Prizmo
3.7.2
Prizmo
3.7.2Scan and OCR documents with your digital camera or scanner.
Overview
Prizmo 3 is a scanning application with optical character recognition (OCR) in over 40 languages with powerful editing capability, text-to-speech, and iCloud support.
No matter what capture device is at hand, Prizmo is the key for scanning and performing OCR. It works with pictures taken with your iPhone, iPad, or digital camera; with documents coming from connected or Wi-Fi scanners; even with screenshots! It will help you easily scan any documents (invoices, receipts, boards, giant advertisements...) of any sizes (US Letter, US Legal, A4, and custom)...
Features:
- Specific processing for business cards, multipage texts and images
- Input file formats: regular photo files including JPEG, RAW, TIFF, as well as PDF files (even multipage)
- Input options: digital cameras (iPhone, iPad, DSLRs...), camera tethering, connected and Wi-fi scanners, screenshots, snap any text from your screen
- State-of-the-art OCR technology in 40 Languages: Afrikaans, Albanian, Basque, Belarusian, Bulgarian, Catalan, Croatian, Czech, Danish, Dutch, English, Esperanto, Estonian, Finnish, French, Galician, German, Hebrew, Hungarian, Icelandic, Indonesian, Italian, Korean, Latvian, Lithuanian, Macedonian, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Serbian, Slovak, Slovenian, Spanish, Swahili, Swedish, Turkish, Ukrainian, and Vietnamese.
- Detection of 3 types of regions: texts, images and numbers
- OCR editing and correction: add, delete, merge and reorder regions on the one hand, edit text for correcting mistakes on the other hand
- Automatic frame detection: lay the page, snap it, and let Prizmo handle the cropping
- Automatic text orientation detection
- Perspective correction to compensate camera point-of-view
- Non-linear lens distortion correction
- Page curvature correction for books and magazines
- Real-time OCR results: moving/resizing region frames immediately updates text results
- Resolution increase: image enhancement allows OCR results where OCR is generally not even thinkable
- Output formats: image files (JPEG, PNG, TIFF), and text files (PDF with Spotlight-searchable text, RTF, TXT) - even multipage!
- Export to the Evernote, Box, Dropbox, Google Drive and WebDAV online services
- Online translation into 23 languages
- Accessibility feedback for the OCR processing (blank, blurry image, etc.)
- Text-to-Speech for reading scanned texts: over 70 high-quality voices in 26 languages with navigation settings and interactive reading progress display (available for free in OS X Preferences)
- Enhanced for the Retina display.
What's new in Prizmo
Version 3.7.2:
Fixed:
- A crash that could occur when moving or resizing a text region
- A crash when using the Snap Screen tool when multiple displays are attached to the Mac
- The page cropping user interface that could become corrupt when selecting another page
Join over 500,000 subscribers.
Subscribe for our newsletter with best Mac offers from MacUpdate.
Prizmo Screenshots
$49.99
App requirements:
- Intel 64
- OS X 10.10.0 or later
Category:
Developer Website:
Install with MacUpdateDownloaded & Installed 89,531 times
App of the week
50 Prizmo Reviews
See allRate this app:
Most helpful
The new version (3.5) seems more accurate when recognizing text in conditions that are less optimal, like when trying to do OCR from photos in lower resolution where text is more pixelated and less clear... it seems to make fewer errors in those cases when compared to the previous version... the new version also seems to recognize more words correctly when dealing with some "calligraphic" fonts where letters can have more "decorations" (for example, the "Old English" and "Old German" font styles... although some characters in some of those font styles can still have recognition issues, especially some of the uppercase characters with more "decorations", which usually make letters more difficult to recognize for OCR apps... but it seems better than it was)...
That new OCR that the new version (Prizmo 3.5) is using seems better for things that were more difficult for the old version to do...
But the previous version (Prizmo 3.1.18) seems to be more accurate when the conditions for OCR are more optimal, like when recognizing text from high-quality Retina screenshots with very clear text that uses some very common font in larger sizes and where the background is just a plain color without some "noise"... or if you have scanned some page in high DPI and the font is very readable... or something like that...
The new version isn't really "bad" at recognizing text from Retina screenshots or high DPI scans, but I've noticed multiple times that it sometimes made some small mistakes when doing an OCR from a high-resolution image in places where the previous version didn't seem to get "confused" with the same image and the same font... like confusing a lowercase "i" in some words with "1" in a text that uses the Georgia font with 18px font size or something like that (and if you take a screenshot and zoom in, since it's a Retina screenshot, that lowercase "i" has about 28px in height on the actual screenshot, out of which 5px is just the dot, with about 4px between the dot and rest of the letter)... and confusing the lowercase "i" for "1" would maybe be expected when dealing with a lower-resolution image or if the image was noisy and it wasn't clear that the dot on the letter "i" was there (the rest of the letter does look similar to "1" in Georgia font, although "1" in Georgia font would still be somewhat taller and wider than "i" without a dot, but on a low-resolution image that may not be very clear), but with a 18px font size on a Retina screenshot it is very clear that it's a letter "i" and not "1" and the previous version didn't seem to get confused with it in places where the new version confused it with "1"... with some other fonts without "serifs", I've also noticed that the new version confused a lowercase "l" (lowercase L) in some words with a lowercase "i" or something like that (which would also be expected if the resolution was low, but I'm writing about text on Retina screenshots, that was clear enough for the previous version to identify same letters correctly)... it doesn't seem like a huge drawback, but I've just seen cases where the previous version recognized all words correctly and when trying to do OCR of the same text with the new version, a few small corrections were needed.
So, for the kind of images that already had very high OCR accuracy when using the previous version, the previous version still seems better for those images.
I hope that there would be some update soon to make the new version as optimized for larger font sizes as the previous version is... and that they would also bring back OCR support for Hebrew in that update.