Summary

The OmniPage OCR Engine cannot process e-sized images.

Resolution

There is a hotfix available for Laserfiche products that include the OmniPage OCR engine. The fix includes updated versions of the following files:

• BPOmniOCR.exe (version 7.2.1.3)
• OmniOCRWrapper.dll (version 7.2.1.3)
• OmniPage32.lfo (version 1.0.0.3)

To update the OmniPage OCR component

1. Close any applications that may be using the OmniPage OCR engine.
2. Click the following link to download a zip file containing the hotfix files.
   Hotfix_SCR27647.zip
3. Replace your existing versions of BPOmniOCR.exe and OmniOCRWrapper.dll with the updated versions contained in the zip file. The files are located at "C:\Program Files\Common Files\Laserfiche\Batch Processor\BPOmniOCR."
4. Register the new version of OmniOCRWrapper.dll.
   1. Click Start and then click Run.
   2. Type the following and then click OK:

      regsvr32 "C:\Program Files\Common Files\Laserfiche\Batch Processor\BPOmniOCR\OmniOCRWrapper.dll"

5. Replace your existing version of OmniPage32.lfo located in your Laserfiche Client installation folder with the updated version. By default, the Laserfiche Client is installed at "C:\Program Files\Laserfiche\Client."

Laserfiche works around this size limitation in the OCR engine by processing an e-sized image in sections. Laserfiche generates OCR text for each section and then merges the results together to form the final OCR text pages. Be aware that this can occasionally result in words being split in the middle.

However, text search functionality will not be affected as the search engine will merge the split words together during searches.

Related Links

BPOmniOCR.exe (version 7.2.1.3) also contains additional hotfixes not related to e-size processing. Please see the following Knowledge Base articles for more information.

1011725 OmniPage OCR Does Not Correctly Use the Image Orientation Displayed in the Laserfiche Client When OCRing.

1011840 Temporary OCR Files Are Not Automatically Removed.

1011841 OmniPage OCR Does Not Correctly Process Multi-Byte Western Languages.