Updates on OCR and Cloud vision...
We worked for months on it, and now it is available: Make searchable PDFs from scans and images for free with our Online OCR form and the OCR API.
This works with PDFs with images inside (such as the ones you get from scanning documents) but also works with images (jpeg/png/gif). The image is converted to a searchable PDF that you can download.
Try it here and select the “Create visible text layer” option. This way you can easily see the text layer that gets added during the conversion.
Our latest Online OCR and OCR API update fixes the “left to right bug” with Arabic OCR that many users reported. Now the API returns the text in the correct order (right to left), and the online OCR forms display the arabic text right-aligned.
Thank you for your patience while we worked on this issue.
Today’s Online OCR and tomorrows upcoming OCR API update adds four new languages to our service, and improves the recognition for all other languages:
New languages:
All languages: Improved character recognition
Release dates:
The German Public TV station “Norddeutscher Rundfunk” reports that they were offered data that includes the surfing habits of three million German citizens. It seems this data was, at least partly, collected by the “Web of trust” (WOT) browser extensions. This is as bad as it sounds. The TV station was able to use this data to identify the browsing habits of individual persons – including high-ranking German and EU politicians.
The November Free OCR API update adds two often requested features:
Easier authentication via request header
Base 64 support for sending images
The API url and format has not changed. Also, we still support the old authentication methods. So, unless you decide to make use of the new features, this upgrade will be 100% transparent to you.
In more details:
