Text recognition v2

AI-generated Key Takeaways

outlined_flag

• The ML Kit Text Recognition v2 API recognizes text in Chinese, Devanagari, Japanese, Korean, and Latin scripts and can automate data entry for documents like credit cards and receipts.

• It analyzes text structure by identifying blocks, lines, elements (words), and symbols, returning bounding boxes, corner points, and confidence scores for each.

• The API supports real-time text recognition on various devices and can identify the language of the recognized text.

The ML Kit Text Recognition v2 API can recognize text in any Chinese, Devanagari, Japanese, Korean and Latin character set. The API can also be used to automate data-entry tasks such as processing credit cards, receipts, and business cards.

iOS Android

Key capabilities

• Recognize text across various scripts and languagesSupports recognizing text in Chinese, Devanagari, Japanese, Korean and Latin scripts
• Analyzes structure of textSupports detection of symbols, elements, lines and paragraphs
• Identify language of textIdentifies the language of the recognized text
• Real-time recognitionCan recognize text in real-time on a wide range of devices

Text structure

The Text Recognizer segments text into blocks, lines, elements and symbols. Roughly speaking:

• a Blockis a contiguous set of text lines, such as a paragraph or column,

• a Lineis a contiguous set of words on the same axis, and

• an Elementis a contiguous set of alphanumeric characters ("word") on the same axis in most Latin languages, or a word in others

• an Symbolis a single alphanumeric character on the same axis in most Latin languages, or a character in others

The image below highlights examples of each of these in descending order. The first highlighted block, in cyan, is a Block of text. The second set of highlighted blocks, in blue, are Lines of text. Finally, the third set of highlighted blocks, in dark blue, are Words.

For all detected blocks, lines, elements and symbols, the API returns the bounding boxes, corner points, rotation information, confidence score, recognized languages and recognized text.

Example results

Photo: Dietmar Rabich , Wikimedia Commons , "Düsseldorf, Wege der parlamentarischen Demokratie -- 2015 -- 8123" , CC BY-SA 4.0