Hi,

if you know what's on the slides and when, the best strategy would be to add an 
MPEG-7 catalog just as the OCR service is doing, containing the time segments 
and the text within each segment. Take a look a the classes 

- VideoSegmenterWorkflowOperationHandler and
- TextAnalysisWorkflowOperationHandler

to see what is happening. If you provide your own catalogs and get the 
flavors/tags right, you can easily implement your own algorithm or way of 
providing segments and texts.

Tobias

On 29.06.2012, at 21:03, Бисер Миланов <[email protected]> wrote:

> Greetings,
> 
> I was wondering do we have control over the OCR that is being done on the 
> slides, apart from inserting new dictionaries? For instance, if I already 
> know what is on the slides, I don't even need OCR in the first place -- I 
> would just fill the information out. To be more precise, can I edit the 
> strings that are included in the media package that are generated from the 
> OCR?
> 
> Regards,
> Biser Milanov
> _______________________________________________
> Matterhorn-users mailing list
> [email protected]
> http://lists.opencastproject.org/mailman/listinfo/matterhorn-users

_______________________________________________
Matterhorn-users mailing list
[email protected]
http://lists.opencastproject.org/mailman/listinfo/matterhorn-users

Reply via email to