Dear AppleVis members,
I wanted to let you know that we (Creaceed) just released a major new version of our scanning app Prizmo, now at version 5.
You can find the app description on the App Store, I'm adding the link at the end of this message.
I'll quickly go through what's new on the accessibility front in Prizmo 5, and if you have questions, feel free to ask.
1. New OCRs. Prizmo comes with two new OCRs, we call them on-device OCR and Cloud OCR, the same as in Prizmo Go. The former runs entirely on the device, while the latter requires an internet connection but typically provides better accuracy. A third one will be arriving soon in Prizmo 5 with iOS 13, as we are integrating Apple's new OCR (English only, for now), hopefully on day one. Prizmo's OCRs are now all based on machine learning, and are more robust to image variations than what was previously available.
2. Spoken guidance and spoken description. Prizmo still has spoken guidance that provides feedback (left, left, up…) while aiming at the document. Spoken description is new, similar to Prizmo Go, and it is available when VoiceOver is active on the iPhone or iPad through a secondary shooting button, on the left of the main one, shaped as an eye. When you tap that button, Prizmo will shoot, and will provide a summary of what it sees to help you reposition the device if needed.
- Prizmo tells you to rotate left or right by a few degrees (like 7 degrees right). This can help as horizontal text is typically better recognized by the OCR.
- It tells the area coverage of the page within the image as a percentage (like, page coverage is 62%). The higher the better typically, as this means document image will get more details and thus will be better recognized.
- It tells the number of lines of text that were detected, as well as the text area coverage as a percentage. This can help if you expect a lot of text (book), but the description tells there isn't much, you know you have to reshoot.
- Finally, it tells the angle of the device with the gravity (like 5° with gravity). This can help position the device parallel with the table or document, in which case angle with gravity becomes 0° with Prizmo's convention.
3. Raw text mode. When VoiceOver is active and you open a page of a document in Prizmo, then go to text mode (top), Prizmo will automatically enter in Raw text mode, and will maximize text display to fullscreen. This means all the text of the scanned document will be available in a continuous flow instead of separate column structure. This makes it easier to read it aloud using VoiceOver. You can end Raw text mode by tapping that button in the bottom toolbar (Raw Text button), in which case you'll return to the standard, per-column document structure. You can change this behavior globally in Prizmo in app settings.
4. Prizmo accessibility settings. In Prizmo app settings (settings button on the top left on the main screen), there is an accessibility section with a number of settings. The first two settings, Maximize Text Display and Raw Text Mode, allow to enable or disable globally the behavior I was describing in point 3. Next, the Spoken Guidance switch lets you disable interactive feedback while aiming, as some users find it disturbing or don't need it. Finally, the Vibrate when Level switch lets you get a vibration when the device is perfectly horizontal while aiming if you want to.
5. Prizmo still provides a full featured Text Reader that can read aloud. If you're scanning a single page, you may not need it because VoiceOver can read that text as described in point 3. But if you scan multiple pages (of a book or magazine), it will be more comfortable to send them to Prizmo's Text Reader, where you can choose the voice, the rate, and you get per sentence or per page navigation through buttons.
We tried to implement the most recurring user requests in this release. Of course, if you have remarks or ideas, please let us know. You can reach us at firstname.lastname@example.org