style of different individuals. This paper deals with recognition of off-line Oriya handwritten characters using quadratic classifier based on the features obtained mainly from directional chain code information. Here, at first, the bounding box of a character is segmented into blocks and directional chain code features are computed in each of these blocks. Next, these blocks are down sampled using Gaussian filter for recognition. Finally, chain code features obtained from down sampled blocks are fed to the quadratic classifier for recognition. We used two sets of feature vectors (one feature vector has 64 dimension and the other has 400 dimension) and their corresponding results obtained from the classifier are reported. We tested our system on 9556 Oriya off-line handwritten characters and we obtained 91.11% accuracy from the proposed recognition system when 400 dimensional feature vector was considered. We have used five-fold cross-validation technique for result computation.

The problem of determining the script of a document image has a number of important applications in the filed of document analysis, such as indexing and sorting of large collections of such images, or a precursor to optical character recognition (OCR). This paper concerns the extraction of texture features and the use of such features for determining the script of a document image, based on the observation that text has a distinct visual texture. Texture features are extracted using wavelet decomposed images of an image and its complement, and their effectiveness is tested on 32 texture category. These features are then used to solve the problem of script identification in document image processing. The scheme has been tested on 8 Indian scripts and found to be robust to the skew generated in the process of scanning. The proposed system achieves an overall classification accuracy of 97.55 % on a large testing set.

We developed a check reading system, termed INTERCHEQUE, which recognizes both the legal (LAR) and the courtesy amount (CAR) on bank checks. The version presented here is designed for the recognition of French, omni-bank, omni-scriptor, handwritten bank checks, and meets industrial requirements, such as high processing speed, robustness, and extremely low error rates.

We give an overview of our recognition system and discuss some of the pattern recognition techniques used. We also describe an installation which processes of the order of 70,000 checks per day. Results on a data base of about 170,000 checks show a recognition rate of about 75% for an error rate of the order of 1/10,000 checks.