The world's first Vihakoa printed document identification system came out

National language and information technology has reached a new level. The world's first Vihako (Chinese-English) A (British) printed document identification system has been published.

The national character recognition field of Tsinghua University is further fruitful.

On August 10th, the world’s first print document identification system supporting Uyghur, Kazakh, Kirgiz (aka Waikow) and Arabic (Abbreviation for Arabic) in the unified system framework was introduced at Tsinghua University, which marks our country The Arabic word system document recognition technology has already ranked among the top in the world. It is reported that the system can accurately and efficiently transform the paper documents of Vihakowen into electronic documents, which greatly facilitates the information processing of the above-mentioned text materials, and will certainly promote the economic and cultural construction and external communicate with. The system also fully supports the recognition of the Arabic language used by more than 200 million people around the world. Therefore, it will also have a profound impact on the cooperation and exchanges between China and Arab countries.

At the system appraisal meeting held at Tsinghua University on August 10th, a number of information processing specialists led by Professor Ni Guangnan and appraisal committees composed of experts in the field of Vahovar language gave a very high evaluation of the system. The appraisal experts agreed that as the first domestic and foreign document recognition system supporting UH, Kazakh, Ke, and Ah in the unified system framework, the recognition rate of the actual text and the major technical indicators have reached the international advanced level. The outlook is very broad.

Experts stated that the technical achievements in the field of ethnic word recognition at Tsinghua University are arguably the golden key to the entrance of ethnic languages ​​into information technology. A nation’s writing records the history and culture of the nation, and a nation must keep pace with the development of the modern information society. The informatization of language and writing is the key. Only by translating a large amount of historical, economic, and cultural information recorded on paper into a computer-readable digital language can the ethnic treasures be promoted, processed, and preserved in order to achieve the goal of sharing and making full use of social and economic information resources. Therefore, the socio-economic and cultural development in ethnic areas can better keep pace with the progress of modern society.

It is understood that the so-called character recognition is based on OCR (optical character recognition) technology, which converts textual image information such as books, newspapers, manuscripts, forms, and other printed materials obtained by scanners, cameras, and other optical input methods into computer-readable and recognizable information. Text information. Because it can greatly improve work efficiency, text recognition technology has been more and more people's attention. Especially in China, due to the comprehensive development of informatization construction, it provides a broad application stage for word recognition technology, such as office automation, banking, taxation, digital library, electronic publishing, public security, transportation, postal sorting and other industries. Has a wide range of applications.

According to experts, Uighur, Kazakh, Kirgiz and Arabic are very similar to the main national languages ​​used in Xinjiang in China. The Arabic text form is very different from the block Chinese character, and it is more difficult to recognize the character because of the multiple characters and similar characters, and the way the text is written from right to left. The Vihako (Chinese-English) A (British) print document identification system jointly launched by Tsinghua University and Xinjiang University solved this problem.

It is understood that Vivaco (Chinese-English) A (UK) printed document identification system not only has a core method of character recognition based on statistical methods for multi-font, multiple-digit Vahovar acronyms, but also has a dimension that supports writing from right to left. Hako Arvin and text segmentation methods with English or Chinese characters written in reverse. According to the character classification results in the text segmentation, combined with the existing Chinese character and English character recognizers of Tsinghua University, the final document recognition system of the mixed arrangement of Vahova and Chinese-English, and Arabic and English are arranged.

Relying on years of technology accumulation in the field of OCR identification, the Department of Electronic Engineering of Tsinghua University has achieved remarkable results in the field of recognition of Chinese, Japanese, Korean, English, and Tibetan nationalities. Many of its text recognition technologies are among the leading international standards. Its Tsinghua TH-OCR core Japanese, Korean and Korean recognition technology won the international competition and was used in Microsoft Office 2003. The Vihako (Chinese-English) A (British and English) printed document identification system jointly launched with Xinjiang University once again brought the identification of ethnic characters in China into the world's leading ranks, and at the same time made China’s OCR technology further into the world. Added an important chip.