Camejo Hospital Medical Records Project
Moderator: peterZ
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Camejo Hospital Medical Records Project
I did a little photogrammetry to find out the thickness of that really thick book, which will become a part of the cradle design for your scanner. Looks like 2.75 inches will be enough.
Re: Camejo Hospital Medical Records Project
Forensic science for book scanning! Excellent idea, and thank you!
-- DonnaA
-- DonnaA
Re: Camejo Hospital Medical Records Project
]A huge thank you to Daniel for creating "Daniel's 5th Generation DIY Book Scanner" for the Camejo Hospital Medical Records Project (http://www.diybookscanner.org/forum/vie ... +5th#p7145)
He shipped the rig to me, in Denver, and our team assembled and disassembled the rig several times to become completely familiar with the process. We also applied polyurethane to each piece of wood to protect the wood from the high humidity weather conditions of Leogane, Haiti. Here is a photo of all of the wooden pieces, labeled and laid out so that we can annotate the photo. We put each piece of hardware in a plastic baggie and labeled those as well.
He shipped the rig to me, in Denver, and our team assembled and disassembled the rig several times to become completely familiar with the process. We also applied polyurethane to each piece of wood to protect the wood from the high humidity weather conditions of Leogane, Haiti. Here is a photo of all of the wooden pieces, labeled and laid out so that we can annotate the photo. We put each piece of hardware in a plastic baggie and labeled those as well.
Last edited by Anonymous on 14 Dec 2010, 01:14, edited 1 time in total.
Re: Camejo Hospital Medical Records Project
Working with the LED lighting strip that Daniel provided, we did some experimenting to determine the height of the lights that would eliminate glare. We concluded with the lighting strip at a minimum of 55" above the table top, positioned as you see in this photo. The lighting strip is supported by microphone stands because that's what I had on-hand here in Denver. When in Leogane, there are several options for supporting the lighting strip, including lamp-stands that are used for examination lights or surgery lights.
Re: Camejo Hospital Medical Records Project
Here is a photo taken with the "Daniel's 5th Generation DIY Book Scanner" rig, using the two Olympus E-P1 cameras with the settings I summarize below. The data that you see here is representative -- it is in my handwriting, using blue ink, on paper that is the size of the copybooks, in the columnar format that is used at Camejo Hospital for record-keeping. The text is from the bible, since the Author has put that text in the public domain
Olympus E-P1 settings that we used:
ISO 200; Fstop 4.8; shutter speed 40; Fluorescent1; Manual Focus; RAW file format; compensation 0.0 .
Olympus E-P1 settings that we used:
ISO 200; Fstop 4.8; shutter speed 40; Fluorescent1; Manual Focus; RAW file format; compensation 0.0 .
Last edited by Anonymous on 14 Dec 2010, 01:14, edited 1 time in total.
Re: Camejo Hospital Medical Records Project
With the Olympus E-P1 camera that we're using, and the configuration of the camera arms, we found that the access door for the camera batteries and the camera memory card was partially blocked by the top corner of the camera arm. It would be ideal if we did not have to move the cameras to change batteries and memory cards, once we have the cameras in great positions relative to the copybook pages. After consulting with Daniel, we decided to trim off that corner on both of the large camera arms.
Here are two photos:
(1) "Before", a close-up showing how the camera arm blocks the access door for the camera, and
(2) "After" we trimmed the corner, Jodel demonstrating the open access door while the camera is in place.
Good teamwork!
Here are two photos:
(1) "Before", a close-up showing how the camera arm blocks the access door for the camera, and
(2) "After" we trimmed the corner, Jodel demonstrating the open access door while the camera is in place.
Good teamwork!
Re: Camejo Hospital Medical Records Project
Looking very nice, Donna! The painted scanner is pretty darn lovely looking.
Is that paper meant to be white? You might want to try a different white balance setting. It looks like it should be perfectly readable regardless of what the white balance is set to, though.
Is that paper meant to be white? You might want to try a different white balance setting. It looks like it should be perfectly readable regardless of what the white balance is set to, though.
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.
Re: Camejo Hospital Medical Records Project
Adjusting white balance ... here's another photo:
Re: Camejo Hospital Medical Records Project
It looks like that's actually adjusting the brightness.
The "white balance" is a bit confusing at first. It took me awhile to figure out the right thing to do when I was getting set up. What that refers to is the colour of the light; the camera has to compensate for that colour in order to make sure that everything else in the scene is the right colour. The different settings on the camera refer to different kinds of lights, which are coloured differently. You can set the white balance in the camera, and if you captured raw then you can also change it in the Olympus Master software or in Photoshop.
It looks like Fluorescent1 is too blue for the lights you're using. Try experimenting with the other settings to see if there's one that works better.
Dan, can you tell me if there's a rated temperature for your LEDs?
I can't find the exact number you need without the raw, Donna, but here's an estimate of what the image could look like with the white balance corrected.
The "white balance" is a bit confusing at first. It took me awhile to figure out the right thing to do when I was getting set up. What that refers to is the colour of the light; the camera has to compensate for that colour in order to make sure that everything else in the scene is the right colour. The different settings on the camera refer to different kinds of lights, which are coloured differently. You can set the white balance in the camera, and if you captured raw then you can also change it in the Olympus Master software or in Photoshop.
It looks like Fluorescent1 is too blue for the lights you're using. Try experimenting with the other settings to see if there's one that works better.
Dan, can you tell me if there's a rated temperature for your LEDs?
I can't find the exact number you need without the raw, Donna, but here's an estimate of what the image could look like with the white balance corrected.
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.
Re: Camejo Hospital Medical Records Project
great work !!! unfortunatelly you might face some difficulties in taking out the data automatically.
Just out of curiousity, why didn't you unbind them and use a cut sheet scanner ? For indexing matters we use Kofax but there is a, also SimpleIndex which might have interesting features. Most of the softwares for indexing purposes, cost a lot of money. But, from my experience they;re really worth it when you can automatise your work.
From the samples you sent, or from what I can see from the two photos, fully manual data retrieval or also combining auto features, will ad up to the same ammount of time.
Batch features are good when you have homogen documents, or with a high grade of homogenuity, but from your photos it seems that data retrieval will be done manually.
Just out of curiousity, why didn't you unbind them and use a cut sheet scanner ? For indexing matters we use Kofax but there is a, also SimpleIndex which might have interesting features. Most of the softwares for indexing purposes, cost a lot of money. But, from my experience they;re really worth it when you can automatise your work.
From the samples you sent, or from what I can see from the two photos, fully manual data retrieval or also combining auto features, will ad up to the same ammount of time.
Batch features are good when you have homogen documents, or with a high grade of homogenuity, but from your photos it seems that data retrieval will be done manually.