OCR vs digital ink recognition

Before going any further, it is important to make the difference between digital ink and optical character recognitions.

Suppose you write something on a screen or a sheet of paper that you want to recognize. You, first, need to represent this ink so that it is suitable for computer processing. There are two common approaches for this: optical character and digital ink recognition.

Off-line recognition or Optical Character Recognition (OCR)

Scanning a written page with an optical scanner or a camera produces a bitmap image, that is to say an array of pixels. Each pixel represents a part of what was captured. OCR applications then use image analysis to extract text or shape information from this data. They rely on a visual representation of ink.

“hello” Bitmap image Recognition Scan orpicture

So they must segment handwriting or typeset from other information contained by the paper. This may lead OCR to confusion and errors.

On-line recognition or digital ink recognition

On-line recognition systems analyze 2D points sequences. Those points describe ink trajectories and are produced by various devices or systems, as:

  • Finger touch or stylus on a screen.

  • Digital pen and paper.

  • Vectorization of a bitmap (although this is pretty difficult to get right)

“hello” Digital ink Recognition Capture

Collected digital ink thus expresses the trajectory of handwritten strokes.

A stroke is the trajectory of a finger or stylus, from the moment it touches the writing area (pen down) until it goes upwards (pen up). It might include information about the pressure used to write as well.

Thus, a sequence of pen events models the stroke. The first event corresponds to a down event. The last one is a up event, and all other ones are move events. Each event also contains spatial and temporal information: (x, y, t, P) . x and y are the pen position, t the absolute time and P the pressure (optional) .

So digital ink recognition refers to a dynamic process. The method steps on where strokes start, where they end, and in which order they were drawn to perform character, text or shape recognition.

MyScript performs exclusively on-line recognition with digital ink as input

So if you want to recognize the content of an image or a scanned file, we cannot help.

MyScript does not capture the ink strokes either. This is the job of the hardware that produces and collects your digital ink. Interacting with these devices is out of MyScript SDKs scope too. It’s up to your application to extract strokes from the devices you plan to support and send them to MyScript APIs. For an easier integration, MyScript provides code samples on how to do so, though.

But if you’d like to have your notes strokes recognized and interpreted, you are at the right place. MyScript Interactive Ink is your best shot! So now let’s move forward and discover what interactive ink is!

We use cookies to ensure that we give you the best experience on our website Read the privacy policy