image » ocr(image,saveVar)

Description

This command performs OCR (optical character recognition) on a specified image (image) and then saving the recognized text to a data variable denoted as saveVar. Note that Nexial is utilizing an external API for the OCR process, which is, by some standard, a fairly basic implementation. It works best on images that contains mostly text formed by block fonts (such as Arial, Verdana, Courier, Times New Romans, etc.). Handwritten text are less likely to be recognized correctly. In addition, images with complex format or nested table structure will less likely be recognized correctly.

Note that as of Release 3.5, this command also supports PDF files. Simply specify a PDF file for the image parameter to activate OCR for PDF file. Note that at this time, this command will only support a PDF file with 3 pages or less.

This command supports the following image formats:

1. JPEG (file extension .jpg or .jpeg)
2. PNG (file extension .png)
3. GIF (file extension .gif)
4. TIFF (file extension .tif or .tiff)
5. Bitmap Images (file extension .bmp)
6. PDF (file extension .pdf, as of Release 3.5)

For images containing pre-formatted section like table, Nexial will try to retain existing layout. However it will not be able to retain successive blank spaces or tabs.

This command is best use to detect key words or key phrases, such as names, addresses, status. title, etc. For text formatting, alignment, font sizes or color, this command will not be able to provide sufficiently towards such ends.

For more advance text scraping automation, consider using vision » saveText(profile,image,var). It uses AWS’ Rekognition, which is paid service.

Parameters

• image - the full path of the image/pdf to scan
• saveVar - the data variable to store the OCR text

Example

Suppose we have the following image:

We can extract text from this image with something like this: Script:

See Also

• vision » saveText(profile,image,var)