Form Recognition Testing
There are two parts to Form Recognition Testing, the Print Preview and the Test Preview. The reasons for recognition testing your template are:
- To make sure your scanned forms will be sharp enough to detect the Form ID Barcode, the Segment barcodes and the checkboxes in the Mark Areas.
- To make sure all response marks in checkboxes will be detected accurately.

The Print Preview and the Test Preview tabs in the Form editor.
IMPORTANT: If a form template has any
undetected components, you will need
to go back to the segment editor or form editor where they were added, and correct any errors.
Recognition - Print Preview:
The Print Preview checks that your form is properly designed before you do a Test Preview.
A Recognition Print Preview can be performed at any time during the designing of your segment or form template.
- Determines that the Barcodes, Segments and Fragments that have been
added to the form template can be detected.
- Check your text - Sometimes if the text area isn't big
enough for the amount of text you typed, some of the text will be
invisible. Just make the text area a bit bigger. It doesn't matter if the edges of text areas overlap with each other.
Step by Step Print Preview
- Click the Recogniton Preview tab.

- The Print Preview is automatically selected.
.png)
The Preview Check boxes are used to show if Barcodes, Segments, Fragments and Marks in Checkboxes are detected.
- Click the boxes in the Preview check Panel to show if the
barcodes (Form ID Barcode and segment barcodes), segment and Fragments
and Marks are detected by checking each box in the Preview Check panel.
- If you want to test how marks in checkboxes are detecting, go
back to the editor and using the line tool, place some marks in some of
the checkboxes. Don't forget to go back and delete the test marks
before you go ahead and publish the form.
Note: Make sure the Form ID Barcode, the two segment barcodes, the segment border and each of the fragments are detected separately.

Screenshot: Recognition Preview of a form template. The detected components are highlighted with colored borders. Possible causes of broken detection
- The Form ID Barcode detected with a pink border
- The segment
barcodes detected with a blue border
- The segment detected with a green
border
- The fragments detected with red borders

Detected Test Marks
Test Marks can be added to checkboxes, using the line tool (Note: Don't forget to delete the marks before adding the segment to the form template or publishing the form.)
Recognition - Test Preview
A Recognition Test Preview is performed when your form template is finished and ready to be published, to recognition test a scanned image of your form using your scanner. (If there is more than one page in your form template, each page should be scanned and viewed separately).
Note: A scanned image of the form template opened in a test preview is not processed into captured data because the template has not yet been published. You wouldn't publish a form template and use it until a test preview has proved that the forms printed from the template can be accurately detected by FormReturn.

The Recognition Preview tab bar in Form Editor.
Step by step Test Preview
- With the form template open in the Form Editor, click the Recognition Preview tab.
- Click the Print Button in the Recognition Print Preview control bar and print the form page from the template.

- Handmark the form yourself by shading in some of the checkboxes and scan it and save the scanned form image as a TIF, GIF, or Png file on your computer.
- Open the image in the Test Preview window by clicking the Browse Image button. Look for the image where it is saved and double click on the file or select it and click open.
- When the image has opened in the Test Preview window, check the quality
of of the scanned image. Look for the detection of all Barcodes and Fragments and make sure your handmarks are accurately detected.
Quality
scanners which produce clear images are an important factor of form processing using OMR software.

- Find the scanned form image on your computer, open it and wait for page image to load onto the test preview window.

Screenshot: Form image with test checkbox marks detected in the Test Preview. Possible causes of broken detection
- In the Preview Check Panel, click Detect Barcodes, Detect Segments, Detect Fragments, Detect Marks.
Broken detection of Barcodes or Fragments:
- Try setting a higher resolution or contrast on your scanner and re-scan the form to get a clearer image.
If any of the hand marks aren't accurately detected:
- Click on the Fragment to open the Mark Detection Details dialog and check the Mark Threshold %. If it is under 40%, that is the reason why the mark didn't get detected. This could be caused if the checkbox doesn't have a very big mark in it; by the Luminance Setting; or by the quality of the scanned image.
Click here>> For more information on Mark Threshold and Luminance in Recognition Settings.
Click here>> For more information on Scanning Forms.
Leaving Test Preview
When you are finished with the test preview the Clear Button can be used to clear the image from the application window, or just go back to the form editor.

Recognition Preview Controls

Zoom: To zoom in or out for full view of a particular area or the whole page.
Set Preview Field Data:
Print: - Prints current page
Save As PDF: - Saves current page as a PDF file.

Zoom Control:
No Image Loaded:

1 Bit View: is selected to convert color pixels to black or white
pixels, so you can see how the black pixels of a scanned image will be
counted.
Browse Image: - Browse for image saved on your computer to open in Test Preview.
Clear: - Clear Image.
Preview Check Panel
.png)
Detect Barcodes shows if all barcodes are detected on the segment or form.
Detect Segments shows if all segments have been detected separately.
Detect Fragments shows if all checkbox mark areas have been detected separately.
Detect Marks show if handmarks have been detected and marked correctly.
Recognition Settings Panel

On a color image, the Luminance setting determines the cut off point at which pixels will be detected as black or white when detecting handmarks in the mark area.
The Mark Threshold setting determines which is the most heavily shaded checkbox, by using the set percentage (default 20) for comparing the percentage of the pixel count in each checkbox against the percentage of the pixel count in the checkbox with the Lowest count.
Fragment Padding is the white buffer between the checkboxes and the fragment border. Increasing the fragment padding allows better detection of handmarks in the checkbox mark area.
Deskew Threshold is used to determine if a scanned image is skewed.
When the Perform Deskew box is checked, FormReturn will automatically perform a deskew on the scanned image.
The Restore Defaults button returns any changed settings back to the default settings.
The Apply Settings button will update any changed settings to take effect.
For more information see Recognition Settings
Possible Causes of Broken Detection
Problem - Form ID Barcode not detected.
Possible Causes
- Part of the Barcode is obscured.
- There are lines or text touching the barcode.
- The Barcode has been accidently squashed or made smaller.
Problem - Segments aren't detected
Possible Causes
- The segment barcodes are partly obscured or touching text, lines or checkboxes.
- One or more of the Mark Areas or a Key Field (OMR ID Grid) is touching text or dark lines.
- One or more of the Mark Areas or a Key Field (OMR ID Grid) is touching text or lines.
- One or more Mark Areas or a Key Field (OMR ID Grid) is to close to the segment border.
- Something added to the segment has been pushed partly outside the segment borders.
Problem - Fragments not Detected.
Possible Causes
- Mark areas have gone outside the borders of the segment.
- Black text or lines are touching some of the checkboxes.
- Black text or lines are touching the Segment Barcodes.
Problem - Hand marks aren't detected.
Possible Causes
Print Preview: Test Marks made using the line tool aren't heavy enough.
Test Preview:
- Scanned image isn't clear enough or the marks aren't big enough. (Check the Mark Threshold of the Marks)
- Luminance setting not set high enough, especially if pencil was used.
- Mark Threshold setting not set high enough.