Docupen R-700 Rechargeable, Portable Scanner Review - OCR Font Size Test
(Page 6 of 7 )
OCR Font Size Test
Like I said earlier, I scanned a sheet with many font sizes to see how they turn out after OCR conversion. I want to know how small is too small for the OCR software to read.
This is the original text of the paper I scanned:
This is a test of the PLANon Docupen R-700 Rechargable and its included OCR software. This was printed using an inkjet printer. Scanning this would be similar to scanning an office memo or a legal document.
There are two OCR methods to export. Like I mentioned earlier, you can open the scan with Word or Notepad. Opening in Word will try to preserve the formatting. Opening in Notpad will, of course, strip all formatting. The software package has a lot of trouble determining what are new formats; it often mistakes scanning imperfections for new fonts or layout changes. However, the basic text is actually quite good. So, here is how Notepad displayed my 200dpi file (I bolded the mistakes):
200dpi OCR into Notepad
20 Point Font: This is a test of the PLANon Docupen
R-700 Rechargable and its included OCR software.
This was printed using an inkjet printer. Scanning
this would be similar to scanning an office memo or a
legal document.
iS Point Font: This is a test of the PLANon Docupen R-
700 Rechargable and its included OCR software. This was
printed us ing an inkjet printer. Scanning thiq would he
similar to scanning an office memo or a legal document.
16 Point Pont: This is a test of the PLANon Docupen R-700
Rechargable and its included OCR ~nftware. Thig wa~ printed
using an inkjet printer. Scanning this would be similar to scanning
an office memo or a legal document.
14 Point Font: This is a test of the PLANon Docupen R-700 Rechargable
and its included OCR software. This was printed using an inkjet printer.
Scanning this would be similar to scanning an office memo or a legal
document.
12 Point Pont: This is a test of the PLANon Docunen R-700 Rechargable and its included
OCR sofiware. This was printed using an inkjet printer. Scanning this would be similar to
scanning an office memo or a legal document.
11 Point Pont: This is a test of the PLANOn Docupen R-700 Rechargable and its included OCR
software. This was printed using an inkjet printer. Scanning this would be similar to scanning an
office memo or a legal document.
10 Point Font: This is a test of the PLANon Docupen R-700 Rechargable and its included OCR software.
This was printed using an inkjet printer. Scanning this would be similar to scanning an office raicrilo ot a
te8al document.
8 Point Font: Ilims isa test of the PLANon Docul~n K-IOU Kecharganle and its iuciutLcd OCR sott~aic Thi~ wa~ printud nun; an
inkiet printer. Scanning this would be similar to scanning an office memo or a legal document.
I am pretty please with how this turned out. The PaperPort software had a very good success rate with font sizes 10 and above. Font size 10 is kind of questionable here since you can't tell what some words are, and 10 point fonts turned more poorly in some other scans. Still, in 200dpi, the software does a pretty reliable job at reading 12 point and above.
PLANon recommends using the 200dpi setting if you plan to use the OCR software. Still, I'd like to see how 100dpi turns out. I am cutting some intermediate fonts so that there is less to look at. Once again, I am bolding the mistakes that occur:
100dpi into Notepad
16 Point Font This is a test of thc PLANon Docupen R-700
Rechargable and its included OCR software. This was printed
using an inkjct printcr. Scanning this would bc ~siiriilar to ~scai1riiiig
an office memo or a legal document.
12 Point Font: This is a test of the PLANon Docupen R-700 Rechargable and its included
OUR software, This was printed using an inkjet printerS Scanning this WOuld be similar to
ocunnins an office mcmo or a Icgal document,
10 Point Pont: This is a test of the P1.ANon Docupen R.700 Rechargable and it~ ~ ~Ll,;~ ~ ~ to sc.an,isn$ an.:.ffte £,,.~k1c. cr a
lugal document.
I guess 100dpi isn't recommended for a reason. It's not really surprising this didn't convert too well after seeing what 100dpi looked like as an image.
Next: Results >>
More Peripherals Articles
More By Developer Shed