Symbol Detection in Video (looking for guidance)

edit

Hi,

I'm just starting out learning openCV and just looking for some guidance on a problem im playing with.

Essentially what I want to be able to do is detect a sentence of text from a video stream(localization). I've learnt how to capture the mat frames and now im onto the bit of actually doing the detection where im a bit lost.

The key thing is that the sentence of text always starts with the same symbol (the text will vary). What I want to do is detect and bound the text, cut it from the image frame and do some post processing on the text.

I know there are various alogrithms available but I dont know which i should be focusing on e.g. classifers, features etc?

If some one could guide me with some practical advice on even how to detect the symbol that would be most appreciated. I'm developing on iOS but I guess it doesnt have to be iOS specific.

Thanks

Mike