Looking for a way to sequentially OCR scan many many photos
August 28, 2024 11:04 AM Subscribe
A friend who's far away from me has many pages of typewritten material and they're willing to snap photos of each page, can I feed them into some program that will perform sequential OCR?
Extended situation:
Someone has a lot of pages that were typed, not handwritten, so I have good hopes that OCR would be able to parse them correctly. They can't mail them to me, and they don't want to risk putting them through a scanner, but they are willing to use their phone to send me a pic of each page against a dark surface.
Ideally, what I'd like is a program or process where I can queue up the photos one-by-one and feed them into a program that will perform OCR so that each screenshot gets it's own separate page in a Word document.
So Photo_001 ---> Page 1
Photo_002 ---> Page 2
Photo_003 ---> Page 3 and so on.
Does a program exist that can do this? I'm using an iPhone and Mac OS X. I would normally receive these photos on my iPhone but if it's critical I can probably arrange to get them another way. Also, I'm willing to pay some money for this, but I'm not looking for a whole enterprise-wide system, just a solution to this one unusual problem. Thanks very much!
Extended situation:
Someone has a lot of pages that were typed, not handwritten, so I have good hopes that OCR would be able to parse them correctly. They can't mail them to me, and they don't want to risk putting them through a scanner, but they are willing to use their phone to send me a pic of each page against a dark surface.
Ideally, what I'd like is a program or process where I can queue up the photos one-by-one and feed them into a program that will perform OCR so that each screenshot gets it's own separate page in a Word document.
So Photo_001 ---> Page 1
Photo_002 ---> Page 2
Photo_003 ---> Page 3 and so on.
Does a program exist that can do this? I'm using an iPhone and Mac OS X. I would normally receive these photos on my iPhone but if it's critical I can probably arrange to get them another way. Also, I'm willing to pay some money for this, but I'm not looking for a whole enterprise-wide system, just a solution to this one unusual problem. Thanks very much!
So as not to abuse the edit window, it does look like Adobe Scan can use existing files to create new scans.
"Scan’s integration with the device photo library allows you to create scans from existing photos or digital documents. Scan uses the document detection feature to distinguish documents from selfies, artwork, and other non-document photos."
posted by papayaninja at 11:27 AM on August 28 [1 favorite]
"Scan’s integration with the device photo library allows you to create scans from existing photos or digital documents. Scan uses the document detection feature to distinguish documents from selfies, artwork, and other non-document photos."
posted by papayaninja at 11:27 AM on August 28 [1 favorite]
Tropy is an open source tool often used by researchers to organize and transcribe photographed documents - it does not have a native OCR capability but there are people who have developed workflows for keeping image files and text transcripts together.
posted by zepheria at 11:39 AM on August 28 [1 favorite]
posted by zepheria at 11:39 AM on August 28 [1 favorite]
Google Stack on Android does this natively. You just take a series of pictures and it all gets added in sequence to the same document and sutomatically gets converted to one large pdf.
posted by MiraK at 11:46 AM on August 28
posted by MiraK at 11:46 AM on August 28
I have a folder of about 500 photos, one of each page of a rare book. I surely would love a way to do this on my Mac, since hand-transcribing it is maddeningly slow.
The full Adobe Acrobat has pretty good OCR, which will embed the text into a PDF -- but you need to manually review & correct the text, which is possibly slower than typing it by hand in the first place.
posted by wenestvedt at 12:07 PM on August 28
The full Adobe Acrobat has pretty good OCR, which will embed the text into a PDF -- but you need to manually review & correct the text, which is possibly slower than typing it by hand in the first place.
posted by wenestvedt at 12:07 PM on August 28
I use a really old version of ABBYY OCR to do batches of OCR like you're describing. At least on the old version I use, the tools for correcting the OCRed text are really great compared to Acrobat. Seems like ABBYY has a trial version which you could see if it fits your purpose.
Not sure how commandliney you are, but there's also Tesseract OCR which is pretty great. I use it for stuff where I want a program I've written to OCR something.
Also just as an aside, a flatbed scanner really is pretty low risk in terms of damage to the documents, and will provide much better images than something shot with a phone.
posted by gregr at 12:55 PM on August 28
Not sure how commandliney you are, but there's also Tesseract OCR which is pretty great. I use it for stuff where I want a program I've written to OCR something.
Also just as an aside, a flatbed scanner really is pretty low risk in terms of damage to the documents, and will provide much better images than something shot with a phone.
posted by gregr at 12:55 PM on August 28
There's a very good (and free! without ads!) iPhone app called QuickScan that can do this for you. It creates a PDF with both the original images and the OCR'd text so it's searchable. (You can also output plain text if you want.)
There's also an image/PDF import feature which I imagine will let you OCR an already-existing document, I haven't tried it though.
It's also quite good for scanning documents, it automatically takes the photos and crops and tweaks the contrast (or doesn't, if you don't want it to).
posted by neckro23 at 4:54 PM on August 28
There's also an image/PDF import feature which I imagine will let you OCR an already-existing document, I haven't tried it though.
It's also quite good for scanning documents, it automatically takes the photos and crops and tweaks the contrast (or doesn't, if you don't want it to).
posted by neckro23 at 4:54 PM on August 28
Okay, on review the OCR feature is no longer free with QuickScan. I guess I got it as an early adopter.
posted by neckro23 at 5:13 PM on August 28
posted by neckro23 at 5:13 PM on August 28
What about the scan document feature that’s built-in to iPhone Notes?
posted by wyzewoman at 12:03 PM on August 29
posted by wyzewoman at 12:03 PM on August 29
« Older What's a fun family trip near the Bay Area, CA? | Seeking cloud-based backup protection Newer »
You are not logged in, either login or create an account to post comments
If so, maybe take a look at this wirecutter guide. Perhaps one of those apps also allows for image uploading, but I'm not sure.
posted by papayaninja at 11:25 AM on August 28