Single-character segmentation for handprinted Thai word

Lohakan, M.; Airphaiboon, Surapan; Sangworasil, Manas

doi:10.1109/icdar.1999.791874

Cited by 7 publications

(3 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ittner [1] convolutes the projection profile with a Gaussian kernel and then uses the minima in the averaged projection to extract text lines. M. Lohakan [2], on the other hand, blurs the image in downward direction before computing the horizontal projection profile. The results of both methods depend on the size of the filter used in the averaging process.…”

Section: Review Of Thai Document Layout Analysismentioning

confidence: 99%

Rule-based middle-level character detection for simplifying Thai document layout analysis

Yingsaeree

Kawtrakul

2005

Eighth International Conference on Document Analysis and Recognition (ICDAR'05)

View full text Add to dashboard Cite

Although research interest in machine printedThai character recognition has been an intense research area in the past decade, there are only a few results available for Thai document layout analysis. In addition, directly using the method proposed for other languages with Thai documents is not possible since Thai documents have a unique characteristic (i.e., Thai characters can be placed in four different levels). This paper proposed an approach to eliminate that characteristic by removing nonmiddle-level characters from the image based on heuristic rules derived from Thai language properties: nonmiddle-level characters are usually smaller than middle-level characters, and the gap between each level is smaller than the gap between two consecutive lines. After removed, one can use any existing methods with Thai documents without any modification. The experimental results show that the proposed method can effectively remove nonmiddle-level characters from 200 test images with 99.46% accuracy even when the image contains various font sizes.

show abstract

Section: Review Of Thai Document Layout Analysismentioning

confidence: 99%

Rule-based middle-level character detection for simplifying Thai document layout analysis

Yingsaeree

Kawtrakul

2005

Eighth International Conference on Document Analysis and Recognition (ICDAR'05)

View full text Add to dashboard Cite

show abstract

“…Among the papers concerning Thai handwritten segmentation, they used the classical approaches to segment character [14], [15], [16] and the demonstrations are based on controlling the writing styles of the writers that is not practical. In addition, all those studies were based on modern Thai language and it is different from ancient Thai language.…”

mentioning

confidence: 99%

Character segmentation from ancient palm leaf manuscripts in Thailand

Chamchong

Fung

2011

Proceedings of the 2011 Workshop on Historical Document Imaging and Processing

View full text Add to dashboard Cite

This paper presents a character segmentation system from ancient palm leaf manuscripts written in ancient Thai language. This aims to develop an automated system for the digitization and processing of ancient manuscripts. In this paper, the preprocessing stage of noise reduction is carried out. An optimal binarization is selected in order to reduce the unrelated noise and background information on the document. The proposed approach can improve the readability of the documents and enable selection of the optimal binarization technique. Text line segmentation is then applied to partial projection profiles, and the characters are separated by using the contour tracing algorithm and a trace of background skeleton. The experiment results have shown that this proposed system can be used to support subsequent steps such as automatic recognition of characters from Thai ancient palm leaf.

show abstract

“…These studies focused on the English language; however, it is highly likely that this concept applies to other character-based languages such as Thai. A simple algorithm for separating units within a Thai sentence or phrase image is developed based on "contour following" (Lohakan 1999). These units may be isolated or connected components composed of characters, sub-characters, words, and sub-words.…”

Section: The Proposed Modelmentioning

confidence: 99%

Off-line Thai handwriting recognition in legal amount

Chatwiriya¹,

Klinkhachorn²

View full text Add to dashboard Cite

Single-character segmentation for handprinted Thai word

Cited by 7 publications

References 3 publications

Rule-based middle-level character detection for simplifying Thai document layout analysis

Rule-based middle-level character detection for simplifying Thai document layout analysis

Character segmentation from ancient palm leaf manuscripts in Thailand

Off-line Thai handwriting recognition in legal amount

Contact Info

Product

Resources

About