Designing a Scannable Form

FormReturn OMR software needs a neatly designed form template, printed using a lazer printer, and quality, clear scanned form images to accurately recognize each page and accurately detect the barcodes and checkbox marks.

A scanned form page may look perfect to your naked eye, but may have imperfections when being read by FormReturn OMR Software. Imperfections in scanned barcodes and checkboxes can be seen by focusing with the zoom feature in the scanned image 'Preview'. Scanned form imperfections cause inaccuracy of OMR detection and Barcode Reading.


Scannable Form Template Design

As the form template is designed, make sure no black text, objects or dark lines are touching the barcodes, Fragments or Segment borders. You want optimal detection of these parts of the forms when they are being read by the form processor.

Give the Form ID Barcode room to expand. If it has been made smaller in any way detection could be broken. It should always be added to the left of the page and then moved over without touching the right hand side of the form.

The default margins of the segment and the form template allow space to ensure any of the components that must be detected by the OMR form processor, won't be obscured from the page if a scanned form image is skewed (out of line).

A quality document scanner with a good automatic document feeder, will decrease the risk of badly scanned form images, and make your captured response data accurate.

A poor quality scanner or incorrect settings, can cause more black or white to show up on the form image, which could cause errors with detection of the barcodes or location of the mark areas. To overcome the risk of low quality form images causing errors in detection, increasing the resolution or the contrast on the scanner's settings could make a big difference. Sometimes it can be a game of cat and mouse to get the scanner's settings correct so that quality images can be produced.

If the respondents will be using pencil to mark the checkboxes, this can cause errors in detection of marks. The luminance and mark threshold settings in FormReturn can be increased to overcome this.

Recognition Preview is used to detect the segment, barcodes and mark areas added to the form template, as you go and when you are finished. There should be no errors in detection of these components. Make sure there is detection of the segment border, the barcodes, the checkbox mark areas and Key Fields such as ID grids and Barcode areas. These are all used for detection and capturing information in the OMR form processing.

Test Preview the form before you publish it, to test the detection of a scanned image from your form template.


Recognition Preview Check

The Recognition Preview confirms, the added form components can be detected by the software.
When the segment and form template are being designed, text, dark lines and objects that are too close to the checkboxes or barcodes can cause breaks in detection. If there is a problem with detection of barcodes, fragments or the segment, go back to the segment or form editor where they were added, and move anything that could obscure the detection of any of these components.
The print preview will confirm the barcodes, the fragments and the segment can be detected by the software, bearing in mind scanned images of a forms can get skewed or black marks can appear where they shouldn't, depending on the scanner being used.

print preview

  • The Form ID Barcode is detected with a pink outline.
  • The segment barcodes are detected with a blue outline.
  • The Segment is Detected with a green outline.
  • The Mark Area Fragments are detected with a red outline. If there are no test marks made in the checkboxes, it will say Marked: none

Test A Scanned Form Image

Because all scanners are different, a test preview of the form template by printing a form and scanning it, tests your scanner's settings with FormReturn. The test preview will pre-determine how handmarks made in checkboxes will be detected by FormReturn. If there are errors in the detection of the hand-marks, you can change the Recognition Settings of FormReturn or your scanner's resolution and contrast settings and try the test preview again until everything on the form is detected accurately.

Most often if there is a problem with form recognition or detection of barcodes or fragments it is because:

The forms weren't scanned with the correct settings.

The form has been designed with something added that obscures the detection of barcodes or mark areas.

The Mark Threshold Setting isn't correct which could be caused by the medium used for shaded in check boxes or the quality of the scanned images, and will need to be re-set.

If the checkbox size and style or fragment padding has been changed from the FormReturn default size, the mark threshold might need to be changed.


Scanned Form Processing

FormReturn OMR Processing is fast and accurate but it can only detect filled in checkboxes from scanned form images it can recognize. Make sure the form template has been Test Previewed before being printed and distrubted to the respondents. A scanner dedicated for FormReturn will ensure the scanners settings remain correct, every time you are scanning forms for FormReturn.

FormReturn uses:

  1. Scanned images of forms that have been designed using the FormReturn template editors.
  2. Detection of the Form ID barcode on each scanned page image, to identify the FormReturn form page and the FormReturn publication it belongs to.
  3. Detection of the Segment barcodes on each scanned page image, for locating the area that contains the Mark Area Fragments and Key Field Fragments.
  4. Detection of Fragments and unbroken checkbox borders, to accurately detect and extract the dark marks made in checkboxes.

Form ID Barcode

In the Recognition Preview, Form ID Barcode is shown as detected with a pink outline. When the barcode is added, care should be taken so as to not squash the barcode, which makes it smaller. A smaller than default size barcode could cause an error with detection and form recognition by the software.

recognition barcode

*When adding the Form ID Barcode, make sure is has plenty of room for expansion. *

Keep the Form ID Barcode clear of all other elements added to the form, to allow for good detection and recognition of the barcode.

When scanned form images are uploaded to FormReturn, if the Form ID Barcode can't be recognized, the image will be rejected by FormReturn and put in the Unidentified Images folder in the Processing Queue. The scanned form images that FormReturn reads, must be clear enough to detect and recognize barcodes. It is really important that the Form ID Barcodes don't touch any other black text or lines on the page and are not squashed or made smaller. **
**

Adding Barcodes


Segment Barcodes

In the Recognition Preview the segment is shown as detected with a green outline and the segment barcodes are shown as detected with a blue outline. Make sure the segment barcodes are kept clear of any checkboxes text, dark lines or other objects added to the segment or the form template.

recognition segment barcode

A screenshot of the top segment barcode detected in the Recognition Preview. If one barcode can't be detected, neither of them will show as detected.

When scanned form images are uploaded to FormReturn, if the segment barcodes can't be detected, the OMR software won't be able to detect and located the mark areas, therefore the form won't be successfully processed into captured data. The reason why segment barcodes might not be detected could be because they are touching other black marks, like text or lines which have been added to the page. **
**


Mark Areas / Fragments

In the Recognition Preview, the Mark Areas are shown as Fragments detected with a red border. Each fragment is detected separately. When you add checkbox Mark areas, an ID Grid or a Barcode Area to the segment, they should all be detected sepatately as Fragments in the Recognition Preview, before you publish the form template.

When scanned form images are uploaded to FormReturn, if any mark area fragments can't be detected, the captured checkbox data will throw an error.