Extract Text

Text extraction in images

Settings

Field	Descriptive
Title	Component Title
Object Identifier Name	Unique name of the object to extract text from the image (variable)
Provider	Which provider will be used for the content detection (Tesseract (free) / Google)
Language	What language the content in the image is in. Currently it is possible to use Portuguese and English.
Capture Mode	Determines where the image will come from. This can be a link of an image on the Internet or a file.
File/URL	Enter the file path or the image link. This setting depends on the selected Capture Mode.

Example

To perform text extraction from an image, a few steps must be followed:

1 - Select the "Extract Text" object from the "OCR" menu, with "Google" as the provider, and drag it to your working area.

• First you must define a unique name for the object in the "Object Identifier Name" field and select the capture mode.

  • If you select file as the way to extract the text from the image, you must enter the file path.

  • Caso selecionado o link da imagem como a forma de captura do texto em imagem, deverá informar o link. - If you select the image link as the way to extract the text from it, you must enter the link.

• Next, if you do not have a utilization plan for the IDE, you should fetch the .json file that contains the key settings parameterized in Google and enter the language acronym. If you do have a utilization plan, you can choose to consume the available credits (which can be tracked directly in the Orchestrator).

If you select "Google" as the provider and do not enter the access key, this object will only be executed if your company has OCR credits available.