Recognition of Merged Characters Based on Forepart Prediction, Necessity-Sufficiency Matching, and Character-Adaptive Masking

Song, Jiqiang; Zuo, Li; Lyu, Michael R.; Cai, Shaobin

doi:10.1109/tsmcb.2004.837588

Cited by 27 publications

(9 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However to get an idea about the segmentation result of the existing system on normal touching of horizontal direction, we compare the results in Table 1. It has been noted that, though the accuracy is less compared to that of result obtained by [10,17], but we think for multi-oriented environment it is a good performance. …”

Section: Comparison Of Segmentation Resultsmentioning

confidence: 51%

“…Although there is some published work to recognize the touching characters of normal horizontal direction [10,20], to the best of our knowledge, there is no work to recognize touching characters of arbitrary orientation. To handle multi-oriented touching string of map documents, in this paper, we propose a scheme to recognize two-character touching patterns of arbitrary orientations.…”

Section: Fig1 Example Of a Map Shows Orientation Of Different Charamentioning

confidence: 99%

“…Touching character recognition methods can be divided into two classes: segmentation based approach and holistic approach [10].…”

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Recognition of Multi-oriented Touching Characters in Graphical Documents

Roy

Pal

Lladós

2008

2008 Sixth Indian Conference on Computer Vision, Graphics &Amp; Image Processing

View full text Add to dashboard Cite

Touching characters are major problem of achieving higher recognition rate in Optical CharacterRecognition (OCR). Present OCR systems do not perform well when adjacent characters touch. If characters are touched in graphical documents (e.g. map) then such touching string recognition is more difficult because in such documents touching characters appear in multi-oriented direction. In this paper, we present a scheme towards the recognition of English two-character multi-oriented touching strings. When two or more characters touch, they generate a big cavity region at the background portion and we used this background information in our scheme. To handle the background information, convex hull is used. In this scheme, at first, a set of initial segmentation points is predicted based on the concave residues of the convex hull of the touching characters. Next, based on the initial points, we select some candidate segmentation lines. Finally the recognition confidence of two sub-images of a touching string, obtained from each candidate segmentation line is computed. The candidate segmentation line from which we get optimum confidence is the actual segmentation line and the corresponding characters in favour of which the two segmentation parts show optimum confidence is the recognition result of the touching string. To compute the recognition confidence, SVM classifier is used. The features used in the SVM are invariant to character orientation. Circular ring and convex hull ring based approach has been used along with angular information of the contour pixels of the character to make the feature rotation invariant. From the experiment we obtained encouraging result.

show abstract

Section: Comparison Of Segmentation Resultsmentioning

confidence: 51%

Section: Fig1 Example Of a Map Shows Orientation Of Different Charamentioning

confidence: 99%

See 1 more Smart Citation

Recognition of Multi-oriented Touching Characters in Graphical Documents

Roy

Pal

Lladós

2008

2008 Sixth Indian Conference on Computer Vision, Graphics &Amp; Image Processing

View full text Add to dashboard Cite

show abstract

“…Here, a number of recognition approaches, for example, template matching, neural network, and hidden Markov model (HMM), are used for the segmentation, which significantly improves the accuracy performance for characters. In Song et al, 11 a merged-character recognition framework is sophisticatedly designed, where a necessity-sufficiency matching is adopted for the segmentation with its characteradaptive template masking. Furthermore, a HMMbased segmentation is successfully presented to deal with Urdu scripts.…”

Section: Introductionmentioning

confidence: 99%

Segmentation of touching characters via tree-indexed demarcation using forward and backward searches

Tang

Zhang

et al. 2017

Advances in Mechanical Engineering

View full text Add to dashboard Cite

In this article, a tree-indexed demarcation using forward and backward searches is proposed to segment the seriously touching characters, which is used as an important procedure for the character recognition in the real industrial scenarios. Unlike the segmentation directly depending on the histogram analysis, the reliable and unreliable valleys are first identified from a smoothed projection curve in our method, where a set of image pre-procedures are carefully carried out to catch this smoothed curve. With these operations, two segment binary trees are sequently constructed via the corresponding forwarding and backward searches of the valleys. Then, the delimiter candidates are sophisticatedly selected with a number of well-designed rules. The optimal segmented paths are finally obtained on the width variation measure of the characters, and the isolated characters are achieved. Experiments show that with our tree-indexed demarcation using forward and backward searches method, the touching character groups can be effectively segmented to better serve the industrial automatic recognition system.

show abstract

“…Many OCR tools are built to recognize individual characters [7,15,22,8]. As a result of this, to achieve handwritten text recognition, we often need to segment a connected word (or words) into individual characters [18], which we call handwritten text segmentation in this paper.…”

Section: Introductionmentioning

confidence: 99%