A month back I posted about my progress with the automatic English Heritage plaque transcription project using Optical Character Recognition (using tesseract) and Python for OpenPlaques. The post mentions a monthly cash prize for progress towards a solution…
A few days back Jonathan Street announced his entry in the challenge’s thread – he’d beaten my initial average error of 709 (it was in part designed to be easy to beat!) and quickly brought it down to 33.4. Jonathan becomes the winner of the A.I.Cookbook’s first challenge, the challenge now rolls on to this month and the same prize is offered.
I’ll be presenting the results at the open day for the Open Plaques project (sponsored by the Royal Society of the Arts) on 25th September and I hope to be able to demonstrate that, for a few plaques at least, we can automatically get a good transcription.
In Jonathan’s write-up he describes the main steps and includes full Python source:
- image pre-processing to find the blue regions
- restricting tesseract’s character set
- spell checking
- word clean-up (to fix things like dates)
He’s taken some of the ideas I listed in the wiki and taken them further – I’m particularly happy with the blue region detection as that felt like an obvious first step that I hadn’t attempted.
In the thread there’s also a note by Andrew Elwell – he’s using OCR to update @lhcstatus (for the Large Hadron Collider – with >1,000 followers!) by screen scraping their graphical update screen.
David Rawlinson also posted in the thread about some ideas taken from his experience with automatic number plate recognition (ANPR) so we can correct mis-recognised characters (e.g. 0O0 and 1lLiI are easily mis-recognised by OCR!).
The competition runs on, the new deadline is Thursday September 23rd so I can present our progress on the 25th at the OpenPlaques event. If nobody beats Jonathan’s result by then then he becomes the winner by default. I’ll be adding more ideas for improving the result into the main wiki page. Join the Google Group if you’d like to offer ideas and get involved.
I’m working with the OpenPlaques folk to create a system that automatically ‘reads’ images of English Heritage plaques and extracts a transcript of the plaque’s text. This is a classic optical character recognition project. Here’s a simple example (thanks Fiery Fred): The text is very easy for a human to read but very hard for …
Read More →
Here we’ll look at building Headroid1 in a few hours – a face tracking 2-axis robot head controlled by Python and open source modules. This is what the finished system will look like: An earlier demo was presented on my blog as Headroid1 – A Face Tracking Robot, here’s a video demo: Requirements: An afternoon …
Read More →
When I first played with openCV I had no idea how good the facial detection would be, or how fast it might run on my MacBook. I’m recording this demo so you’ll know what to expect… pyOpenCV is the Python binding to the open source openCV (originally created by Intel for vision research). It comes …
Read More →
This is a quick progress report on my webservice for optical character recognition using the open source Tesseract engine. This builds on my post a month back ‘Tesseract OCR to read plaques‘. The immediate goal is to let the OpenPlaques folk have an automatic service which machine-reads English Heritage Plaques (blue plaques – very common …
Read More →
Here is a rather neat demo of the advantage of tracking a face whilst performing speech recognition – if the user is looking at the computer then the computer knows to listen. This is common sense to a human but for a computer with just a microphone input it has to listen to everything, not …
Read More →
Back at Christmas I was speaking to John Graves about his Open Allure DS PhD project – a conversational interface written in Python. The project has moved wonderfully forward over the past few months, I’ll summarise some of the features of Open Allure here. Sidenote – if you prefer podcasts then John was recently interviewed …
Read More →
Here’s the new logo for the wiki: This uses the OCR A font, designed in 1968 to allow both machines and humans to read text (available by free download). An OCR B was released in Europe at the same time which is easier for humans to read but I didn’t think it was as pretty …
Read More →
I had a tricky time installing openCV 2.1 on my 10.5 Mac (Leopard) recently, the build instructions aren’t brilliant in the wiki. Here are my notes. I used the Mac notes as a guide and followed the new-style CMake system to build from scratch. I made sure during ‘cmake -G’ that new-style Python support was …
Read More →
I’ve just setup the A.I. Cookbook Google Group, the goal will be to bring like-minded souls together so we can chat about ways of making smarter software. If you came here from the original project description post (New project – a practical Artificial Intelligence Cookbook) on my blog then you’ll know what’s in store. Do …
Read More →