Has anyone here ever gone to the trouble of scanning a whole factory service manual with OCR so that it was a seachable pdf file?
I was thinking about it as I find the factory manuals a bit difficult to navigate as compared to other manuals. Sure would make finding what I was looking for a lot easier.
IP: Logged
04:57 AM
PFF
System Bot
infinitewill Member
Posts: 2344 From: Overland Park, KS U.S.A. Registered: Mar 2009
Interesting thought, that would be a huge undertaking though. Assuming you could scan and OCR a page every 5 minutes, probably optimistic (no OCR software is perfect) is would take about 75 hours to do.
Will
Adobe CTI
[This message has been edited by infinitewill (edited 03-31-2009).]
IP: Logged
07:01 AM
sjmaye Member
Posts: 2468 From: Hendersonville, TN USA Registered: Jun 2003
Interesting thought, that would be a huge undertaking though. Assuming you could scan and OCR a page every 5 minutes, probably optimistic (no OCR software is perfect) is would take about 75 hours to do.
Will
Adobe CTI
It does not take anywhere near 5 mins per page. You simply scan page by page until you are done. Then use Acrobat Pro to do an automated OCR for the entire document. I would guess scanning would be ~20 seconds per page including turning the page and repositioning on scanner. The automated OCR might take an hour, but it does this automatically so i don't care. I have found the OCR to be pretty reliable if you have a good scan. For example I took the Fiero P22 manual and did this to it.
IP: Logged
07:13 AM
Oreif Member
Posts: 16460 From: Schaumburg, IL Registered: Jan 2000
It does not take anywhere near 5 mins. per page. You simply scan page by page until you are done. Then use Acrobat Pro to do an automated OCR for the entire document. I would guess scanning would be ~20 seconds per page including turning the page and repositioning on scanner. The automated OCR might take an hour, but it does this automatically so i don't care. I have found the OCR to be pretty reliable if you have a good scan. For example I took the Fiero P22 manual and did this to it.
I respectfully disagree. I assume you would be doing no type/suspect misspelling correction. My estimate is likely a little high as most people don't know how to use Acrobat in the most efficient manner but there is no way a 900 page document will take, according to your calculations (20 sec X 900 pages / 60 sec per minute + 1 hour for the OCR) 6 hours corrected. I think it is extremely unlikely. Even if one didn't make corrections I rather doubt it could be done in twice that time frame. I'm not saying it would be impossible, clearly it's not. I'm just saying that if you are going to make it searchable (typographically correct) it will take one heck of a lot longer than just compiling a bunch of scans into a multipage pdf.
By the way I am will do put this together given how much all of the people on this forum have helped me. I could never repay all of the time savings you folks have given! The knowledge base on here is amazing.
[This message has been edited by infinitewill (edited 03-31-2009).]
IP: Logged
08:05 AM
sjmaye Member
Posts: 2468 From: Hendersonville, TN USA Registered: Jun 2003
I am one of those people who do not know how to Acrobat to it's fullest potential. I can say I took the P22 manual which was already in pdf format ran I think was called paper capture(not sure as it has been a while) . After running I was now able to do searches. Sure it was not perfect. The bad scans on the p22 manual hurt recognitions, but it was worlds better than nothing at all. At least I could perform a search for a part.
The "real hands-on time" is scanning each page. That would be scanning, repositioning manual and scanning again. I just don't see it taking 5 minutes to do it. As for the paper capture/ocr . I don't care if it took overnight. I would just leave it running.
Forgetting exactly how much time it sure would be great to do a search.
I will take a look at the link above at some of the ones already completed. Thx.
IP: Logged
09:41 AM
BMTFIERO Member
Posts: 1187 From: Beaumont, TX Registered: Dec 2007
I am currently doing this for my '88 Service Manual. I am using Adobe Acrobat Pro 9 and a my HP Scanjet 7400 and I will say that it is taking longer than I expected. I believe the overall speed of the scan depends on the person and the system they are using
IP: Logged
12:52 PM
infinitewill Member
Posts: 2344 From: Overland Park, KS U.S.A. Registered: Mar 2009
I am one of those people who do not know how to Acrobat to it's fullest potential. I can say I took the P22 manual which was already in pdf format ran I think was called paper capture(not sure as it has been a while) . After running I was now able to do searches. Sure it was not perfect. The bad scans on the p22 manual hurt recognitions, but it was worlds better than nothing at all.
