Xiangji Logo

Image Translation API

Translate Image

Recognize and translate the text in an image, directly return the translated image where the text is translated into the specified language, while preserving the original image's layout and style.

POST https://api.piclaza.com/task/v1/image/translate

Request body

Content-Type: application/json

image_url string required[1]
Image link

image_urls array of string required[1]
List of image link

image_file_b64 string required[1]
Image file content, base64 encoding (format: data:[<mediatype>][;base64],<data>)

image_name string
Corresponding file name of image_file_b64

source_language string required
Source language. Currently support:

  • CHS: Chinese (Simplified)
  • CHT: Chinese (Traditional)
  • ENG: English
  • ESP: Spanish
  • JPN: Japanese
  • KOR: Korean

target_language string required
Target language. See Supported languages for more information.

translation_vendor string
The text translation engine, currently supported:

  • Google: (Default) Widely used in multiple languages around the world
  • Papago: Good Korean translation
  • DeepL: Available in multiple languages, excellent in slang and dialect translation
  • ChatGPT: Use the best Large Language Model(gpt-4o), with accurate translation
  • Aliyun: Suitable for Chinese translation

qos string
Quality of Service, currently supported:

  • BestQuality: (Default) Preference for good translation quality
  • LowLatency: Preference for fast translation speed

1 image_url/image_urls/image_file_b64 are three forms of input image parameters, and one of the three parameters should be set

Example

{
    "image_url": "https://m.media-amazon.com/images/I/71Y0duLL0jL._AC_SL1500_.jpg",
    "source_language": "ENG",
    "target_language": "KOR",
    "translation_vendor": "Google"
}

Response body

Content-Type: application/json

code integer
Status code, non-0 indicates an error. See Response status code for more information

message string
Error message

data map
Returned data

results array of Task object

Example

{
    "code": 0,
    "data": {
        "results": [
            {
                "code": 0,
                "task_id": "827e1d0a0a68771e",
                "type": "image-translation-google",
            }
        ]
    }
}

The Task Object

code integer
Status code, non-0 indicates an error. See Response status code for more information

message string
Error message

task_id string
Task id

type string
Task type

Query the results of image translation

POST https://api.piclaza.com/task/v1/image/translate/result

Parameters

Content-Type: application/json

task_ids string required
List of task ids

verbose boolean
Return detailed OCR data or not, default is false.

Example

{
    "task_ids": ["863ab3836328f4ec"],
    "verbose": true
}

Responses

Content-Type: application/json

code integer
Status code, non-0 indicates an error. See Response status code for more information

message string
Error message

data object
Returned data

results array of Image Translatioin Result object

Example

{
    "code": 0,
    "data": {
        "results": [
            {
                "code": 0,
                "task_id": "863ab3836328f4ec",
                "request_id": "863ab3836328f4ec",
                "source_language": "ENG",
                "target_language": "KOR",
                "original_image": "https://i.tosoiot.com/1cbb77123add49a2.jpg",
                "translated_image": "https://i.tosoiot.com/1cbb77123add49a2-f.jpg",
                "text_removed_image": "https://i.tosoiot.com/1cbb77123add49a2-i.jpg",
                "ocr": [
                    {
                        "id": 1,
                        "source": "External function",
                        "target": "외부 기능",
                        "ori_bounding_box": "[[173,133], [617,133], [617,177], [173,177]]",
                        "bounding_box": "[[173,133], [617,133], [617,177], [173,177]]",
                        "font_size": 35,
                        "direction": "h",
                        "color": "255,255,255",
                        "stroke_color": "255,255,255",
                        "line_count": 1,
                        "align": "left"
                    },
                    {
                        "id": 5,
                        "source": "The watch comes with an external speaker, which can play the recording directly without other equipment.",
                        "target": "시계에는 외부 스피커가 있어서 다른 장비 없이도 녹음 내용을 직접 재생할 수 있습니다.",
                        "ori_bounding_box": "[[ 93,208], [697,208], [697,297], [ 93,297]]",
                        "bounding_box": "[[ 93,208], [697,208], [697,297], [ 93,297]]",
                        "font_size": 24,
                        "direction": "h",
                        "color": "255,255,255",
                        "stroke_color": "255,255,255",
                        "line_count": 2,
                        "align": "center"
                    }
                ]
            }
        ]
    }
}

The Image Translation Result Object

code integer
Status code, non-0 indicates an error. See Response status code for more information

message string
Error message

task_id string
Task id

user_id integer
User id

type string
Task type

status string
Task status (ok/running/failed)

original_image string
Original image link

translated_image string
Translated image link

text_removed_image string
Text removed image link

ocr list of Image OCR object
List of recognized text information

The Image OCR Object

Coordinates representation: the upper left corner of the image is (0, 0); the x-axis is from left to right; the y-axis is from top to bottom

(0,0)---x--->
  |
  |
  y
  |
  |
  v

id integer
id

source string
Recognized original text

target string
Translated text

ori_bounding_box string
Original text box border

bounding_box string
Translated text box border

font_size integer
Text font size

direction integer
Text writing direction (h: horizontal, v: vertical) Some languages ​​have a tradition of vertical writing, such as Chinese and Japanese

color string
Text color (RGB)

stroke_color string
Text stroke color (RGB)

line_count integer
Line count of text

angle integer
Rotation angle (clockwise rotation from the positive direction of the x-axis, default 0)

align string
Alignment (left/right/center)

Getting Started
Notes