Text Line Segmentation Research Articles

A two-step framework for text line segmentation in historical Arabic and Latin document images

A two-step framework for text line segmentation in historical Arabic and Latin document images

Segmentation of text lines using multi-scale CNN from warped printed and handwritten document images

Segmentation of text lines using multi-scale CNN from warped printed and handwritten document images

结合文字核心区域和扩展生长的藏文古籍文本行切分

结合文字核心区域和扩展生长的藏文古籍文本行切分

Accurate Fine-Grained Layout Analysis for the Historical Tibetan Document Based on the Instance Segmentation

Accurate layout analysis without subsequent text-line segmentation remains an ongoing challenge, especially when facing the Kangyur, a kind of historical Tibetan document featuring considerable touching components and mottled background. Aiming at identifying different regions in document images, layout analysis is indispensable for subsequent procedures such as character recognition. However, there was only a little research being carried out to perform line-level layout analysis which failed to deal with the Kangyur. To obtain the optimal results, a fine-grained sub-line level layout analysis approach is presented. Firstly, we introduced an accelerated method to build the dataset which is dynamic and reliable. Secondly, enhancement had been made to the SOLOv2 according to the characteristics of the Kangyur. Then, we fed the enhanced SOLOv2 with the prepared annotation file during the training phase. Once the network is trained, instances of the text line, sentence, and titles can be segmented and identified during the inference stage. The experimental results show that the proposed method delivers a decent 72.7% average precision on our dataset. In general, this preliminary research provides insights into the fine-grained sub-line level layout analysis and testifies the SOLOv2-based approaches. We also believe that the proposed methods can be adopted on other language documents with various layouts.

Learning-Free Text Line Segmentation for Historical Handwritten Documents

We present a learning-free method for text line segmentation of historical handwritten document images. This method relies on automatic scale selection together with second derivative of anisotropic Gaussian filters to detect the blob lines that strike through the text lines. Detected blob lines guide an energy minimization procedure to extract the text lines. Historical handwritten documents contain noise, heterogeneous text line heights, skews and touching characters among text lines. Automatic scale selection allows for automatic adaption to the heterogeneous nature of handwritten text lines in case the character height range is correctly estimated. In the extraction phase, the method can accurately split the touching characters among the text lines. We provide results investigating various settings and compare the model with recent learning-free and learning-based methods on the cBAD competition dataset.

An adaptive text-line extraction algorithm for printed Arabic documents with diacritics

The performance of document text recognition depends on text line segmentation algorithms, which heavily relies on the type of language, author’s writing style, pen type, and document quality. In this paper, we present a novel unsupervised text-line segmentation algorithm for printed Arabic documents with and without diacritics. The presented approach employs a projection profile along with connected components in an iterative manner to detect text-lines. The primary benefits of the presented algorithm are (i) it is not threshold dependent, (ii) it is not required a training phase for threshold selection, and (iii) it is robust towards page rotation, font type, size, and style variation for both with and without diacritics documents. The extensive computational simulations on manually collected dataset prove the efficiency of the proposed scheme compared with several baseline and states of the art methods, including, Voronoi, X-Y Cut, Docstrum, Smearing and Seam-carving methods. Computational time analysis also presented.

A Multi-Agent Approach to Segment Arabic Handwritten Text Lines

In text line segmentation, there are three classes of methods: either by sorting physical units such as pixels or connected components (CC) constituting a line or by searching for the baseline of each word and grouping together those who participate in the same line. The third class analyzes the separation locations between the lines. After an overview of lines segmentation approaches, the authors introduced a new method emphasizing its simplicity, speed, and originality. The proposed approach detects the starting components of the lines in the first step. In the second step, it defines a number of agents that start the segmentation process from their starting points between the starting components of lines. Each agent aims to reach the left edge of the document through the correct path. The algorithm used by the agents is based on the morphological process, characteristics of the Arabic manuscript and a communication system. The experimental results on an Arabic dataset show that this approach is an effective solution for the segmentation of lines from different Arabic manuscripts.

