With the release of version 1.43.0 of the Bipsync Notes iOS app (available for download here) we introduced a new useful, and exciting feature – the Document Scanner.
Getting reports, articles, receipts, meeting minutes, business cards, or any other documents into your Bipsync notes has never been easier. The scanned documents are captured as note attachments, and then filtered through Apple’s new VisionKit and Machine Learning frameworks to extract texts via Optical Character Recognition (OCR), forming sentences and paragraphs from scanned documents.
I’ll be walking you through the technologies, as well as explaining how to use this feature, and share some tips and tricks along the way.
What is OCR exactly?
Optical Character Recognition is a technology used to recognize text inside images, such as scanned documents or photos. OCR can take any kind of image containing typed, handwritten, or printed texts and re-produce machine-readable text.
Where do I find it?
Users can access the Document Scanner in one of two ways.
- From a note list by long-pressing the new note icon and selecting “Scan”.
- From the note editor by tapping the ellipsis menu button and selecting “Scan Documents”.
After hitting “Scan Documents,” the device’s camera will open and the app will ask you to point it at the document you’d like to digitize.
Pay attention to the row of icons in the camera view, because the default settings will capture documents in color. If you’d prefer grayscale or black and white, just tap the icon of three circles to change it. If you change your mind later you can also adjust the document once it has been scanned/captured.
We also given you two options: “Auto” and “Manual”. In Auto mode, simply focus on the document for a few seconds and it’ll automatically be captured. If you want to capture manually, tap the shutter button or one of the volume buttons. Once the scan completes, the camera view returns giving you the chance to scan additional documents.
Some Tips and Tricks
The app does a fantastic job of figuring out the edges of a scanned document. It will automatically correct the perspective and straighten everything out. So if you’re in a situation where you don’t have time to square things up, or you snapped a picture of your document from an angle, it will work just fine.
However, to guarantee best results, it’s important to try and position your document as well as possible. Below I have outlined some tips to keep in mind:
- Sufficient Lighting: Avoid light flares or distortion that could potentially make the texts difficult to detect.
- Avoid folds/creases: The app has been optimized to detect paragraphs, but it’s inevitable that you may notice some misalignments or mix-ups in your OCR results when documents are curved or creased. Try to keep the documents as flat as possible with little to no creases.
- Simple formatting: Formatting such as bullet points, multiple columns, many line breaks, large font chars at the beginning of paragraphs, etc. may cause some misalignments or mix-ups as these features are not explicitly supported. Unexpected behavior can be avoided if you position your device’s camera over a smaller area rather than the whole document, to make the subject as simple as possible.
Here are some examples of what we’d consider good and poor documents to scan:
The document on the left is considered the ideal candidate to form perfect paragraphing, as it is relatively flat with proportional text alignments and simple formatting. The document on the right however, is creased, poorly positioned and not flat.
Once all your documents are scanned and you are happy with how they look, hit “Save” in the bottom right corner to perform OCR. You will then be taken to a view where you can swipe through the scanned documents, investigate issues, and edit the OCR texts by tapping on the card view interface.
Once you are happy with the OCR texts simply hit “Save” in the top right corner. All the extracted texts are then consolidated within a new note or added to an existing note, depending on how you launched the scanner earlier.
The app not only stores the OCR texts, but also saves the scans of your documents as attachments on the note.
We’ve identified a few improvements to make in the near future, but we’d love to hear any feedback you may have about additional requirements or ideas, so we can factor them into future updates.
The Document Scanner is now available in Bipsync Notes iOS so try it out and if you have any questions, as always please get in touch.