# Printed character recognition API documentation

# Interface description

• The end-to-end text recognition system based on the deep neural network model and iFlytek's self-developed industry-leading optical character recognition technology can convert printed fonts in pictures (from sources such as scanners or digital cameras) into scanned bodies and complex natural scenes. Text recognition is directly converted into editable text. Supports 32 languages including Chinese、English、Hungarian、French、German、Spanish etc.

• Demo
  java
  python
  go
  nodejs
  C#

• When you use printed character recognition, please follow these requirements:

# Authentication Description

When using the business interface, the requester needs to sign the request, and the server verifies the validity of the request through the signature.

# Authentication Method

Add authentication-related parameters after the request address.Please note that the values that affect the authentication result are URL, apiSecret, apiKey, and date. If you want to debug the authentication, you must debug according to the values given in the example.，The specific parameters are as follows:：

Authentication parameters:：

• Format of authorization parameter generation:

1）Get the interface keys APIKey and APISecret.
After creating an account in iFLYTEK open platform，please visit the console page to obtain 32-bit strings.
2）The format of the parameter authorization base64 before encoding (authorization_origin) is as follows：

api_key="$api_key",algorithm="hmac-sha256",headers="host date request-line",signature="$signature"

Where API _ key is the APIKey obtained on the console, algorithm is the encryption algorithm (only hmac-sha256 is supported), and headers is the parameter involved in the signature（see the note below）。
A signature is a string that uses an encryption algorithm to sign the parameters that participate in the signature and uses base64 encoding. See below for details.

3）The signature origin field (signature_origin) rule is as follows:

The original signature field is formed by splicing three parameters of host, date, and request-line according to the format.
The format of the concatenation is (\ n is a newline character with a space after ’:’)：

host: $host\ndate: $date\n$request-line

If

Requested url = "https://me-east-1.aicloudapi.com/v1/ocr"
date = "Wed, 07 Dec 2022 08:18:43 GMT"

Then the signature origin field (signature_origin) is:

host: me-east-1.aicloudapi.com
date: Wed, 07 Dec 2022 08:18:43 GMT
POST /v1/ocr HTTP/1.1

4）use that hmac-sha256 algorithm and combining the signature of the signature_origin by the apiSecret to obtain a signature dig signature_sha.

signature_sha=hmac-sha256(signature_origin,$apiSecret)

Where apiSecret is the APISecret obtained at the console.

5） Encode signature_sha with base64 encoding to get the final signature.

signature=base64(signature_sha)

If

APISecret = "apisecretXXXXXXXXXXXXXXXXXXXXXXX"	
date = "Wed, 07 Dec 2022 08:18:43 GMT"

Then the signature is

signature="J0D7cz4s+6lQpzNtT03BiZN1QEIhqZrSBKvk6W6nK5s="

6） Encode signature_sha with base64 encoding to get the final signature：

api_key="apikeyXXXXXXXXXXXXXXXXXXXXXXXXXX", algorithm="hmac-sha256", headers="host date request-line", signature="J0D7cz4s+6lQpzNtT03BiZN1QEIhqZrSBKvk6W6nK5s="

7）Finally, the authorization_origin is base64 encoded to obtain the final authorization parameter.

authorization = base64(authorization_origin)
Example result is：
authorization=YXBpX2tleT0iYXBpa2V5WFhYWFhYWFhYWFhYWFhYWFhYWFhYWFhYWFgiLCBhbGdvcml0aG09ImhtYWMtc2hhMjU2IiwgaGVhZGVycz0iaG9zdCBkYXRlIHJlcXVlc3QtbGluZSIsIHNpZ25hdHVyZT0iSjBEN2N6NHMrNmxRcHpOdFQwM0JpWk4xUUVJaHFaclNCS3ZrNlc2bks1cz0i

# Authentication Result

If the authentication fails, different HTTP Code status codes will be returned according to different error types, and the error description information will be carried. The detailed error description is as follows:：

# Request Parameters

When calling a business interface, the following parameters need to be configured in the Http Request Body. The request data is a JSON string.

example：

{
	"header": {
		"app_id": "your appid",
		"status": 3
	},
	"parameter": {
		"ocr": {
			"language": "language=de",
			"ocr_output_text": {
				"encoding": "utf8",
				"compress": "raw",
				"format": "json"
			}
		}
	},
	"payload": {
		"image": {
			"encoding": "jpg",
			"image": "iVBORw0KGg······",
			"status": 3
		}
	}
}

Request parameter description:

# Return Result

Return parameter example:

{
	"header": {
		"code": 0,
		"message": "success",
		"sid": "ocr000e583f@hu1847a3af5cd05c2882"
	},
	"payload": {
		"ocr_output_text": {
			"compress": "raw",
			"encoding": "utf8",
			"format": "json",
			"seq": "0",
			"status": "3",
			"text": "ewogICAiY2F......"
		}
	}
}

Returned parameter description:

The decoded information of the payload. OCR _ output _ text. Text field base64 is as follows, please pay special attention to:

pages.lines field**

pages.tables field

pages.checkboxes field

pages.seals field

pages.fingerprints field

pages.graphs field

pages.headers field

pages.footers field

pages.blocks field

pages.page_numbers field

pages.expressions field

pages.barcodes field

# language feature parameter list:

# Frequently Asked Questions

# What is the main function of printed character recognition?

Answer: Convert the printed text in the picture into text that can be encoded by the computer.

# What application platforms are supported for printed character recognition?

Answer: Web API application platform is currently supported.

# Are there any requirements for pictures in printed character recognition ？

Answer: The image format supports JPG, JPEG, PNG, BMP, webp and tiff formats, and the image file size shall not exceed 4MB after base64 encoding.