Text line segmentation based on local baselines and connected component centroids for Tibetan historical documents

Text line segmentation is one of the key contents in document image analysis and recognition. Due to the complex situation of text images in Tibetan historical document, such as the coexistence of slanting and distortion of text lines, especially the adhesion between lines, text line segmentation has become a challenging task. In this paper, a new method of text segmentation for Tibetan historical document is proposed. This method first obtains the local baseline information of the text line, and then performs the detection and segmentation of the adhesion area. Finally, according to the barycentre coordinates of the connected component and the local baseline information, each connected component is assigned to the corresponding text line. The method avoids the effect of text line distortion on the segmentation of the stuck text line and greatly reduces the error rate of the adhesion text line segmentation and can effectively deal with the segmentation of slanted and distorted Tibetan sticky text lines.

Line and word segmentation of handwritten text document by mid-point detection and gap trailing

This paper presents the text line and word segmentation from unconstrained handwritten documents based on horizontal projection histogram (HPH) to detect mid-points and gap trailing between lines. The midpoints are estimated from the HPH for the first 100 to 200 columns of the whole document. Then, considering the mid-points, the gap is tracked between two consecutive lines from locally computed HPH for a block having k rows and j columns. The HPH block is examined for various cases to locate optimal rows that separate adjacent lines. The proposed method segments curve, touching and skew-lines and is robust to writing variation and language independent. Word segmentation is not treated as a separate problem and goes efficiently alongside the line segmentation. As the trailing of space between neighboring lines goes on, the vertical projection Histogram (VPH) of t columns is monitored between the above and below separator of a line and find the optimal word separator. The algorithm is evaluated on two isolated datasets of different languages (Meitei Mayek and English). Text-line and word segmentation on Meitei Mayek handwritten documents achieve 91.84% and 88.96% accuracy respectively. Similarly, the handwritten English document meets 94.18% and 87.73% accuracy for line and word segmentation.

Automatic tracing and extraction of text‐line and word segments directly in JPEG compressed document images

JPEG is one of the popular and efficient compression algorithms supported in the consumer electronics world. Excessive usage of mobile phones and e-governance applications have all resulted in a huge collection of JPEG compressed document images. The major challenge with these images is that its processing becomes expensive as it requires repeated decompression and recompression operations. Recently, it has been proved that developing algorithms to operate directly on the compressed data is one of the solutions in overcoming the above issue. This research study investigates a novel algorithm for segmentation of text-lines and words directly from JPEG compressed handwritten document images. Segmenting a handwritten document is challenging due to the presence of uneven spacing, variable font sizes, overlapping and touching components, and it becomes much more challenging if it is to be done directly in the compressed image. The proposed technique virtually fixes a vertical stripe at the beginning of the document to detect starting points of text-lines. Then a moving window-based space penetration algorithm is used for tracing the exact line boundary between two text-lines, resolving the issues of space and font variations, touching and overlapping components. Subsequently, a word boundary tracing algorithm is used to segment words.

Building an efficient OCR system for historical documents with little training data

As the number of digitized historical documents has increased rapidly during the last a few decades, it is necessary to provide efficient methods of information retrieval and knowledge extraction to make the data accessible. Such methods are dependent on optical character recognition (OCR) which converts the document images into textual representations. Nowadays, OCR methods are often not adapted to the historical domain; moreover, they usually need a significant amount of annotated documents. Therefore, this paper introduces a set of methods that allows performing an OCR on historical document images using only a small amount of real, manually annotated training data. The presented complete OCR system includes two main tasks: page layout analysis including text block and line segmentation and OCR. Our segmentation methods are based on fully convolutional networks, and the OCR approach utilizes recurrent neural networks. Both approaches are state of the art in the relevant fields. We have created a novel real dataset for OCR from Porta fontium portal. This corpus is freely available for research, and all proposed methods are evaluated on these data. We show that both the segmentation and OCR tasks are feasible with only a few annotated real data samples. The experiments aim at determining the best way how to achieve good performance with the given small set of data. We also demonstrate that obtained scores are comparable or even better than the scores of several state-of-the-art systems. To sum up, this paper shows a way how to create an efficient OCR system for historical documents with a need for only a little annotated training data.

