Which OCR package for text scanning is the best?
in [Scanners]
|
Prev: Scanning suddenly stopped working after Windows update
Next: Cannon Scangear issue - clip organiser
From: CSM1 on 1 Mar 2010 09:27 Jethro Pull <jpull(a)hotmail.com> wrote in news:RTNin.38804$K81.13522(a)newsfe18.iad: > Interesting ... no, I have found no spelling errors in the dictionary. > How would you correct them if you knew an obvious error or > accidentally added a misspelled word to the dictionary? > > On 2/28/2010 3:18 PM, isw wrote: >> In article<1kuin.53095$qJ.29950(a)newsfe10.iad>, >> Jethro Pull<jpull(a)hotmail.com> wrote: >> >> >>> As to ABBY vs. Omni Page: I have used both and prefer ABBY. Why? >>> ABBY is simpler to use and is accurate enough for me. When I import >>> into MS-Word, misspelled words are underlined and I can fix them. >>> >> Unless the error results in a different word also in the dictionary. >> >> Have you ever found spelling errors in the MS dictionary? I have. >> >> Isaac >> > > MS word has a custom Dictionary which you can edit in MS word. It is found at Tools > Options, on the Spelling & Grammer tab. Click the Dictionaries button to get to Edit. -- CSM1 http://www.carlmcmillan.com
From: spam on 1 Mar 2010 11:19 On Sun, 28 Feb 2010 17:36:05 +0100, "Edward Kroeze" <edwardkroeze(a)home.nl> wrote: >"CSM1" <nomail(a)nomoremail.com> wrote in message >news:Xns9D2C78F9917E5nomoremail(a)74.209.136.91... >[snip] >> >> No OCR is 100% accurate. 90% is doable. >> > >Just curious what exactly you mean by 90% ? > >1) 90% of the pages are flawless? (and 10% contain one or more mistakes?) >2) 90% of each line is correct? (and 10% (appr. 5 to 8 characters) PER LINE >are wrong?) >3) something different? > >The first option might be workable, but the second for sure is not. > >Cheers, > >Edward Different people define OCR accuracy in different ways. I prefer to go by characters misrecognised. I am amazed that anyone quotes 90%. They must have a truly useless scanner working on something they found in a muddy puddle. It is much much lower than anything I get. I have just looked at a simple three-column press cutting, chosen at random, scanned in B&W at 300 dpi. (I have a Canon Canoscan LiDE 600F.) And recognized with OmniPage Pro. The cutting is 257 words and has 1591 characters. It has one word misspelt. 'Natufe' rather than 'Nature'. One character awry in >1500 words is more like 99.9%, which is my general experience, confirmed by looking at another cutting. That one has no errors. Note, I am talking practice rather than theory. I neither know nor care about the subtleties of edge detection. MK
From: spam on 1 Mar 2010 11:21 On 28 Feb 2010 19:34:44 GMT, CSM1 <nomail(a)nomoremail.com> wrote: > >PDF Image on text works pretty good, you get the image that is the exact >copy of your sheet, and the OCR result for searching. You only miss the >words that don't OCR well. > >I have seen a 100% OCR result on really clear original copy of just >text. -----End Quoted (and cut) Message----- This reflects my own experience. MK
From: isw on 1 Mar 2010 12:38 > > Have you ever found spelling errors in the MS dictionary? I have. In article <RTNin.38804$K81.13522(a)newsfe18.iad>, Jethro Pull <jpull(a)hotmail.com> wrote: > Interesting ... no, I have found no spelling errors in the dictionary. > How would you correct them if you knew an obvious error In the dictionary provided with Word 6 for Mac, the word "continuous" also had the spelling "continuos". I managed to get the attention of a MS employee, who after telling me three times that the spelling was correct (because he persisted in using Word's dictionary to check it instead of Webster's!), he finally admitted that it was in error. There is evidently no way to actually correct the error by fixing the dictionary, but he told me how I could get the word flagged so I could manually correct it. Create an "exclude" dictionary, which tells the app to flag a word even if it is in the built-in dictionary, and then when you see that it is flagged, you correct it manually. A real hack, but that just makes the solution consistent with the rest of the application. Actually, an "exclude" dictionary can be useful: suppose you wanted to spell the word "color" when corresponding with folks in the US, but "colour" when the letter was going to England. By putting the word in the "exclude" list, it would always be brought to your attention. > or accidentally added a misspelled word to the dictionary? > Words you add do not go into the main dictionary, but into an auxiliary "custom" one which you can edit. Isaac
From: isw on 1 Mar 2010 12:41
In article <Xns9D2E5611CD20Bnomoremail(a)74.209.136.89>, CSM1 <nomail(a)nomoremail.com> wrote: > Jethro Pull <jpull(a)hotmail.com> wrote in > news:RTNin.38804$K81.13522(a)newsfe18.iad: > > > Interesting ... no, I have found no spelling errors in the dictionary. > > How would you correct them if you knew an obvious error or > > accidentally added a misspelled word to the dictionary? > > > > On 2/28/2010 3:18 PM, isw wrote: > >> In article<1kuin.53095$qJ.29950(a)newsfe10.iad>, > >> Jethro Pull<jpull(a)hotmail.com> wrote: > >> > >> > >>> As to ABBY vs. Omni Page: I have used both and prefer ABBY. Why? > >>> ABBY is simpler to use and is accurate enough for me. When I import > >>> into MS-Word, misspelled words are underlined and I can fix them. > >>> > >> Unless the error results in a different word also in the dictionary. > >> > >> Have you ever found spelling errors in the MS dictionary? I have. > >> > MS word has a custom Dictionary which you can edit in MS word. > It is found at Tools > Options, on the Spelling & Grammer tab. > Click the Dictionaries button to get to Edit. Words in the custom dict do not (or did not with W6 for Mac) override an incorrect spelling in the main dictionary; you can't correct it that way. Basically, if a spelling is found in the main dict, the custom one is not searched. Isaac |