Offline character recognition using online character writing information

Nishimura, Hiromitsu; Timikawa, Takehiko

doi:10.1109/icdar.2003.1227653

Cited by 7 publications

(3 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, handwritten signatures may prove unreliable as a means of personal verification. Hence, there is a need for personal verification based on writing pressure data [4,5]. In verification using writing pressure, handwriting forgery becomes difficult, which leads to better security.…”

Section: Introductionmentioning

confidence: 99%

Character recognition by time–frequency analysis using writing pressure

Watanabe

Katsura

2011

Elect Comm in Japan

View full text Add to dashboard Cite

SUMMARYWith the development of information and communication technology, personal verification becomes more and more important. In the future ubiquitous society, the development of terminals handling personal information will require personal verification technology. The signature is one of the personal verification methods; however, the number of characters is limited in the case of the signature and therefore false signatures are easily used. Thus, personal identification from handwriting is difficult. This paper proposes a "haptic pen" that extracts the writing pressure, and shows a character recognition method based on time-frequency analyses. Although the figures of characters written by different writers are similar, the differences appear in the time-frequency domain. As a result, it is possible to use the proposed character recognition for personal identification more exactly. Experimental results show the viability of the proposed method.

show abstract

Section: Introductionmentioning

confidence: 99%

Character recognition by time–frequency analysis using writing pressure

Watanabe

Katsura

2011

Elect Comm in Japan

View full text Add to dashboard Cite

show abstract

“…Compared with online recognition, offline character recognition is more challenging in that it does not have the trace of the writer's pen as well as the order of writing in the input. It has been shown [10] that such pen dynamics information can help to obtain better recognition accuracy than static scanned images alone.…”

Section: Introductionmentioning

confidence: 99%

Interpretable Distance Metric Learning for Handwritten Chinese Character Recognition

Dong¹,

Varde²,

Stevanović³

et al. 2021

Preprint

View full text Add to dashboard Cite

Handwriting recognition is of crucial importance to both Human Computer Interaction (HCI) and paperwork digitization. In the general field of Optical Character Recognition (OCR), handwritten Chinese character recognition faces tremendous challenges due to the enormously large character sets and the amazing diversity of writing styles. Learning an appropriate distance metric to measure the difference between data inputs is the foundation of accurate handwritten character recognition. Existing distance metric learning approaches either produce unacceptable error rates, or provide little interpretability in the results. In this paper, we propose an interpretable distance metric learning approach for handwritten Chinese character recognition. The learned metric is a linear combination of intelligible base metrics, and thus provides meaningful insights to ordinary users. Our experimental results on a benchmark dataset demonstrate the superior efficiency, accuracy and interpretability of our proposed approach.

show abstract

“…(Plamondon & Srihari, 2000) Even though there is less data, the recognition rates reported are much higher for the online case in comparison to the offline case. (Nishimura & Timikawa, 2003) In this thesis, only online recognition techniques are considered because the shorthand outline is an online pseudo cursive script and we are seeking a real-time, online method for fast text entry. The literature reviewed spans preprocessing methods, character recognition algorithms, shorthand recognition techniques and Korean character recognition.…”

Section: Introductionmentioning

confidence: 99%

Machine compatible script for fast text entry of text using handwriting

Ma¹

View full text Add to dashboard Cite

There is a wish to be able to enter text into hand-held devices at the speed of speech (around 180 words per minute). Accurate unconstrained speech recognition remains an unresolved research challenge. Commonly used handwriting methods, like block capital letters and cursive script, are relatively slow and can only achieve a fraction of the speed of spoken speech (30 words per minute would be considered fast handwriting). Only handwritten shorthand schemes, which can, with training, be used at speeds in excess of 100 words per minute, can achieve a rate close to the speed of normal speech. This research focuses on techniques of recognizing and translating two shorthand systems-Pitman shorthand and Renqun shorthand, which represent English and Chinese respectively. According to the review of the latest on-line handwriting recognition techniques, most techniques are so sensitive to geometric features of characters that they are not able to achieve a satisfactory recognition accuracy for the recognition of shorthand. Shape recognition techniques recently demonstrated for the recognition of Korean script are applicable for recognizing shorthand handwritings. Based on this, an overall solution to the recognition of both Pitman shorthand and Renqun shorthand is proposed. Recognition approaches of Pitman shorthand components are firstly discussed in detail. A two-stage approach (segmentation & classification) is proposed for the recognition of the consonant outlines, which is the most difficult part in Pitman notation. A template-based matching approach is proposed for the shortform classification. Hausdorff Distance is introduced to measure the similarity between the input outline and the corresponding templates. Compared with Pitman shorthand, both vocalized outlines and shortforms of Renqun shorthand are composed of rhymes and consonants and have similar shape features. Due to the inherited geometric features of Renqun shorthand from Pitman

show abstract

Offline character recognition using online character writing information

Cited by 7 publications

References 5 publications

Character recognition by time–frequency analysis using writing pressure

Character recognition by time–frequency analysis using writing pressure

Interpretable Distance Metric Learning for Handwritten Chinese Character Recognition

Machine compatible script for fast text entry of text using handwriting

Contact Info

Product

Resources

About