Setting Recognition Parameters

Applies to TestComplete 15.65, last modified on July 17, 2024

In the Set Recognition Parameters dialog, you specify recognition parameters for one or several images in the Image Repository.

To call the dialog, do any of the following:

  • Right-click the selected item(s) on the Items list on the left of the Image Set editor and select Set Recognition Parameters from the ensuing context menu.

— or —

  • Right-click the selected image(s) in the Image Strip panel of the Image Set editor and select Set Recognition Parameters from the ensuing context menu.

In the dialog, you can set the following parameters:

  • Color tolerance - Specifies an acceptable color difference at which TestComplete will treat two pixels as identical. The color difference is represented as an integer value within the 0…255 range that specifies an acceptable difference for each color component (red, green and blue) of the compared pixels. TestComplete considers two pixels identical if the difference in each color component does not exceed the specified value.

    If Color tolerance is 0 (the default value), TestComplete will consider two pixels identical only if they have the same color. When Color tolerance is 255, TestComplete will consider pixels of any color identical.

  • Pixel tolerance - Specifies the allowed number (as percentage) of dissimilar pixels. If the number of such pixels is less than or equal to Pixel tolerance, TestComplete considers the images to be identical. The default value is 0. This means that all pixels must coincide.

To set the parameter:

  1. Select the Apply check box next to the parameter you want to assign.

  2. Specify the needed value in the edit box.

The value will be assigned to all selected images in the Image Repository.

Setting the recognition parameters to high values may cause your tests to run slow.

To save the changes and close the dialog, click OK. To close the dialog discarding any changes, click Cancel.

See Also

Image Set Editor
Image Set Item Properties
Image-Based Testing

Highlight search results