Science and technology
make cultural relics "alive"

Protecting historical and cultural heritage is to record and inherit the history of civilization development. The digitization of ancient books based on OCR technology is one of the best measures to protect cultural heritage. Phoenix Fire is the world leader in artificial intelligence OCR technology for various Brahmic scripts and Chinese minority scripts.

Company mission

We have been committed to "inheriting and spreading culture with technology and innovation". Using AI technologies such as computer vision, neural networks and machine learning, to promote the construction of digital humanities in multiple languages.

Industry leading

  • OCR for Khmer

    Khmer is the oldest and most complex script in Southeast Asia. Phoenix Fire has in-depth cooperation with Beijing Foreign Studies University to realize the OCR for Khmer first in the world

  • OCR for all Tibetan Characters

    Supports all single characters included in the GB/T character set, expansion set A and B, and additionally supports various multi-layer stacked characters in Tibetan transliteration of Sanskrit (The Sanskrit written in the Tibetan script), as well as Tibetan abbreviations and composite characters

  • OCR for various script fonts

    Based on 66 Tibetan fonts training, supports various regular script ,cursive script and running script fonts, including: Uchen, Betsu, Dutsa, Tsuing, Tsutong, Tsumachu, and other Tibetan fonts. Based on 42 Khmer fonts training, supports Khmer OS, Moul, Metal Chrieng, other Khmer fonts

  • various book formats and layouts

    Supports OCR for various book formats and layouts, including various ancient books of palm-leaf scriptures, woodblock editions and manuscripts (including Chinese Pothi Binding, Sewn Binding, Concertina Binding, Butterfly Binding, Scroll, etc.) and various modern books

  • Digital Library System

    Provides Visual Restoration Reader, supports books information retrieval and contents full-text search, automatic collation of multiple editions, multiple Romanization and transliteration, intelligent proofreading, massive thesaurus, intelligent segmentation and online translation

Core Technologies

  • AI OCR engine

    Based on AI core technologies such as computer vision, neural network, and machine learning, we can perform high-precision OCR for pictures of books in various languages and scripts, including Chinese, Latin alphabet scripts, Brahmic scripts (Tibetan, Khmer, etc.)

  • Brahmic scripts OCR technology

    Based on in-depth research on Brahmic script, we have developed a universal Brahmic script recognition technology (invention patent has been applied). Currently supported: Tibetan and Khmer, under development: Manchu, Mongolian, Sanskrit, etc.

  • Components recognizing

    By intelligently splitting and independently recognizing all the components (letters, marks, modifiers, signs) of single multi-layer stacked character, reorganizing and merging the Unicode character encoding, we finally realize OCR to support all complex characters in Tibetan and Khmer

  • ID OCR and MRZ Recognition

    Supports OCR and MRZ recognition of multi-national ID documents, passports, and driver's licenses based on computer vision technology and intelligent analysis of MRZ specifications.

  • Computer vision AI technology

    Provides Visual Restoration Reader, supports books information retrieval and contents full-text search, automatic collation of multiple editions, multiple Romanization and transliteration, intelligent proofreading, massive thesaurus, intelligent segmentation and online translation.

Partners