PlanetPress Capture ICR

The term "ICR", which means "Intelligent Character Recognition" is an evolution on the popularly-known "OCR", which is "Optical Character Recognition". The difference between the two is easily explained: While OCR can only recognize characters using the finished shape (for example, in scanned documents and pictures), ICR relies on much more data which is provided by the Anoto Digital Pen: the path that the pen takes, the exact timing of this path, start and stop points, etc. This extra information boosts the recognition rates of characters by a wide margin.

It's important to note that both OCR and ICR are relatively loose terms - that is to say, they can have different specific meanings depending on the technology used, but in their general sense mean the same thing. When using the term ICR, we use the above definition for convenience.

Note: The PlanetPress Capture ICR engine is only available with PReS Workflow version 7.5 and higher.

An ICR Workflow

The ICR engine in PReS Workflow is used in conjunction with PlanetPress Capture, translating the ink from the Anoto Digital Pen into separate characters, or text, that is readable by the suite. Multiple components are required in order for the ICR to work:

  • In PlanetPress Design, a Capture Field Object must be added and the Perform ICR option must be activated (See the Capture Field page in the PlanetPress Design User Guide). This must be either a Multi-Area Field or a Text Field.
  • The Capture Fields Processor must have the Perform ICR Recognition option checked, and language needs to be selected.
  • Once the ICR data is available, do something with it. This is done by reading the ICR data that is available in the metadata generated by the Get Capture Document task.
  • The metadata is also readable by the Capture Condition task, including the captured text and the reliability of this text.

The Workflow as such is the following:

  • A Capture field is setup for ICR in a PlanetPress Design document.
  • The document is sent to PReS Production
  • The Capture Field Generator is used to produce one or more print-outs using this document.
  • The physical sheets are written on using an Anoto Digital Pen
  • The pen is docked and the data is sent to PReS Production
  • The pen data goes through the Capture Field Processor, where the Capture Field ink is sent through the ICR engine.
  • The captured ICR data is retrieved with the document using the Get Capture Document task.
  • Conditions are applied if necessary with the Capture Condition task.

Caution: ICR, just like OCR, has its limitations. Please refer to the PlanetPress Capture ICR Best Practices page for more information.

Terminology and Definitions

In regards to our ICR technology specifically, the following terminology applies:

  • ICR: "Intelligent Character Recognition", or the engine that will read the pen data and attempt to recognize the text written using the pen itself. The ICR engine uses the path of the pen, its movement speed as well as the overall shape of each character to determine which character was written.
  • ICR Value: The alpha, numeric or alphanumeric value that was determined by the ICR engine.
  • ICR Confidence: A percentage value that the ICR engine gives to any specific value, when comparing the pen data with it's character database.
  • ICR Resemblance: A percentage value that defines how closely the value resembles the "average" character shape. Both the Confidence and Resemblance can be used together to make an informed decision on the contents received by ICR.