Does OCR work with JPEG or BMP image format??

[Bauzi]

Hi,
We are evaluating GdPicture.Net with OCR PlugIn in order to use it in our development activity.
Running your sample C# OCR we can OCR only the TIFF format and not the other image formats (like JPEG or BMP), but transform it to a PDF file the sample runs well enough.
The images are created from a screen shot. Is this the problem? But the same images are correctly OCR by another OCR engine.
For us is important to understand if GdPicture.Net is usefull o not to our development because we'll be able to support the OCR process for all kind of image formats.
To help you to undarstand our problem, I attach an image file that GdPicture doesn't OCR.
Thanks
Anna

[Loïc]

Hi,

The OCR can work wit all image format supported by GdPciture technology: http://www.gdpicture.com/solutions/Image-Formats.php

For JPEG, the accuracy can decrease because this format destroy some data especially characters ! So I suggest you to use high quality compression rate for JPEG.

Also, print-screen usually provide very good quality image but too low resolution for image analysis (90 DPI). What I can suggest you is to scale the image before starting OCR process. In GdPicture.NET this can easily be done using the Scale() method.

IE:

Code: Select all

oGdPictureImaging.Scale(ImageID, 200, Drawing2D.InterpolationMode.HighQualityBicubic)

A simple call to this function can considerably multiply the accuracy of the engine. Also, there is a lot of pre-processing function which can help you to gain quality. You can discover these functions browsing the reference guide or trying the code sample provided in our package.

Hope this helps !

With best regards,

Loïc Carrère

[Bauzi]

Thank you Loïc. Now my sample works better.
Anna