Image and Document Quality Control

IFP deskews, registers, and cleans up each image. It also performs quality control checks to ensure image integrity. If an image is skewed, has shifted too much, or is too light or dark, the image is flagged and sent to an exception process. IFP also verifies that the document is complete and cohesive.

Form Dropout

For each form type, form dropout removes the preprinted form from the image, leaving only the filled in data. Removing lines, captions, and instructions not only improves ICR accuracy but also produces an image that requires less storage space and less bandwidth to transmit across your network. Form dropout can also repair gaps where characters cross lines, and can detect whiteout areas on the form, such as labels and alterations. IFP can also separate documents via a barcode or 2D barcode on the primary form.

Optical Character Recognition/Intelligent Character Recognition (OCR/ICR) recognizes machine printed or hand printed data respectively. Using OCR/ICR, IFP reads the variable data from the form and:

• Performs preliminary validation of the data using confidences and context
   
  
• Performs table and database lookups
   
  
• Checks that all required forms are included in the document
   
  
• Prepares image files for archival storage
   
  
• Stores captured data in the database for subsequent application processing

To achieve optimal results, IFP lets you select one or more ICR engines to capture handprint, machine print, and OCR-A data. IFP supports IBM and other commercial ICR engines. Noise filtering, image enhancement, and other parameters improve the results on messy forms you need to scan.

Depending upon your unique requirements, IFP uses other character engines, such as Oce’ Technologies’ RecoStar Professional, or Mitek’s Quickstrokes to improve data recognition accuracy.

Context Checking

After ICR, Forms Processing can apply numerous context checks to improve and reformat results. For instance, many ICR engines know that amount fields can contain decimals and commas, but do not enforce their syntax. And large commas are often read as 1’s. IFP context checking corrects these and other characters for noticeably improved results.

Data Validation

After ICR and context checking, IFP validates the captured data against predefined edits and against user-provided dictionaries, tables, and business rules. For instance, you can verify that an account number is valid or that a tax amount has been correctly captured.

Statistics Collection

IFP collects and outputs recognition statistics so that you can monitor performance. You can easily see how quickly and accurately IFP is capturing data and be alerted to any changes in performance.

Remittance Processing

IFP can also capture both MICR and courtesy amount data from checks. IFP uses Mitek’s Quickstrokes engine for check recognition and courtesy amount and legal amount recognition.