Ah, now I understand Yea, I guess I am a bit of a perfectionist. I used to work for ILM and was on the development team for "Display" which was later acquired by Adobe and renamed Photoshop. I have spent the last 20 years as a special effects photographer and product evangelist/trainer for Adobe Photoshop and now the Creative Suite. (CTI = Certified Training Instructor). When I think of scanning and converting manuals I generally think of getting things exact because dolts like me are often not sure what we are searching for so I want to make sure that EVERYTHING is correct
BTW, I am doing a Creative Suite seminar in Nashville on June 23rd, drop in and I will put you on the guest list! I really enjoy meeting the people on this board. Some amazing rides on here, most of you have forgotten more about Fieros then I will ever know about them!
Best regards,
Will
Adobe CTI
IP: Logged
01:49 PM
30+mpg Member
Posts: 4061 From: Russellville, AR Registered: Feb 2002
Ah, now I understand Yea, I guess I am a bit of a perfectionist. I used to work for ILM and was on the development team for "Display" which was later acquired by Adobe and renamed Photoshop. I have spent the last 20 years as a special effects photographer and product evangelist/trainer for Adobe Photoshop and now the Creative Suite. (CTI = Certified Training Instructor). When I think of scanning and converting manuals I generally think of getting things exact because dolts like me are often not sure what we are searching for so I want to make sure that EVERYTHING is correct
BTW, I am doing a Creative Suite seminar in Nashville on June 23rd, drop in and I will put you on the guest list! I really enjoy meeting the people on this board. Some amazing rides on here, most of you have forgotten more about Fieros then I will ever know about them!
Best regards,
Will
Adobe CTI
quote
Originally posted by infinitewill:
I already have one I was offering to convert it to a pdf for the good of the cause to make it searchable.
I am glad to know that I am not the only one who believes that "good enough" is not always "good enough". I for one would appreciate any pointers I am working with CS3 and I am loving it
------------------ White 88 GT 5 spd (Soon to be a Shelby S1 powered 6 spd) Red 88 GT Auto (Soon to be Shelby S1 powered)
Black 88 coupe 5 spd (For Sale)
IP: Logged
09:56 PM
infinitewill Member
Posts: 2344 From: Overland Park, KS U.S.A. Registered: Mar 2009
I am glad to know that I am not the only one who believes that "good enough" is not always "good enough". I for one would appreciate any pointers I am working with CS3 and I am loving it
I am teaching a Photoshop seminar in Houston on Apr 27 then over to Gulfport, Baton Rouge, Layfayette and ending up in Alexandria on May 1. PM me and I will get you new on the door list of any of them. I use mostly CS4 now and am beta testing CS5 and alpha testing CS6 (very flakey).
Will
IP: Logged
10:15 PM
rourke_87_T-Top Member
Posts: 1347 From: Toronto Ontario Canada Registered: Jan 2009
The P22 guide is not complete. It's the rear section with the exploded parts digrams and parts numbers list, which is good enough for looking up most parts.
There was someone on here a bit back that was selling an older copy of the GM parts manual that the dealers used. Works good and cross references to other models as well when you do a search.
Alldata also has an electronic version of the service manual on DVD as well - Mitchell On Demand is another one (I kind of like the AllData one a bit better for the older cars (pre-90))
[This message has been edited by Mickey_Moose (edited 04-01-2009).]
I am teaching a Photoshop seminar in Houston on Apr 27 then over to Gulfport, Baton Rouge, Layfayette and ending up in Alexandria on May 1. PM me and I will get you new on the door list of any of them. I use mostly CS4 now and am beta testing CS5 and alpha testing CS6 (very flakey).
Will
Update. My schedule has changed I will be in Wilmington, DE then Baltimore, Altoon, PA then Erie, PA and end up in Cincinnati the week of Apr 27th.
Will
IP: Logged
11:27 AM
infinitewill Member
Posts: 2344 From: Overland Park, KS U.S.A. Registered: Mar 2009