Historic Hood River
OCR and Script
Notes
This scan is just the reverse of a nice 1907 Barrett School class photo we saw in this post, but it provides a good illustration of some important technology.
When we started scanning our collection in 2011 I made the decision not to spend significant time transcribing names, trusting technology would eventually be able to automate that tedious process. That day is rapidly approaching. Recent versions of MacOS and iOS have started automatically doing OCR (optical character recognition) on images as they are displayed, including cursive writing. It doesn’t do a great job yet, but I have already used it to help prepare HHR posts. For example, this is the text which the software read from this image:
190%
8th qrads/-Ween Davidson
7./ Lent Shoemaker
3. Lavona Moore
11 – Claude Thompson
5 – Lester Murphy
6- Earl Moses
7- Ellen Moses
5. Professor murphy
9- grace Sherrieb.
MUSEUM
1907
I’d rate this at about 50% right now, but it’s not hard to imagine it being >90% in a year or two. Even at 50% this would save quite a bit of time if I used it as the base for a transcription. More importantly, this capability can feed into searches. If I search my image folder for “Sherrieb” this image would match, even though it does not have a “Sherrieb” keyword. All the writing on our images will eventually be available for keyword searches without needing to explicitly type keywords, which is what I was hoping for in 2011 when we started the arduous task of scanning and keywording >20,000 images. Though the image has the keywords “1907, Barrett, Barrett School, Group Photos, People, Schools, Students”, it could now be returned with a search for any of the students’ last names. Carefully assigned keywords are still the main tool for finding images in our collection, but a powerful new tool is emerging.
L. E.
It is disappointing that google-docs does not have a good key word search.
I have been thinking about using SmugMug for some of my history photos just because it does have a word search capability.
I use your tags here on HHR all the time. It is a great feature. I know it is time consuming for you Arthur, but in the long run, it saves time.
Arlen L sheldrake
I like the reformatted welcome page and importance the Museum has placed on this service.