Computer vision results for different color backgrounds

Does computer vision scope results/activities give a difference in behavior when the background color of the target application changes?

The CV scope works in terms of identifying elements based on attributes it has already learnt. For example, text boxes based on the size and shape, labels based on the text position, place etc. So. theoretically, if the elements are similarly placed/located/distinguishable between different environments, the scope should be able to pick them up in a similar manner and if all the elements which were available in the lower environments are established successfully, the interaction would work in the same way. However, one will have to be careful in cases, where the presence of a different color on the screen, creates distortions in the elements being identified. The recommendation would be to follow the below steps to validate this between environments:

1. Run a sample job that simulates all the interactions between both the environments to confirm if the interaction works as expected between the environments.

2. Another way to confirm this is to create a project on Studio, having the computer vision scope, pointing it to the target application and validate if the elements are appearing successfully and similarly in both the environments. This can be validated by using the “Show Elements” option in any of the CV activities to display the identified elements as shown in the below screenshot. Also, additional details are also available here .