An adaptive threshold algorithm for offline Uyghur handwritten text line segmentation

This paper presents an effective text-line segmentation algorithm and evaluates its performance on Uyghur handwritten text document images. Projection based adaptive threshold selection mechanism is implemented to detect and segment the text lines with different valued thresholds. The robustness of the proposed algorithm is admirable that experiments on 210 Uyghur handwritten document image including 2570 text lines got correct segmentation by 97.70% precision and 99.01% recall rate and outperformed the compared classic text-line segmentation algorithm on same evaluation set. Additionally, the proposed algorithm is tested on the public handwriting dataset and get 98.05% correct segmentation rate which is robust and promising.

Segmentation and Recognition for Historical Tibetan Document Images

As a shining pearl in traditional Tibetan culture, historical Tibetan documents have received extensive attention from historians, linguists and Buddhist scholars. These documents are converted into digital form using Tibetan document segmentation and recognition methods. The document digitization is of great significance for the research, protection and inheritance of Tibetan history. This paper proposes an overall segmentation and recognition framework for historical Tibetan document images. Firstly, the historical Tibetan document image is preprocessed to correct imbalanced illumination, tilt and noises, and is further transformed into the binarized image. Secondly, we propose a layout segmentation method based on block projection to segment Tibetan document images into texts, lines and frames. Thirdly, in order to solve the problems of touching strokes between text-lines and curvilinear text-lines, we present a text-line segmentation method based on graph model for historical Tibetan text-line segmentation. Lastly, we present a touching segmentation method to segment touching Tibetan character string, and then recognize Tibetan characters. Experimental results show our proposed methods on layout segmentation, text-line segmentation and touching character string segmentation, achieve the satisfactory performance. The proposed methods can also be applied to other fonts in Tibetan font family.

Line segmentation of handwritten text using histograms and tensor voting

Line segmentation of handwritten text using histograms and tensor voting

A Hybrid Method for Mathematical Expression Detection in Scientific Document Images

Mathematical expressions have been widely used in scientific documents. In order to analyze the documents, automatic detection of mathematical expressions is a crucial step. The paper presents a unified system for the detection of mathematical expressions including both inline and isolated expressions in scientific document images that usually consist of heterogeneous components (e.g., figures, tables, text and expressions). In the system, a hybrid method of two stages is proposed for the effective detection of mathematical expressions. First, the layout analysis of entire document images is introduced to improve the accuracy of text line and word segmentation. Then, both isolated and inline expressions in document images are detected. Both hand-crafted and deep learning features are extensively investigated and combined to improve the detection accuracy. Furthermore, a generic performance metric is applied to evaluate the system comprehensively. The proposed method has been evaluated on two public benchmark datasets (Marmot and GTDB). The obtained accuracies of isolated and inline expressions in the Marmot dataset are 91.18% and 81.35% while those in the GTDB dataset are 89.51% and 80.20%, respectively. The performance comparison is carried out with the conventional methods to show the outstanding effectiveness of the proposed system. Moreover, extensive experiments have been performed in order to point out the effect of document image resolution and post processing techniques on mathematical expression detection.

