The past few decades have witnessed an intensive research on optical character recognition (OCR) for Roman, Chinese, and Japanese scripts. A lot of work has been also reported on OCR efforts for various Indian scripts, like Devanagari, Bangla, Oriya, Tamil, Telugu, Malayalam, Kannada, Gurmukhi, Gujarati, etc. In this paper, we present a review of OCR work on Indian scripts, mainly on Bangla and Devanagari—the two most popular scripts in India. We have summarized most of the published papers on this topic and have also analysed the various methodologies and their reported results. Future directions of research in OCR for Indian scripts have been also given.

/pdf/a-survey-on-optical-character-recognition-for-bangla-and-3tmybx9xa2.pdf

A survey on optical character recognition for Bangla and Devanagari scripts

The number of programming tutorial videos on the web increases daily. Video hosting sites such as YouTube host millions of video lectures, with many programming tutorials for various languages and platforms. These videos contain a wealth of valuable information, including code that may be of interest. However, two main challenges have so far prevented the effective indexing of programming tutorial videos: (i) code in tutorials is typically written on-the-fly, with only parts of the code visible in each frame, and (ii) optical character recognition (OCR) is not precise enough to produce quality results from videos. We present a novel approach for extracting code from videos that is based on: (i) consolidating code across frames, and (ii) statistical language models for applying corrections at different levels, allowing us to make corrections by choosing the most likely token, combination of tokens that form a likely line structure, and combination of lines that lead to a likely code fragment in a particular language. We implemented our approach in a tool called ACE, and used it to extract code from 40 Android video tutorials on YouTube. Our evaluation shows that ACE extracts code with high accuracy, enabling deep indexing of video tutorials.

Extracting code from programming tutorial videos

Chinese spelling check is an important component for many NLP applications, including word processor and search engines. However, compared to checkers for alphabetical languages (e.g., English or French), Chinese spelling checkers are more difficult to develop, because there are no word boundaries in Chinese writing system, and errors may be caused by various Chinese input methods. In this paper, we proposed a novel method to Chinese spelling checking. Our approach involves error detection and correction based on the phrasal statistical machine translation framework. The results show that the proposed system achieves significantly better accuracy in error detecting and more satisfactory performance in error correcting.

Chinese Spelling Checker Based on Statistical Machine Translation

This paper proposes an OCR post-processing approach based on multi-knowledge, which integrates language knowledge and candidate distance information given by the OCR engine. In this approach, statistical language model and semantic lexicon are combined, and candidate distance information is used to reduce the size of the search space. The experimental results show that this approach is very effective. After post-processing, the recognition accuracy rate on the test set increases from 58.45% to 83.73%, which means 60.84% error reduction.

An OCR post-processing approach based on multi-knowledge

機器翻譯為本的中文拼字改錯系統 (Chinese Spelling Checker Based on Statistical Machine Translation)

This work describes an effective spelling check approach for Chinese OCR with a new multi-knowledge based statistical language model. This language model combines the conventional n-gram language model and the new LSA (latent semantic analysis) language model, so both local information (syntax) and global information (semantic) are utilized. Furthermore, Chinese similar characters are used in Viterbi search process to expand the candidate list in order to add more possible correct results. With our approach, the best recognition accuracy rate increases from 79.3% to 91.9%, which means 60.9% error reduction.

/pdf/a-chinese-ocr-spelling-check-approach-based-on-statistical-12sfd7q1wo.pdf

S. Naoi

Papers

A Chinese OCR spelling check approach based on statistical language models