I presume someone has tried the old trick of extracting the pictures
from PDF to make sure they didn't just apply the redaction over the
top of the pictures? Sounds unlikely, but it's been done before...
On 19 Jun 2009, at 16:23, Brian Butterworth wrote:
2009/6/19 John O'Donovan <[email protected]>
We did something similar, but it's the low tech version... :o)
http://news.bbc.co.uk/1/hi/uk_politics/8106044.stm
with things people have found being published here...
http://news.bbc.co.uk/1/hi/uk_politics/8106650.stm
I like what the Guardian have done with this - been playing with it...
After an hour I start wondering about using OCR software...
Does anyone know of a command line OCR tool that I could use?
Something that works with PHP perhaps?
As it seems very easy to get at the images from the http://mps-expenses.guardian.co.uk/page/X/
pages as there is only one image in the whole document. Running
OCR will generate lots of crud, but it could be matched against the
human input to act as validation.
Cheers,
John O'Donovan
Chief Technical Architect
BBC Future Media & Technology (Journalism)
BC3 C1, Broadcast Centre, 201 Wood Lane, London
http://news.bbc.co.uk/
http://news.bbc.co.uk/sport/
http://news.bbc.co.uk/weather/
From: [email protected] [mailto:[email protected]
] On Behalf Of Brian Butterworth
Sent: 19 June 2009 09:46
To: [email protected]
Subject: [backstage-developer] Nice bit of crowd-sourcing
Nice bit of crowd-sourcing I think here:
http://mps-expenses.guardian.co.uk/
Shame the app's not AJAX, would been easier to use that way, but
generally a great way of checking 77252 pages of documents.
Kind-of-wondering why Auntie didn't do it first, but...
All the best
Brian Butterworth
--
Brian Butterworth
follow me on twitter: http://twitter.com/briantist
web: http://www.ukfree.tv - independent digital television and
switchover advice, since 2002