Text line segmentation and binarization of handwritten historical documents using the fast and adaptive bidimensional empirical mode decomposition

Text line segmentation is a basic step that can be used to serve several processes in the recognition stage. Image segmentation is thus inevitable. A necessary step for text line segmentation is the image binarization. However, it is challenging due to complex background in such images. In this paper, we present a new text line segmentation and banarization method based on fast and adaptive bidimensional empirical mode decomposition (FABEMD). To achieve this, we use the FABEMD method to decompose each image into two components. The binarization threshold is calculated taking into account the characteristics of each component. The inspiration for the new binarization method was drawn from the standard Niblack's and Sauvola's Algorithms. Then the text line segmentation is performed after eliminating the BIMF1 (bidimensional intrinsic mode functions) which could disrupt the segmentation process. The results show that the proposed approach is less sensitive to noise and provides improved text line segmentation and binarization compared to iterative methods.

A novel method of text line segmentation for historical document image of the uchen Tibetan

Text line segmentation is a key step in Tibetan historical document recognition. A novel method for text line segmentation was proposed based on the baseline in uchen Tibetan, and a new dataset was released, which was used to evaluate the results of text line segmentation of uchen Tibetan historical documents. In this paper, there were two steps for the proposed method: baseline detection and text line segmentation using the baseline. In baseline detection, the upper edges of all characters in the document were obtained by a horizontal gradient operator, then an edge connectivity definition was proposed by which the upper edge set was divided into disjoint subsets. Eligible sets were selected from these subsets, and the edges in these sets were joined in turn to obtain the baseline. In text line segmentation, the document image was truncated at the baseline position, then the adhesion regions were segmented again. Each connected region in the image was assigned to its nearest baseline. All connected regions belonging to the same baseline formed a text line. Experiments on the proposed dataset showed that the method could effectively avoid document distortion, the accuracy of text line segmentation was high, and the text line adhesion could be handled.

An Efficient Algorithm to Find the Height of a Text Line and Overcome Overlapped and Broken Line Problem during Segmentation

Line segmentation is a critical phase of the Optical Character Recognition (OCR) which separates the individual lines from the image documents. The accuracy rate of the OCR tool is directly proportional to the line segmentation accuracy followed by the word/character segmentation. In this context, an algorithm, named height_based_segmentation is proposed for the text line segmentation of printed Odia documents. The proposed algorithm finds the average height of a text line and it helps to minimize the overlapped text line cases. The algorithm also includes post-processing steps to combine the modifier zone with the base zone. The performance of the algorithm is evaluated through the ground truth and also by comparing it with the existing segmentation approaches.

Handwritten Arabic Documents Segmentation into Text Lines using Seam Carving

Inspired from human perception and common text documents characteristics based on readability constraints, an Arabic text line segmentation approach is proposed using seam carving. Taking the gray scale of the image as input data, this technique offers better results at extracting handwritten text lines without the need for the binary representation of the document image. In addition to its fast processing time, its versatility permits to process a multitude of document types, especially documents presenting low text-to-background contrast such as degraded historical manuscripts or complex writing styles like cursive handwriting. Even if our focus in this paper was on Arabic text segmentation, this method is language independent. Tests on a public database of 123 handwritten Arabic documents showed a line detection rate of 97.5% for a matching score of 90%.

Text line Segmentation in Compressed Representation of Handwritten Document using Tunneling Algorithm

In this research work, we perform text line segmentation directly in compressed representation of an unconstrained handwritten document image. In this relation, we make use of text line terminal points which is the current state-of-the-art. The terminal points spotted along both margins (left and right) of a document image for every text line are considered as source and target respectively. The tunneling algorithm uses a single agent (or robot) to identify the coordinate positions in the compressed representation to perform text-line segmentation of the document. The agent starts at a source point and progressively tunnels a path routing in between two adjacent text lines and reaches the probable target. The agent's navigation path from source to the target bypassing obstacles, if any, results in segregating the two adjacent text lines. However, the target point would be known only when the agent reaches the destination; this is applicable for all source points and henceforth we could analyze the correspondence between source and target nodes. Artificial Intelligence in Expert systems, dynamic programming and greedy strategies are employed for every search space while tunneling. An exhaustive experimentation is carried out on various benchmark datasets including ICDAR13 and the performances are reported.