Have you given the possibility of automatic link recognition any thought? It would sure be handy if you could just click a link (like you can do on the `real' website) to whiz to a new page... It should not be to hard, since the character set is mostly fixed (at least within the same teletekst-channel) and easy to scan (you only need to decode decimal numbers, the rest is not useful anyway). If nobody has an interest in pursuing this, I might give it a go.
Cheers//Frank
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
we have had contact with the NOS about the protocol which they use in their windows version of this teletext program. their version uses indeed link recognition, but they said everything was done by the HTTP protocol, so it would be some kind of image map.
It would be nice if someone could figure out how to read this information. Probably is's a server-side image map. In that case, clickable links would be a piece of cake and wouldnt require character-recognition algorithms.
And secondly, this protocol could be used too with other teletext feeds, whereas a ocr algorithm would require mucho config for every new feed (other font, etc).
cu, Arjan
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Have you given the possibility of automatic link recognition any thought? It would sure be handy if you could just click a link (like you can do on the `real' website) to whiz to a new page... It should not be to hard, since the character set is mostly fixed (at least within the same teletekst-channel) and easy to scan (you only need to decode decimal numbers, the rest is not useful anyway). If nobody has an interest in pursuing this, I might give it a go.
Cheers//Frank
we have had contact with the NOS about the protocol which they use in their windows version of this teletext program. their version uses indeed link recognition, but they said everything was done by the HTTP protocol, so it would be some kind of image map.
It would be nice if someone could figure out how to read this information. Probably is's a server-side image map. In that case, clickable links would be a piece of cake and wouldnt require character-recognition algorithms.
And secondly, this protocol could be used too with other teletext feeds, whereas a ocr algorithm would require mucho config for every new feed (other font, etc).
cu, Arjan